+++ plato iv group notes +++

general interest notes 1975

notes beginning dec 1, 1975

file onotes41 printed at 10:00 am on june 12, 1976

---------- note 0 demo
  12/01     00.13     cppy       research

      A reminder that a demonstration of a speech
recognition system for PLATO will be given at 3:00 pm on
Tuesday, December 2 in Room 351 of CERL.

      Further details can be found in general note ¬$173.

---------- note 4 X-blank
  12/01     05.32     bonnie       matha

Would it be possible for lesson "search" (the option off the
block page) to show all lines with blanks in the command
field that follow a line _beginning_ with the searched-for
object?

For example, I was X-searching someone else's lesson for
-zero- commands. The author used:
zero    var,5
        var2,3
in the lesson, but X-search did not show the second line.

I never realized til now what I may have been missing!
Thankx!

---------- note 5 datai/o
  12/01     05.48     john r       reading

The new -datain- and -dataout- have been sorely needed.
at least by the reading project. Thank you, Mr. Blomme.

---------- response 1
  12/01     07.43     keith s       matha

Thanks much, Rick. This will cut elmath ECS use by
10K words at peak load!

---------- response 2
  12/01     08.09     david       hebrew

Thank you, rick. It makes TUTOR more elegant by far.

---------- response 3
  12/01     09.15     berger       mfl





and saves us much storage/common/student variable
manipulation and interaction.

Thanks!

---------- note 16 more touch
  12/01     10.10     judy     pso

I would like to continue the -touch-discussion ....
Please read Gen. Notes 425 and 445 for background.

Topic One:
I am against having -touch- numbered from 1-256 because
1) there are already two screen-location grids to remember
2) whereas both coarse-grid "1217" and fine-grid 200,200
   have physical meaning which I can picture in my mind,
   -touch 149- (for 149↑ˆth↑ˇ square) has no meaning to me--
   I cannot readily convert numbers modulo 16.
3) -touch- locations are almost always coordinated with
   text and/or line drawings, as in
TOUCH HERE
TO CONTINUE
The -touch- address should be easy to relate to the
text address.

---------- response 1
  12/01     10.17     judy       pso

Topic Two:
I would like to see the present one- and two- tag -touch-
commands dropped. I feel that most people use the three-tag
form and that many users do not understand how the one-
and two- tag forms work.

Then, (necessarily after a suitable "waiting period")
I would like to see -touch- changed so that a one-tag
form would be equivalent to: -touch location,1,1-

Topic Three:
A -touch- initiated "help" and "helpop" would be very nice
indeed. But rather than Gary's suggested format, wouldn't
you need
thelpop location,unit $$ to specify touch position
or
thelpop location,xtolerance,ytolerance;unit

And how would you write it when the "unit" needed to be
conditional?

---------- response 2
  12/01     10.51     michael       english

I agree with Judy's observations on the 1,2 and 3 argument
forms of the current -touch- command. It must be kept in
mind that the current -touch- is a judging command, not a
definable function key as I am proposing. In line with the
current treatment of function keys, a touch-based student
branching key should set a pointer to a specific unit when





the command is encountered, as does -help: this is the
reason that argumented function key jumps are not allowed.
A possible format for such a command might be:

thelp condition;unit1,201,2,1;unit2,(601,1,1;801,4,1)

In this case, if "condition" is true, the touch unit pointer
is set to "unit1"; a later touch at 201,2,1 will cause a
help-type jump to it. If it were false, the pointed to unit
would be "unit2". which would be reached by a touch at 601,
1,1 OR 801,4,1.

CONTINUED

---------- response 3
  12/01     10.53     silver       ve

An alternative would be a command which set certain touch
locations equivalent to a key:
touchk help,1205,5,1;data,505,1,1
etc. leaving present branching commands unchanged.
(in format). Would this work?

---------- response 4
  12/01     11.07     michael       english

I realize that the proposed format allows three levels of
separators, but they are needed -- and many of the answer
judging commands provide a precedent for it.

Two problems arise:

1) Touch areas may conflict. The execution of a touch-
jump should involve a left-to-right scan of the subarguments
of the command, and the first area that includes the just
touched coordinate satisfies the search. The -keytype-
command is analogous to this.

2) A -thelp- may be in effect at an -arrow- which also has
-touch- commands below it. The search for a branching touch
key should occur first, then the arrow can be processed
normally if no t-jump was executed.

Lastly, the above procedure could be easily extended to
allow the commands to operate on defined touch groups.

---------- response 5
  12/01     11.10     michael       english

PS: John Silver's idea seems quite reasonable also.

---------- response 6
  12/01     11.21     larry north       p

I am working on establishing a link between touch input and
control key branching. Any suggestions about this problem
are welcome, since the exact implementation has not been





determined yet.

---------- response 7
  12/01     11.37     jp       af1

what about???
I like the touch grid numbered from 1 thru 16ˆ2.
Problems with using the present grid system include:
touch is counted from the bottom left corner like fe
fine grid.
Using course grid coordinates adds undue complexity
to using it in the same way,,so,,users have to learn
another set of grid components anyway.(sorry judy)
   What about the following:

Pause  touch=101,1,2;data;324,1;5,1,1;help;touchno
OR
Pause  touch=101,1,2¬,data¬,324,1¬,5,1,1¬,help¬,touchno

---------- response 8
  12/01     11.45     jan       park

*concerning the touch grid...the exercise of learning
an entirely unique grid just for touch is seemingly
unnecessary. as with memorizing multiplication tables...
once one has tussled with the touch locations and gotten
the picture firmly stuck in the brain, a new system is
not very attractive. (unless some super benefits can
be reaped by installing a new system)

*the use of 1,2 or 3 argument -touch- statements is a
similar example of usage breeding familiarity...
i for one, heavily use the 2 and the 3 argument touches
and the differences between them are significant enuf
to spur me to ask judy, why she would favor getting
rid of them...in favor of some other locus touch system.
most of the touch routines done here, are in relation to
diagram that have structures to be identified that some-
times don't conveniently land on text lines...thus the
2 and 3 argument touches are used for more flexi

---------- response 9
  12/01     11.49     jan       park

cont....
2 and 3 argument touches are used for more flexible
definition of the touchable areas.

*i do hope that the branching can be accomplished, it would
greatly help future programming in touch routines.

---------- response 10
  12/01     11.58     b sherwood       s

Judy was suggesting that the one- and two-argument forms
be dropped, and the three-argument form emphasized. She
didn't ask to get rid of the the three-argument form.





---------- response 11
  12/01     12.15     barr       econp

touch location=key
I would prefere a command which sets touch locations equal
to keys,the command would be in effect untill cleared.
All touches would the have the keycode of the key they are
equal to. No special commands would be needed for
for pauses, help units,etc.

For example.

unit   test
tkeys  location1=help,location2=a
helpop helpu
arrow  1010
long   1
answer a
...

Pressing location1 or the help key would cause unit
helpu to be done. Pressing location2 would couse a
to be written at the arrow and judged ok.

---------- response 12
  12/01     12.40     stan smith       chem

Discussion of the form of touch commands needs to consider
the very different character of the student interaction
with touch compared to keyset imput. For example, when
you ask the student to touch something or point at something
an arrow plotted on the screen has little meaning.
Hence code contains inhibit arrow. Also, ok and no
are not as usefull for touch and we always write specs
nookno. If the student makes an error, it is nice to
have a message of some kind which goes away when he does
the right thing. But, the automatic -no- answer stuff
is of little value because you tend to just touch the
right answer after an error instead of pressing NEXT.
So -no- is followed by judge ignore and some author
generated code to do what is done automatically for keyset
inputs. A structure which makes it easier to generate
student screen touch interactions is more important than
branching.

---------- response 13
  12/01     13.08     judy       pso

Stan... What you are wanting is something like
-force firsterase-, which would work "reasonably"
for -touch-es? Me, too!

Jan: No, I definitely do not propose dropping the 3-tag
-touch-! The problem which arises is that the 2-tag
touch encompasses an area AROUND the tag given, so that
touch 1233,2 is FOUR squares wide and FOUR squares tall
but
touch 1233,2,2 is only TWO squares by TWO squares.





This seems confusing to me, and the single extra tag not
an excessive burden to bear in order to make it less
confusing.

---------- response 14
  12/01     14.28     chabay       s

I agree that using coarse grid to specify touch locations
is extremely cumbersome. I also agree that in many cases
where the default options active at an arrow are not
necessary, it would be useful not to have to inhibit the
arrow, turn off ok/no, judge ignore, etc. For cases like
this, weVe discussed something like:

pause   keys=touch
keytype (tkey),touch(0,4),touch(1,5)

where the touch coordinates are given in touch grid.
Possible additions would include allowing touch
areas (more than one touch square) to be specified in the
-group- command.

I agree that a crucial question is how easy it is to
position text at a appropriate screen locations -- I think
that simply having the touch grid available in SD takes
care of some of this.

---------- response 15
  12/01     16.06     silver       ve

I would rather stick with the arrow and branching machinery.
After much -inhibit-ing of arrows, I put them back on the
screen again, because they still come in handy!
There are still 200+ terminals which will not have touch
panels in the foreseeable future. Also -- sometimes touch
panels don't work (not often, but it happens). One of the
biggest complaints about a particular set of diagnosis pro-
grams was that no keyboard equivalents were given for touch
commands. So I number diagrams and stuff like that --
Finagle will get you, if you don't. There are situations
where not having a touch panel takes all the fun out of
the lesson, but this is a rarer situation than I used to
think.

---------- response 16
  12/01     23.51     blomme       s

   Since it is conceivable that the "fineness" of
the touch equipment might sometime change, putting in
some other grid based upon current capabilities is
just asking for trouble (as well as the confusion that
will be generated by yet another co-ordinate scheme). We
want the specifications for touch areas somehow linked
to current ways of specifying display position.

---------- response 17
  12/02     09.20     bruce iehl       mtc





Aw, booo... You don't want to create another coordinate
system, but you are already _mixing_ them in the -touch-
command -- you don't refer to the width or height of the
area in terms of course grid unit! Why can't we approach
the use of the touch panel as if we are interested in using
the touch panel for it's own sake, not just as an added
option to existing lessons? I am assuming the panel will
be checked out before entering the lesson-- and that those
who need them will have them. Also I am assuming that
the problem of remembering coordinate locations for those
who are doing standard types of things will be handled
as it is now -- with the computer generating the code from
a Service type routine.

---------- response 18
  12/02     11.06     jan       park

perhaps what we can do is consider what is already
available in the touch machinery...e.g.
use a combination of what the "key" variable receives
as a value when a location is "touched" ...and
also using that as a tag for the -touch- command

thus: -touch- 511 (would render the top right corner
                   active, and then branching via
                   key=511 would be useful immediately)

at first glance this useage has at least one drawback...
the variable sizes of the touch square will have to be
generated some way, since one touch square alone is
hard to hit by most fingers!

to see "key" values...lesson "tess" has an illustration
that may help

---------- response 19
  12/02     13.32     j predmore       mtc

have you ever tried judging some 2 square touch areas
which are randomly generated, and giving feedback on
the incorrect responses?
   That is really a bucket of worms while using the
keycodes...There seems to be a hangup with trying to
use a touch area and give it the designation of a
keycode. If touch doesn't get easier, it will not
be used and the pannels will rot away.
   We should think of the touch pannel as an alternate
input method into the computer, not a slave of the
keyboard...It is very useful because it is an alternative
to spelling and typing.It is not useful when it is more
difficult than the keyset...for students and authors alike.

---------- response 20
  12/02     15.44     jp       af1





I want to have all but a small specific portion of the
touch panel disabled so that I get no beeps or anything.
Will a more selective enable likely to be implemented
in the future or will I have to work around the plato
disability by using the unhandy keycodes?

---------- response 21
  12/02     16.22     michael       english

Since the beep is a hardware function that takes places
whenever an enabled panel is touched, it would seem im-
possible to do what you want.

---------- response 22
  12/03     00.30     john r       reading

and also bad practice: a non-beep could mean (1) touching
a non-enabled area _or_ (2) touching an enabled area without
breaking a beam pair _or_ (3) a broken touch panel.

---------- response 23
  12/03     08.53     j predmore       mtc

Why does a touch input have to break through a pause
command?(blank tag)

---------- response 24
  12/03     09.30     friedman       csa

ANYTHING breaks through a -pause- with a blank tag.↓
Unless you really want to process ALL keys (including,
for example, TERM), you should put an appropriate tag
on your -pause-; then you can decide (in the tag) whether
you want touch, or whatever, to break the pause.↓
Blank-tag -pause- should be thought of as an exceptional
use of -pause-, not the normal use.

---------- note 18  party
  12/01     10.25     celia       pso

Plans for the PLATO Christmas Party have been finalized.
The cost is $1.75 per person. The time, 8:30 pm, Friday,
December 19. Place, Levis Faculty Center, 2nd floor.
Please pay Sheila (room 258 ERL) by Tues, Dec. 16.
Remoters, please tell Sheila by Dec 16 if you're coming,
and pay us at the party.

Everyone is invited. The Mighty Plato Art Players have
planned some delightful entertainment; there will be
goodies to eat, and a cash bar.

---------- note 20  site/cours
  12/01     10.48     hinton       ssu





Both the recent note about site directors in Colorado and
the problems noted last week about a stolen (I presume)
signon which resulted in a number of very offensive notes
prompt me to ask

        Could there be a lesson giving signons/names/course
of Site Directors and ≤/or Course Directors ?

   I know that Tebby supplies this information willingly
when asked, but I don't see the need to send her a pnote
every time there is reason to contact a Site and/or a
Course Director. "Network" gives mailing addresses and
phone ¬$s of Site (but not Course) Directors, but surely it
would be easier to contact them direct ≤≤≤ ly through pnotes.
        And it would save time lost by contacting whoever
is listed in Users under that course name, which requires
all sorts of back-and-forth-ing.

ndh

---------- response 1
  12/01     11.05     avner       s

The major problem is that such names are under control of
the course directors (and hence subject to rapid change).
The churning of course names is wonderous to behold in
itself. Thus, in the absence of a routine that gallops
through all courses to compile an up-to-date list at each
request, it is easier for Tebby to serve as an information
resource.

---------- response 2
  12/01     11.59     b sherwood       s

Hmmmm.....but surely the most responsible people (site
directors) don't change their sign-ons periodically?

---------- response 3
  12/01     12.40     avner       s

Unfortunately, the most responsible (site directors) are
not always the ones that people want to contact. There are
many one-semester courses that have in the past tended to
become pirates after the real students were done. Some of
these changed course directors on a daily basis.
They have also been used at sites which have nothing to
do with the original use of the course.
In addition, a geographical site (we have about 200 of them)
often has several people responsible for different things.
Some of these people (believe it or not) don't talk to each
other. Tebby is aware of who should be seen, who is on
sabbatical, etc. or can track this information down in a
specific case. To maintain the same information on all
sites would take more of her time than could be justified.

---------- response 4
  12/01     12.59     lombardo       ed

That is too bad because when the problem arises it would





be nice to have someone to leave a p/note for that would
handle the problem as soon as he/she returns. I realize
Tebby will take care of it asap, but the responsibility
really shouldn't be on her shoulders for everything that
goes wrong. All things considered she does an unbelievable
amount of work!

---------- response 5
  12/01     13.56     dowling       ames

        This problem would be solved for courses by having
the information page available without inspect code. A
similar change for lessons would let one know the name of
the author. I presume that a similar info page is provided
for a site. though not being a site director, I do not know.
It may be necessary to change the info pages slightly to
provide the signon name as well as the full name and address
of the individual. The responsibility of keeping this up to
date would fall where it should - on the director.
        I would have no objection to a shortened info page.
Security code info, for example, should not be needed, nor
references to other lessons (router, instructor file, data
file, use reference, etc.). The "last editied by" info
would be useful to contact an alternate person if the
director were unavailable. ¬↑ˇW¬≤↑ˇC¬≤↑ˇD

---------- note 28 6500 cpu?
  12/01     11.43     mike b       cornell

        I've heard some rumors concerning the capabilities
of the new CPU which PLATO is getting in January. Specific-
ally, I understand that the new CPU will be a 6500, rather
than a 6600, and that the 6500 does not have the hardware
to do the -move- command. Therefore, the machinery will have
to be simulated by software, with a subsequent degradation
in service. Is it true that the new CPU will be a 6500,
and, if so, just what will the effect be?

M¬ ike≤≤≤≤ˇB

---------- response 1
  12/01     12.05     golden       s

There was never any plan/hope for 6600! That machine is
not well suited to PLATO. The question was a cyber 73 or
a 6500. They are very nearly identical, but yes there does
have to be a slight change in the way move is done. I think
the change has already been made. It will not be noticeable
to users.

---------- response 2
  12/01     13.22     frankel       p

The real problem is that the 6500 is a different color from
our Cyber 73-24...

---------- response 3





  12/01     13.53     berger       mfl

Gee, dave... IBM makes the 370 in red, yellow, and blue.
Why don't we switch to an IBM machine?

---------- response 4
  12/01     14.04     oberpriller       arizona

you mean the cyber 73 that runs plato ISN'T
ORANGE-AND-BLACK?????

gasp!

---------- response 5
  12/01     14.56     dave fuller       uimc

No, blue plastic and woodgrain contact paper.

Very tacky, actually...

---------- response 6
  12/01      15.27      hinton        ssu

We should offer spray cans of paint instead of cookies
henceforth until the new machine is color-compatible.
ndh

---------- response 7
  12/01      15.32      k mast        p

Cookies will still do nicely.

---------- response 8
  12/01      19.26      fay        o

Gee, Dave, you mean it won't match your baby blues?

(Or, are they baby browns?)

---------- response 9
  12/01      22.23      dave fuller        uimc

        Brown... but cookies are nice. Maybe we ought
to tile the thing with plasma panels...

---------- response 10





  12/02     11.02     holle       uimatha

That option is totally unacceptable, dave...
If you tried to use these plasma panels, you could
  be charged with plotting against the system.
If not, the new processor would be subjected to
  conditions from one ex-screen to another.

---------- note 47 12/20-1/10
  12/01     14.35     golden       s

Reminder: PLATO service will be unavailable from noon on
December 20 until 8:00 am on January 10. If it becomes
possible to resume service before the 10th, we will, but
cannot promise that now.

---------- response 1
  12/01     14.53     hody       med

does anyone know if it is possible for remote users to
put their costly dedicated lines on a "holiday" rate
or to suspend them altogether for that period? that
might be enough dough to bake another piece of hardware
for plato or another staff member or two??

---------- response 2
  12/01     22.30     ostuni       iu

Even though I'm just an anybody on this system...there is
something (as refered to in last response) that intrigues
me at times. And that would be: just what is the cost of
transmission from CERL to a far away land like arizona?
How about an answer in the form of cost per terminal per
month.
Needless to say...but anyway...
THANKS...As I remain,
A.John O'Tuna

---------- response 3
  12/01     22.42     golden       s

Costs vary, but a good rule of thumb is $1.00 per mile per
month. One line can carry signals for four terminals. So,
$0.25 per terminal per mile per month. Yes, it is expensive

Costs here in Urbana-Champaign are only about $2.75 per
terminal per month via microwave.

---------- response 4
  12/02     08.03     stephanidis       iumed

Is this expensive or cheap? I think to have many

terminals would be cheap to operate at these prices

today. But doesn't the terminal itself cost a pretty





penny?

$Interested≤≤≤≤≤≤≤≤≤≤ Interested

---------- note 48 edit bug
  12/01     14.45     patrick       mathw

Execution error in lesson edit.

        unit - preset
        state - regular
        last command - stoload
        2nd command
        error type was blank
        Join sequence      start   0
                           ieu     13
                           vocabret 2

I got this error while trying to return from my lesson
after editing a vocab block. This error is repeatable.

---------- response 1
  12/01     15.26     midden       p

Fixed, I hope.

---------- note 56  audio
  12/01     16.04     jim       pso

the audio device description written by john risken has
been added to AIDS. For access enter AIDS, press DATA, and
enter:

audio device

thanks john ---- this is a needed addition to AIDS.

---------- note 59  -eraseu-
  12/01     17.47     agg       iowa

2 requests involving the -eraseu- command that might make
it more useful:

1)  Could the unit specified be done in mode erase by
        default?
2)  Could normal erasing be automatically inhibited if
        there is an -eraseu- in effect?

I have needed these changes repeatedly and have had to
resort to all sorts of weird code to simulate it.

Have others?

---------- response 1





  12/01     19.21     jim       pso

no to both ----- i would have to change many existing
lessons.

these can easily be done when desireed by adding a -mode-
command for case 1) and by adding a -write- (no tag) for
case 2).

---------- response 2
  12/01     21.26     agg       iowa

I did not intend to discuss present lessons. I intended
to present a more consitant form of the command.
If my suggestions will yeild more consitancy to TUTOR
then that is what I would like comment on.

---------- response 3
  12/02     08.35     judy       pso

mode erase
¬+
write (no tag)

are not "wierd code"! Can you be more specific about the
"wierd code" you required?

---------- response 4
  12/02     13.52     agg       iowa

  Excuse the use of the word "weird". "Redundant" is
more like it. Every instance that I have used the
-eraseu- command has required that I use a mode erase,
write blank tag or inhibit anserase. Perhaps I miss
the point of the -eraseu- command and perhaps I see its
function more clearly than most others. Is a mode erase
and write blank tag used with virtually every application
of -eraseu- or not?
  If so, there is useless redundancy in TUTOR lessons. If
not, the -eraseu- command has applications I have been
unable to see myself. Please advise.

---------- response 5
  12/02     16.01     b parrello       med

-eraseu- enables one to gain control between the time the
user breaks the -no- pause and the time judging is begun
again.

---------- response 6
  12/02     20.33     david       hebrew

If you look at unit $ersu in block drive2 of
lesson hdriver, you will see just how important that
"getting control" is. In fact, it is the only thing
that makes the Hebrew arrow work. I do, however,





agree with you that most of the stuff done in an
erase unit is either in mode erase, or consists of
-erase- commands.

---------- response 7
  12/03     21.31     judy       pso

The only -eraseu-s I have used have been block erases
which clear a section of the screen:
erase 15,10
This is a lot faster than re-drawing my cute little
pictures in -mode erase-.

---------- note 68  security
  12/01     23.49     rader       s

Sometime in October, at the time when there was an error
that permitted instructors to edit any course, someone
managed to get hold of system level passwords (using,
reputedly, the -readr- command ??), and has in fact used
some system records since that time. All system level
passwords/ relevant codewords have now been changed; we
would recommend that course directors change the codes
on their courses and authors change their passwords and
codewords.

We have not seen any evidence of malicious use of inform-
ation thus acquired, but feel that users should be aware
of the situation.

---------- response 1
  12/02     02.13     sellers       arizona

That there was no apparent malicious use of this situation
speaks well of the plato community. The lack of malicious
use of this information also speaks well of the people
who designed this system. Is it not true that if the
system wasn't worthy of respect that those who had the
chance would have been more likely to use their knowledge
for distructive purposes?

---------- response 2
  12/02     09.48     berger       mfl

Well, John... the problem is that the person DID ues
this information, even if not maliciously, for a long
time.

---------- response 3
  12/02     14.33     volpe       mfl

Why not have a 'last looked by ¬0 ¬1.' or last edited
by ¬0 ¬1.' Just like normal everyday PLATO lessons?????
     tom volpe

---------- response 4
  12/02     15.00     michael       english





Huh?

---------- response 5
  12/03     02.45     stone       pfw

hahahahahah.....

---------- response 6
  12/04     10.28     ken       cornell

It seems to me that the lack of malicious use of this in-
formation speaks well of no one other than the person who
stole it. On the other hand that someone would discover
such an error and not report it IS disturbing and I wonder
just what the systems people have done to determine the
identity of this non-malicious plato user. (After all
shouldn't he receive some sort of award for his virtue?)

---------- response 7
  12/04     13.46     mike b       cornell

        The last respondent should re-read the note. The
error has been corrected and apparently was corrected quite
some time ago. The note did _not_ say that Systems was not
informed of the error that permitted this breach of
security. Apparently the only information that was with-
held from the systems programmers was that fact that
security codes had been gotten.

M¬ ike≤≤≤≤vB

---------- note 75 seminar
  12/02     09.39     whansen       csa

ACSES Seminar
R.Doring - The LEKTOR Portable CAI system.
115 DCL 1 PM Tuesday Dec 2 (Urbana)

---------- note 77 bio index
  12/02     10.05     thrasher       iu

A general query?
  Some companions of mine (biology major type) were
wondering if there is a general bio index and
if so how do you access it.
Chem major who doesn't know the answer.

---------- response 1
  12/02     10.50     hagstrom       iuchem

Try 'bioindex' from the author page.





---------- response 2
  12/02     15.30     jordan       english

The multiple sign-on "biology" of course "biology also
will access the bioindex if you don't have author records.
for additional info contact boggs/biocc or herrick/biocc
who are coordinating the community college demonstation
and field test of plato in the biology area at malcolm x,
kennedy-king, and wright colleges in chicago. they are
both located at cerl in urbana--phone 333-4405.
pauline jordan

---------- note 78 -sum-
  12/02     10.39     keith s       matha

What kind of CPU advantage would result from a system
implementation of a -sum- command with a format similar to
the following?

sum var1,¬$ of vars,result var,opt'l step size,opt'l mask

Example:

sum n1,150,n2,2,077

would sum the last 6 bits of the odd-numbered n variables
from n1 to n149 and place the result in n2.

The savings I see lie perhaps in the elimination of bounds
checks that must be done every time a -calc- is executed
in a -doto- loop. Correct? Any others? It seems this
application occurs often enough to warrant some discussion.

---------- response 1
  12/02     11.57     tenczar       s

Have you looked at the -Sum(A)- function in the calc array
machinery? Go to AIDS and type in "sum"

---------- response 2





  12/02     11.58     b sherwood       s

Some of these features, and others, are available thru
the "define array" procedures. It is hoped that eventually
you would be able to handle bytes as well as whole words
with that machinery.

---------- response 3
  12/02     14.36     keith s       matha

I am aware of the array machinery, but the array Sum
function does not seem as general as might be desirable,
e.g., as far as I can tell does not allow summation of
every nth element of a vector, where n>1. This might
be useful when dealing with y-word packages, where the
xth word of each package is a quantity to be used in a
summation.

---------- note 79 oldsysrout
  12/02     10.47     golden       s

Reminder:
          The old system router will no longer be available
after the present academic term. Be sure to make
preparations to switch to the new system router, "mrouter".

---------- note 89 phone bill
  12/02     14.52     golden       s

It seems as though we can save you some money on phone
lines by canceling them while we are down later this month.

It is not clear if it will apply within Illinois nor just
what the savings will be; but it looks good.

The one hitch is, if we cancel, you will be disconnected
until January 10 even if we come up several days earlier.

We will tell more as negotiations with A. T. T. are
concluded.

---------- response 1
  12/02     16.44     golden       s

Ok, we are going to cut the phone lines. Let us know if
you want your line kept active.

---------- response 2
  12/03     01.18     hody       med

about how much money (total) is involved?

---------- response 3
  12/03     12.03     dr weller       mtc





do you think this might cause line error problems?
remote sites often seem to get "bad" phone lines for
a period of time, resulting in unacceptable red light
problems. sometimes it requires much negotiation with
ma bell, including re-routing lines, changing equipment
along the line , etc. before the problems are solved.
unless the remote sites get exactly the same routing as
before the cutoff, i would expect several sites to have
to go through this line cleanup process starting the
tenth of jan. not that i think that avoidance of such
problems would be worth the cost of keeping the lines up.
i was just wondering if this is a potential problem?????

---------- response 4
  12/03     18.03     hofstetter       unidel

If there is a possibility of messing up the phone lines
again, I would much rather pay the extra phone bill than
go through weeks of red error lights.

I am sure you will take this into account in deciding
whether to go ahead and disconnect, and I will go along
with what you decide.

---------- response 5
  12/04     07.53     aikin       cornell

I agree with others who have mentioned problems with
phone lines as a possible result of cancelling the
lines duration the down period to save money. I hope
that in your negotiations with MABEL you have
considered this. Further, since the cost of the line
can vary depending on the routing used (i.e. how
much milage gets GSA retes, how many terminations, etc.)
what gaurentee can you get from ATT that line charges
on a monthly basis will not increase when the lines
are reinstalled? As one of the PLATO sites with a
high phone bill, I'm certainly interested in saving
money...BUT I don't want to see interference with
service or higher monthly bills later...

---------- response 6
  12/04     10.45     straayer       telcomc

     AT+T would not be the cause of higher rates, but GSA
(Federal General Services Administration) might be, as they
allow sharing of their lines by the state govt's with the
Feds getting priority, GSA routing incurrs the chance
of getting "bumped" in any case, temporary disconnect or
not. GSA routing is roulette at all times. Lines routed
via GSA are routed that way because experience has shown
that those certain routes have been bumped less in the
past.

---------- note 97 network?
  12/02     17.16     m stecyk       uni





Just a litle question here: When was the last time lesson
"network" was updated? It doesn't seem to me it is quite
up to date.

PLAT¬t

---------- response 1
  12/02     17.21     tebby       pso

"network" is now the responsibility of Jack Greve since
George Myers has left. Address complaints and questions
to "greve of o". "network" was last edited on 11/28/75
which is last Friday. We try to keep it as accurate as
possible. Are you sure you have later information than we
do? Sometimes people think sites exist that do not.
Tebby Lyman

---------- note 100 touchw
  12/02     22.27     larry north       p

The touchno command is now the touchw command.

---------- note 101 the editor
  12/02     22.32     midden       p

The communication system between various parts of the
editor has been changed. (i.e. no two argument jumpouts)
Please report any problems that arise with the editor"
Marshall Midden

---------- note 105 consult
  12/03     08.34     judy       pso

I have the uneasy feeling that people are getting lost
from the "consult" list. Sometimes I get a HELP request,
go and look at the list, and nobody is there. It happened
just a few minutes ago.

If you think you have been forgotten by the consult people,
could you send us a pnote? We try to cover from about
8:30 am to 6:00pm. Our worst performance record seems
to be between 11:30 and 1:00--lunchtime blues, I guess.

---------- note 108 xsearch
  12/03     10.00     errol       kka

Maybe it's my terminal, but I cannot seem to use the
Xsearch from the block page by pressing "X."

---------- response 1
  12/03     10.03     keith s       matha

No, it's not your terminal. The -jumpout- between the
editor and the search lesson is operational right now.





---------- response 2
  12/03     12.29     midden       p

Fixed soon, hopefully...
M.Midden

---------- note 109 iu pnotes
  12/03     10.01     warner       iu

   All personal notes to courses iu and iumed will
be destroyed early in the morning of Sunday, December 7.
   This will be done in order to enlarge the note-file
capacity of courses iu and iumed.
   Please use the SHIFT-S option on any pnotes you
wish to save, and place them in your lesson spaces
before Sunday.

---------- response 1
  12/05     08.53     mont       csa

this might better be advertised through a course message
in the appropriate courses.

---------- note 111 lost set
  12/03     10.31     marty       mfl

Have just noticed that the "Chanukah" display makes
me lose my character set. Is this normal? Did not
seem to happen with the other traditional displays
(Thanksgiving, etc.)

---------- response 1
  12/03     10.32     fritz       ames

Anytime new chars a loaded into the terminal, the old
ones are overwritten, and thus lost. Besides, whenever
you sign off, your charset flag is cleared and you will
have to reload your charset anyway. The computer can't
know what your terminal has been doing since you were
signed in -- for all it knows, you may have done something
like loaded in a mencrah or something!

---------- response 2
  12/03     11.54     marty       mfl

You don't have to reload your charset if you are
checking lessons and sign in and out repeatedly.
Am only saying that today -whenever the display is
shown- I must wait for the reloading to take place.

This is particularly annoying when one is giving
a demonstration using several lessons (all with
the same charset). Having to wait for the charset
to reload does interfere with the demonstration.
This is a complaint I have had before; the displays
ought not to slow down the sign-in procedure or





interfere with one's normal working habits.

---------- response 3
  12/03     12.41     mike b       cornell

        The only way you can avoid reloading your charset
after signing off and back on is if the lesson you enter has
the STOP key active at the charload (or if there is no loop-
back to the load on zreturn>-1). Otherwise, since signing
off _does_ clear the charset pointer, you would be forced to
reload. However, the above assumes that no new characters
are loaded during the period you are signed off. In that
case, although the STOP key will allow you to bypass the
charload, your charset will be screwed up. The reason the
Thanksgiving and Holloween (and Mickey's Birthday) displays
don't screw up your charset is that they are line-drawn
figures, not characters.

M¬ ike≤≤≤≤ˇB

---------- response 4
  12/03     13.48     tenczar       s

yes...your complaint is understood to be a real one...we
will try to figure a way around destroying the charset in
the terminal with the "special effects" display on the
WELCOME TO PLATO page.

---------- response 5
  12/03     14.20     judy       pso

One thing that CAN be easily done in to put the special
display characters way,way at the end of the charset
instead of into a,b,c,d as the Chanukah set is stored.

_Very_ few people use the Not linked" characters. If they
were overloaded from the Welcome page, it would not matter
to people who want to sign in-and-out. For larger displays,
the font-access-shift-letters (14 of them) are not very
popular, since they are rather a pain to type.

---------- response 6
  12/03     17.59     elston       rhrc

Then too you might get some one to do it in a lineset.
Although the loading time would be about the same.

2c≤/

---------- response 7
  12/03     08.00     b sherwood       s

But....why is it necessary to sign all the way off the
system in order to look at another lesson?

---------- response 8





  12/03     21.33     judy       pso

He may be entering through student records....

---------- response 9
  12/03     22.33     warner       iu

Anything can happen when the terminal is signed off the
system. The terminal can be turned off, or signed on
somewhere else, or even changed to another port --
losing charset information in any case. When the terminal
is signed on, at least its condition can be usually
sensed.
        (Note: signing out of a _lesson_ to the
AUTHOR MODE page should not destroy the charset info.
Only going all the way back to Press NEXT to Begin
should clear the charset.

---------- response 10
  12/04     17.23     broadus       css

I decided to try and make up a solution to the blank
screen problem before the WELCOME TO PLATO page while
the menorah (I hope it is spelled right) is in use.
This temporarily resides in "logoproc," There is no
inspect code, so you can see my stupid way of doing
it. If it is faster, or even the same speed as loading
the characters at first, and you like the method, tell
your favorite system programmer you would like to see
it in the next version.

Thanks,
Bob Broadus

p,s. Notice that it only uses one character. I thought
very few lessons use the arrow any more so currently
that is what I am using. It could be changed to
any character. This takes care of the ruined charset
problem too.

---------- response 11
  12/06     20.28     dowling       ames

        I second the suggestion of broadus of css. Using
only one character would reduce the nuisance of the reload.
        I would suggest, however, that the one character be
character number 0, the space, and that after the display,
one additional char command, -char 0,0,0,0,0,0,0,0,0- be
executed to restore this character to its normal empty self.
Few charsets override this char., and so the nuisance would
be minimized, Furthermore. I would presume that on loading
a charset, the chars are normally loaded in order. Thus if
the loading were terminated by pressing STOP, this char.
would nearly always be loaded.
        There should be no need for the inhibit charclear,
for the flag is already cleared by the signoff - signin.
        I would further suggest this technique on the
exploding not in P-notes.





thanks. ¬↑ˇW¬≤↑ˇC¬≤↑ˇD

---------- note 112 notes err
  12/03     10.51     gravel       chema

There seems to be a small repeatable error in the old
general notes file.

Enter old general notes (option f), and press 'a' for notes
after 11/19/75. Go to note ¬$348. Press LAB, and the first
response appears; press LAB a second time, and second
response appears. Fine. Press LAB a third time to go on
to the next response (or note) and you are bounced out to
the note index. There it says the note ¬$348, of 11/25
has three responses. However, when you attempt to access
the third reponse, you are bounced out. This also happens
when you attempt to access the third response by pressing
'3' from the note itself.

If there is no third response, or if it is not available,
then it seems that pressing LAB should take one onto the
next note.

.....Phil
---------- response 1
  12/03     12.30     woolley       p

Fixed.

---------- note 113 data-pg
  12/03     10.51     buz       phar

Speaking of security codes and the like, it would seem
rather logical to be able to access the data page of a
lesson even if you dont have the inspect code. Not all
authors leave their names and/or sinons in the lesson,
and if you should want to see the lesson, this way you
could at least find out who to contact.
Another 'Ʃ' for this is that if someone gets into your
lesson and changes the security codes, you could find
out "who dun it".

*buz*

---------- response 1
  12/03     11.21     jim g       reading

Please with a promise of cookies/pizza/beer?!?

---------- response 2
  12/03     12.44     mike b       cornell

Perhaps you should be able to see the "last editor"
of a lesson, But I think that the rest of the DATA page
information should remain secret on INSPECT-inhibited.





After all, there is information on that page which is
private, such as telephone numbers, addresses, and
department affiliation.

---------- response 3
  12/03     13.11     hanoka       ed

   I agree with mike, such information should
   remain to be seen only if you have the inspect
   or change code.

   mike

---------- response 4
  12/03     13.14     dr stutz       mtc

But what good is that data if no one can read it?

Surely, if someone knows the inspect code, they are
likely to know how to get in touch with whoever wrote
the lesson.

It's like denying access to a telephone book to every-
one except those who already know the number they want!

---------- response 5
  12/03     13.22     fy       o

The data is good to the system people for updating
lessons like "catalog", "authors", etc.

Any person registered in "s", "o", "p", or "pso" can get
the signon info for you.

---------- response 6
  12/03     13.27     jim g       reading

The essentials would be:

last editted by
owners name
owners phone number

I would propose that this info be displayed by typing the
lesson name and then pressing -HELP-. If someone in a policy
making position feels this is an invasion of privacy would
they please be kind enough to explain here.

---------- response 7
  12/03     17.25     berger       mfl

Before getting into this again, please note that it
was discussed in great detail about a year and a
half ago.

It seemed at the time that most people were in favor





of such information being made available without
codewords.

I, for one, am still in favor of it - it is easy to
keep "sensitive" information off my data pages, and
I like to make it fairly easy to people to comment
to be about my lessons.

Please consider making this change.

---------- response 8
  12/04     09.39     dowling       ames

One more vote in favor.

I would add to jim g's list the owner's signon and course.
Then a person could contact the owner (who is not always
the same as the last editor) through P-notes.

I would favor a modified information page. The user without
the inspect code probably has no need to know which security
codes are "none - open to all", or the name of the use
reference.

Perhaps those who need more security could be given the
privilege of restricting the display of this info, in which
case (and only in this case), the s, o, p, or pso people
could be asked to help, as at present.
There are certainly more important things for these
valuable people to do than serve as information operators
for those who have no phone directory. ¬↑ˇW¬≤↑ˇC¬≤↑ˇD

---------- response 9
  12/05     08.34     mailman       ustaf

        Another vote in favor, with perhaps giving authors
the option of restricting the information.
       At the _very_ least, someone without the inspect code
should be able to find out the signon name and course of the
author without having to bug cerl staff for it. I'm not
sure I understand the need to keep real names, phone numbers
and department affiliations "private" (what are people
worried about? crank calls and threats to kidnap their
lessons?), but if any author has a need to do so (like say if
he/she has an unlisted number anyway) the author should be
given the option.

---------- response 10
  12/05     20.06     warrens       uw

I definitely agree with jim g's suggestion, using a third
access key(next and lab being the other two).

and I do mean THE ACCESS KEY!

anyone could then get the info by typing in the lesson name,
followed by the access key.

Another possible key might be the ANS key for the answer





to one's problems, these mnemonics keys may help.
Thereason that I don't go for the help key is that it
is in use for a help sequence at present.

---------- note 115 partial +-
  12/03     10.59     zark       cornell

Suggestion Department

   Would it be possible to implement the 'Ʃ' and '-' keys
when doing a shift-P on the directory page? It would save
a lot of hassle for those with many part lessons.

paul zarnowski

---------- response 1
  12/03     12.45     mike b       cornell

I have often wished for this feature. Could we
please have it? Cookies? Pizzas?? Illicit payoffs???

---------- response 2
  12/03     14.21     dave fuller       uimc

I think Frankel goes for Coors now...

---------- response 3
  12/03     18.00     frankel       p

Yeah, but I'm going to Colorado this Xmas anyway, so I can
get some myself. Thanks anyway.

---------- response 4
  12/03     21.15     little t       research

Well while we are on the subject, how about shortening
that worthless paragraph that gets plotted when P is
pressed. I find many times I press P then letters
of the blocks then NEXT and STOP1 to condense only to
find that the letters got missed due the lengthy message
that is currently written. Same goes for HELP1.
Todd

---------- response 5
  12/03     21.20     dave fuller       uimc

        Can't be blamed for trying, Dave... In regards
to Little's excellent point, mebbe the demon HELL-POP
could be used to advantage?

---------- response 6
  12/04     09.45     dowling       ames

One more vote in favor.





 Would it also be possible to make the STOP1 active to
condense the lesson from the P option page? Usually the
partial flags are set just prior to a condense, and this
would save the return to the main block index.

It's only one keypress saved, but one that is usually
needed. I am sure many users would find this more
convenient.

Thanks. ¬↑ˇW¬≤↑ˇC¬≤↑ˇD

---------- response 7
  12/04     12.10     iezek       ames

A vote in favor of dowling's suggestion. That keypress
is annoying and there are few times that I use the P option
when I am not immediately going to condense.
M. Iezek

---------- response 8
  12/04     13.37     dave       infe

     Coors Co. unfair to Pipefitters!
  Boycott Coors!

     Borden's unfair to organized labor!
  Boycott Bordens!

     Alberto-Culver kills strikers!
   Boycott Alberto-Culver!

     California growers kill farmworkers!
   Boycott lettuce, boycott grapes, boycott the wines that Gallo makes!

     Spanish government kills anarchists!
   Boycott Spain!

---------- response 9
  12/04     17.25     broadus       css

I thought california got deleted.

Bob Broadus

---------- note 124 CVs
  12/03     12.59     rikel       med

  We are presently updating curriculum vitaes in our
department and several plato authors here (USC) would like
to include plato lessons they have written. Could those of
you who have included your lessons on curriculum vitaes.
please give me a few examples?

Jim Rikel

---------- response 1
  12/03     17.56     b sherwood       s





I just list the titles of major lessons I have written,
under the heading "PLATO lessons". And of course I do
mean "titles", not file names, which would be gibberish.
If the vita is sent to a place where there is no PLATO,
you'll have to explain somehow what a PLATO lesson is.
What problems do you find in referencing such lessons?

---------- note 125 plato up?
  12/03     13.04     ken       conn

   One question about the recent note(s) on
phone line disconnection....

   Just what are the possibilities of PLATO
coming up early after the Dec. 20 shutdown??

   I would think that this would be a factor in
deciding on whether phone lines for certain sites
[especially really remote ones(like us!)].

---------- response 1
  12/03     13.28     rader       s

The earliest conceivable date for software testing after
the machine is turned off is Jan 7th. It is not likely
that absolutely all the 100 different tasks to be completed
will be done on time, however.

---------- response 2
  12/03     13.43     ken       conn

Thanks...

---------- note 135 IDisze
  12/03     15.12     stratton       pfw

I just wish to say that among the many additions to lesson
editing that have been made, a few facts have been over-
looked. for example, i am very fond of using the rdraw in
ID mode. But, the other day I used the ID for rdrawing a
rahter large figure, and it started going off the screen.
So, I inserted a size command in the SD mode thinking that
after the SD switched over to the ID mode, the size would
allow additions to the drawing in proportion to the rest.
I found that I was wrong. It sure would be nice to allow
the last size command to be used as a magnification factor
for the x-,y-coordinates in the ID mode. Don't you agree?
Sincerely Yours,
John L. Stratton

---------- response 1
  12/03     15.22     rick       mfl





It sounds like you want something similar to -rdraw-.
SD/ID is capable of generating and displaying -rdraw-s,
which are drawings which CAN be sized. Go into AIDS
and type DATA ¬D rdraw.

---------- response 2
  12/03     15.59     stratton       pfw

Nope. that's not what I mean.....
I meant to say that after rdrawing in one size, a person
cannot change size and continue adding to the drawing in
the proper proportion.(i.e.,draw half of a large drawing
in size 0, rdraw mode. Then size it down, look at it in
the SD mode. Then, when it automatically flips over to
the ID mode, try adding to the drawing in the new size.
It won't be in proprotion to the first part when looked
at again in SD mode. What I meant to ask was, ""can a
change be made in the ID part of the editor so to use
the size as a magnifier of the coordinates placed in the
program to rdraw the figure?!

---------- response 3
  12/03     16.12     midden       p

Yes, I know about this. The new SD (presently being written)
has this feature. (and don't think I didn't argue with
dave about this either! I guess you could say I won.)
M. Midden

---------- response 4
  12/03     18.02     frankel       p

I don't remember that argument -- I only explained why I
did it the way I did (which is clearly sometimes confusing)!

---------- note 137 audio use
  12/03     15.44     gilpin       peer

The audio device now appears (see gen note 57) to be usable
on a routine basis. The purpose of this note is to suggest
that this fact has significance far beyond the special uses
where an audio is absolutely necessary (foreign languages,
beginning reading, etc.). When/if the audio becomes common,
I believe it can transfore PLATO as a teaching medium. It
can do this by relieving the learner's eyes of some textual
information (eg, directions) so that he can attend more
effectively to other screen info of to a practical situa-
ion (eg, in a lab or shop). Though cost will restrict the
availability of audios for some time, I think exploratory
work might profitably begin now. The possibilities are
wide ranging, including, as Marty Siegel pointed out to me,
the training of authors and instructors. Look at your les-
sons: how could they be different if they exploited audio?

---------- response 1
  12/03     16.37     oppy       research





        In combination with a speech input device the audio
output could essentially allow users to be free of the
terminal and to use their hands to perform some other task.
One person has suggested to me the our voice input could
be used in conjunction with the audio output to enable
PLATO to act as a workbook for somebody servicing or
learning about a piece of mechanical equipment. The input
would allow one to ask for help or turn pages while actually
handling some equipment.

---------- response 2
  12/03     17.18     bruce iehl       mtc

Also the touch panel has been somewhat neglected -- how
many lessons could eliminate the use of the keyset
altogether? It seems we are simply flooded with hardware!

But there are still many logistics problems associated
with the various hardware. Also, many people fail to
recognize that without the system software support
we have grown accustomed to, the hardware will
remain unused. This is by no means a criticism of the
systems programers! But as new hardware comes of age, and
new computers are added, does anyone think to order
more programers? ... and more authors to take over the
existing chores?

---------- response 3
  12/03     17.29     berger       mfl

But let's warn the potential audio user of one thing:

We are still awaiting delivery of the FIRST of our
order of 50 audio devices at the language lab, and
they were promised for last July.

Anyone considering using audio has a long wait ahead
of them.

---------- response 4
  12/03     18.10     silver       ve

Well, some of have to save some nickels and dimes before
placing our orders, anyway. I think Gilpen's point is
well taken.
I think some sort of computer generated speech device will
turn out to be the cheap, every-terminal-has-one device,
because such gadgets need not have moving parts. I think
a well recorded voice sounds better than current "voders"
but they are improving. No need to wait on them, though...

---------- response 5
  12/03     19.45     john r       reading

Perhaps this is a good place for a discussion and a distinc-
tion...





I will discuss three kinds of audio response systems:
analog, digitized, and synthetic. For each system I will
discuss (1) the form of storage of a message, (2) the making
of a message, (3) the delivery of a message, (4) some of
the positive and negative implications of (1) thru (3), and
(5) the current state of the art.

continued...

---------- response 6
  12/03     19.55     john r       reading

Analog Audio Respnse Devicesz

The simplest analog audio response device is a tape recorder
that is turned on and off by the terminal. The simplest
practical device is one which allows random access to
messages instead of the serial access that a simple tape
recorder offers. As far as I know the least expensive
such practical device now available is PLATO's.

The message for such a device is made by a human recording
his/her voice on the disk, drum, or tape that the device
'reads' from. The message is reproduced for the user by
sending the machine a request to start playing at a
particular location.

In some applications, a number of these respone devices
are kept at a central location (with the computer), and
requested messages are sent over telephone lines (separate

continued...

---------- response 7
  12/03     20.04     john r       reading

from the data lines) to headphones at the terminal. The
advantage of centralized analog audio is that the number of
audio devices needed can be kept to a small number. The
disadvantages are that (1) too many people wanting too many
different messages at the same time will cause long queuing
delays, (2) separate communications lines are required. Such
centralized analog systems work passably only with limited
numbers of terminals in a limited range of lessons.

PLATO currently uses a distributed analog audio response
system with each terminal having its own device. The
advantages to such a system are that immediacey of response
is never a function of system load, and that the commands to
access the audio can share the same data lines the terminal
uses. The disadvantages are a higher per terminal cost in

continued....

---------- response 8
  12/03     20.10     john r       reading

the short run, and the need to maintain a disk publishing
and distributing service.





A general advantage of analog response systems is fidelity
of reproduction. Not only speech, but also music, heart
sounds, and special effects can be used for instruction. A
general disadvantage is that no analog systems (that I know
of) have sufficiently rapid random access that sentences can
be built from single words (though they can be built from
phrases). This disadvantage is probably a chimera since
even with rapid access, sentences constructed from words
would lack good intonation patterns.

next: digitized audio

---------- response 9
  12/03     20.18     john r       reading

Digitized Audio Response Devices

'Digitized' audio refers to human speech that began as human
speech and was electronically converted to a form that could
be stored by and in digital devices like computers and disk
packs. The source of a message is the human voice. The
voice is attacked by electronics and stored in an altered
form, usually on a disk pack. It is reproduced by reading
the data from the disk pack, running it through a digital-
to-analog converter, and shipping this converted data to
a speaker or headphones.

Digitized audio is almost always stored at a central site.
In some applications¬ the¬ D¬ to¬ A converters are also centrally
located, adn telephone lines like them to each terminal. In
such applications, the only advantage over centralized
analog audio is that the system can handle more people in 

continued...

---------- response 10
  12/03     20.23     john r       reading

somewhat more lessons with somewhat fewer queuing problems.
I understand that this approach is being used at Pat Suppes'
CAI lab at Stanford, and I'm awaiting a report from that
lab.

The D to A converters can also be distributed to each
terminal. Since the data is digital until it reaches the
converter, it can be sent over the same data lines that the
terminal uses. The disadvantage to this is that while the
audio data is being transmitted, visual data must wait.
Since at its most compact (as far as I know) this data runs
at about 900 bits per second, PLATO terminals (where the
communication scheme handles only 1200 bits/second) would
not be able to handle intermixed audio and visual displays
- a distinct disadvantage. Bolt, Beranek, and Newman are

continued...

---------- response 11
  12/03     20.30     john r       reading





currently working on such a system.

An attractive advantage of digitized audio is that no
distribution/production network needs to be maintained
(as must be done with distributed analog systems). Another
is that message response time (on a distributed digitzed
system) can be quite good (not true with a centralized
analog system).

The chief disadvantages of digitized systems are (1) the
very high cost of storage hardware, and (2) the high
bandwidth of the channel from storage device to terminal.
Number 2, especially, has serious implications in the
PLATO environment.

next: synthetic audio

---------- response 12
  12/03     20.39     john r       reading

Synthetic Audio Response Systems

Synthetic audio is speech produced by algorithm rather than
by people. The source of a synthetic audio message is
(in the most sophisticated systems) a written phrase or
sentence. Through programmed consultation of dictionaries,
sentence grammar rules, pronunciation rules, pronunciation
lexicons, inflexion rules, and the like, the computer
produces a string of phoneme and stress codes - and these
are stored until the message is called for. When the ≤ 
message is called for, these codes are sent over data
lines to a decoder which (like our human speech equipment)
turns them into speech.

In such systems, the intermediate code (the phoneme and
stress code string) is centrally stored. The reproducing
equipment is always distributed.

continued...

---------- response 13
  12/03     21.02     clark       lawyer

Because of the high interest on the subject of the
audio device, teh speech recognition system, and other
external devices, I suggest that a notes lesson be
created (if there already isn't one) to enable authors
to express their views on potential uses of such devices
as well as any suggestions conerning related software.

Comments??

J. Clark Kelso

---------- response 14
  12/03     21.03     john r       reading

A major advantage of synthetic over digitzed systems is





that synthetic systems will probably require only 300
bits per seond between the computer and the reproducer to
generate speech: in the PLATO environment this means that
the terminal could have coordinated audio/visual displays.

A major disadvantage of synthetic in relation to digitized
or analog audio is that it _is_ synthetic and nobody has any
real experience with the psychological/instructional effects
of such a system on users of various ages. If the only
bad effect is esthetic, synthetic audio will be an important
future tool.

John Allen's group at MIT is doing advanced work with
synthetic audio.

---------- response 15
  12/03     21.09     john r       reading

The preceding discussion has been long and monopolistic.
I think there is enough general interest to justify
immediate additions and corrections here. The note lesson
'peripheral' is a good place to conduct debates and long-
term inquiries.

---------- response 16
  12/03     22.34     bowery       comm

And when I saw there were 15 replies, I thought there
would be a lot of information here.

Ah me....

  A similar problem exists with both audio devices and
projector/microfiche systems. The information which is
to be presented cannot be modified by software. On the 
other hand, the charset memory (software modifyable) is
widely used because it is available at no cost/student
beyond that of the base-line terminal/connect cost.
  Once a slide or audio disk is made, enough copies
must be made to support all simultaneous students in the
lesson. The local cost of utilizing PLATO becomes a linear
function of the number of simultaneous students. To 
minimize local cost, the temptation is toward "cafateria"
style CAI which is antagonistic to global concerns.

---------- response 17
  12/04     15.42     volpe       mfl

Ok, I understand all that. Now can you tell me when FLB
will be getting their audio devices?
tom volpe

---------- response 18
  12/04     18.30     capter       comm

John, thanks for the discussion. I think you should
transfer it to a more accessible place. If some information
regarding the fidelity-bandwidth tradeoff for digitized





audio were added, it would be a nice supplement.

Example (figures not accurate);
telephone fidelity speech (2khz)---(digitized) 3.2khz
FM quality ------ (digitized) 18.3khz
barely understandable-----(digitized) 500hz

---------- note 174 8050?!?
  12/04     12.28     warner       iu

        How come, when you have only 2 blocks remaining
in your lesson, and you cerate a common, it still says
How many words in this common? ¬6 (Max=8050)
¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬¬ ¬¬¬↑ˆ
¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬¬ ¬¬¬↑ˆ
¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬¬ ¬¬¬↑ˆ
8050?!?!?!?!
Warner

---------- response 1
  12/04     13.36     woolley       p

Just because the editor doesn't bother to see how much space
you have left until later. 8050 is the absolute maximum
common length.

---------- note 175 keys
  12/04     13.10     darlene       pso

I am getting my keys scrambled for me. Letters are numbers
next and back being pressed for me.
Erasing just gives me more junk keys.
Problem was acute while writing pnote.

---------- response 1
  12/04     13.44     john r       reading

Try lesson -repair-.

---------- response 2
  12/04     15.55     lieber       m

       Only if it happens at one terminal.

---------- note 178 clac
  12/04     13.42     stone       pfw

         is it possible....
to make term -calc- and term -clac- synonomous?

please?????????????

        Scott Stone

---------- response 1
  12/04     13.48     berger       mfl





I hope this is meant as a joke.

---------- response 2
  12/04     13.48     frankel       p

Me too!

---------- response 3
  12/04     13.54     marty smith       mtc

what scott'l probably do is come back and say that
somebody ripped off his sign-on. if you'll believe
him after that, would you like to buy the
brooklyn bridge? send me a p-note for details.

your-ptbarnum-SCHMUCK!

---------- response 4
  12/04     16.56     hinton       ssu

If "clac", why not "≤ c-lc" ? (Or "xalc" ??) ??
ndh

---------- response 5
  12/04     17.04     broadus       css

why not just an okspell for terms in general
haha

Bob Broadus

---------- response 6
  12/04     17.27     berger       mfl

Or on the signon password page...or enable the ans key.

---------- response 7
  12/04     17.34     broadus       css

Mr. Berger,

   I think I have heard this suggestion before. Has
this possible feature been studied by the systems staff?
It sounds like a truly helpful addition to the PLATO
system. Especially when using csstudent.

Bob Broadus

---------- response 8
  12/04     18.36     stone       pfw

sure would make it a whole lot easier if every arrow
had a specs okspell..





how about sign-on restricted lessons?
um- lets think of some other stupid thing...
and marty, i just bought the brooklyn bridge
from kemp...so how come your selling it?

        Sol..

(heh heh)

---------- response 9
  12/04     23.22     joe       reading

I kna't cpell well. Kan plato give me
a speling leson so i kna lern how two tip the rite les-
on?

---------- response 10
  12/05     08.14     john r       reading

Joe. We have some lessons tha may help you.

---------- response 11
  12/05     11.50     frye       mfl

I sure do miss the good old "specs best"....

It would work perfectly in this case!

Dave

---------- response 12
  12/05     13.01     s zweig       iumed

Well folks, in the Bloomington Indiana phonebook,
you can find a listing for Ma Bell under 'fone company'....

---------- response 13
  12/05     14.09     dave       infe

        For sure, the editor should be able to handle just
this one very common typo!!

---------- note 185 timel key
  12/04     14.09     dave       matha

System variable "key" is apparently set to 0310 when the
time limit expires on a -timel-. Is this always true?

If so, is there a system-defined constant equal to 0310?

---------- response 1
  12/04     15.03     white       p





The setting of "key" is a side affect of -timel-, and
should not be counted on. No, there is no system
defined constant for that key value.

---------- note 187 +transfr+d
  12/04     15.13     bowery       comm

I notice that the form of the new dataset commands is
very similar to the form of the -transfr- command. I was
wondering if the idea of generalizing the -transfr- command
to include disk accesses had been considered. ie:

transfr d,dloc;c,cloc;len $$ etc.. for datain
* ¬+
transfr n1;d,dloc;len $$ etc.. for dataout
* or
transfr d,dloc0;d,dloc1;len $$ disk to disk

I am aware that anyone of these forms could imply more
than one real disk access (the last one necessitates),
however, these types of accessing occur anyway in most
programs and simply take up more lesson binary, cpu,
and lesson space than a generalized form of the -transfr-
command would. Too many disk accesses could still be
dealt with in the same way that they are presently.
Further, the population explosion of commands could
be cut down.↓If it was considered, why was it discarded?

---------- response 1
  12/04     15.51     elston       rhrc

That does look nice (costly perhaps) and neat for some
data manipulations. Of course, you would still have to
worry about zreturn values as in dataset commands but it
certainly seems to have merit. The only draw back I could
see is if it interfered with the normal speed or efficiency
of the present -transfr- command.

2c≤/

---------- response 2
  12/04     16.00     bowery       comm

That could be taken care of in the condensor with a
user transparent command that would be condensed if
there were a "d" argument in the tag field.
This is commonly done now if I am not mistaken.

---------- response 3
  12/05     15.35     dave       infe

        In fact, the whole requirement that disk accesses
begin on record boundaries and be in whole-record
increments could apparently be eliminated.
        If you specify origins of dataset objects as base-¬+-
displacement, the number of 'extra' disk accesses needed to
find your origin on disk would remain the same; all that





would be needed would be a slight extension of the PP
routine that is being written to handle flexible-sized
records.
        Similar considerations apply to length specification
in terms of words.
        Suggested syntax:
transfr d,n1(n5);n12;words,n34
would move (n34) words from the disk location beginning
(n1) words after the start of dataset record number (n5),
spanning record boundaries if necessary.
transfr n1;d,0(1);2 $$ equivalent to dataout 1;n1;2
        A reasonable requirement would be that the
displacement be smaller than the record length.

---------- response 4
  12/06     21.46     blomme       s

   Reserve/release will be a problem if you want to
deal with partial records. No PPU program changes
are needed and the routines are already "flexible."
A smaller minimum record has been considered and is
possible; my own view is that we are encouraging
poor program design and mis-use of the system if we start
allowing disk accesses for only a few words at a time.
 Somehow "hiding" the disk accesses inside of -transfr-
has some interesting aspects, but the fact that
a dataset command would still have to precede any disk
access, the need for the reserve/release mechanism still
"external" to the transfer, etc., together with the fact
that this offers no new capabilities at a time when so
many new things need doing makes any such change extremely
unlikely. As to the demand as to why someone's new pet
idea was or wasn't conceived of the system staff, I
confess I am at a loss for words.

---------- note 193 articles
  12/04     16.03     tebby       pso

ANNOUNCING:
        "articles" - a file containing an alphabetical
listing by author name of articles about the PLATO system.
The file is seen in the "author mode". The content is the
same as that of the chronological listing appearing in the
report "PLATO Highlights" (updated version, November 1975).
The articles cover the time span from 1960-1975.

Please send citations for any article not appearing in the
list by p-note to "tebby of pso".

---------- response 1
  12/04     18.48     jim       pso

for a reminder about this lesson and other such "goodies",
refer to the author resources section of AIDS (choice g
from the main index) or enter such file names as:

articles
topics
authors





crash

there are several lessons which contain lot's of nice info.

thanks ---- tebby ¬+ liz, articles is a nice addition to
the resources available to the PLATO authors.

---------- note 200 blank scrn
  12/04     18.29     broadus       css

I decided to try and make up a solution to the blank
screen problem before the WELCOME TO PLATO page while
the menorah (I hope it is spelled right) is in use.
This temporarily resides in "logoproc." There is no
inspect code, so you can see my stupid way of doing
it. If it is faster, or even the same speed as loading
the characters at first, and you like the method, tell
your favorite system programmer you would like to see
it in the next version.

Thanks,
Bob Broadus

p.s. Notice that it only uses one character. I thought
     very few lessons use the arrow any more so currently
     that is what I am using. It could be changed to
     any character. This takes care of the ruined charset
     problem too.

---------- response 1
  12/04     18.55     frankel       p

Actually, most of that delay is caused by some new (deep-
down) system software -- it should be sped up greatly in
the next few weeks.

---------- response 2
  12/04     19.23     b sherwood       s

I'll second that. It is an unfortunate coincidence that
the pretty menorah is being blamed for a one-second delay
that is not its fault. We are sending a question to the
terminal, asking what kind of a terminal it is (in
particular, it could be an intelligent terminal of some
kind). We found that as a side effect we are generating
a one-second delay, which we think we can get rid of.
Sorry for the trouble!

---------- response 3
  12/04     20.20     broadus       css

There is a delay on the Press next to begin page,





which occurs whether the characters are loaded
or not. I am speaking of the blank screen delay
while the characters are loading. Also it was
mentioned earlier, that someone was upset that
his character memory was over written. the version
where lab is pressed (in logoproc) only uses
the three 'not linked to plot' slots. In that
the menorah regularly takes up 9 characters
(a,b,c,d,e,f,g,h,i), this would be an improvement.
With keypresses getting lost or worse yet delayed,
it is not to hard to get to the Press next to begin
page. If you are lucky enough to get the menorah,
you will have to load all the charset again, where
as with the method in logoproc, one could hit stop
while loading characters, and still have all the
important ones.

Bob Broadus

---------- response 4
  12/05     00.56     little t       research

The blank screen is caused exactly by what Bruce was saying.
They now send TWO echos to the terminal which are padded
with delays so the key buffers don't get over written.
That delay was there before the menorah was even added!

Todd

---------- response 5
  12/05     02.59     broadus       css

I think that we are talking about two different delays.
I am speaking of the one you get when you hit next
on the WELCOME... page( to get to the WELCOME..page)
if you get the menorah, there is a second or two
of blank screen, in that the erase is done before
loading the 9 characters.
I think you are talking of the delay after next is
pressed on the Press next to begin page. I am also
aware that this happens, and I realize that this
is independent of the character loading. I saw some
interesting results of this in FLB the other day.
a young lady was going from one terminal to the next
pressing next (nothing happened so she went on to the
next empty terminal.) So a second after she left each
terminal, the WELCOME... came on, but she was long
gone. I don't know if she ever found a terminal that
she thought was working.
Bob Broadus

---------- note 207 backspace
  12/04     19.31     b sherwood       s

Would anyone be annoyed if we changed the action of the
backspace (shift-space) to NOT be affected by the left
margin? At present, a backspace from the left margin
moves up one line and over to the right edge of the screen.
This capability seems not very useful, and it causes all





sorts of problems. It seems more reasonable simply to move
left one position, independent of margin. An exception is
the case of a long student answer containing backspaces
for superimposed accent marks, which could be handled as
a special case (but only if the student's answer has in
fact wrapped around one line already -- an initial backspace
by the arrow would just move toward the arrow character).
Comments?

---------- response 1
  12/04     19.50     clark       lawyer

I would think this would be more logical. Moving up a
line and to the right side of the screen certainly is
strange.

---------- response 2
  12/04     20.14     curly       iu

  Would love it...

curly≤≤≤≤≤¬↑ˇ↑ˆ↑ˆ↑ˆcurly

---------- response 3
  12/04     22.04     david       hebrew

It would cover a multitude of sins, as well as
solving a multitude of problems.

Obviously, I am prejudiced in this matter, but
nevertheless, ONE VOTE FOR!

E.

PS Will the implementor accept bribes of pizza and/or
beer?

---------- response 4
  12/04     22.14     whansen       csa

Maybe easiest and most straightforward to not backspace
past left margin. It depends whether you consider
the text string or the display line as the unit of
information. Anything less than a text string
treatment might as well be as simple as possible.

---------- response 5
  12/04     22.28     dowling       ames

Another affirmative vote! ¬↑ˇW¬≤↑ˇC¬≤↑ˇD

---------- response 6
  12/05     10.38     fritz       ames

 Yes!! PLEASE!





---------- response 7
  12/05     12.47     hinton       ssu

Aye.

---------- response 8
  12/08     16.14     shirer       s

Yes to bs past left margin..

what happens when reach left side of the screen? wraparound?

---------- note 211 readset
  12/04     21.15     dave fuller       uimc

        3 questions:
1)      Could all the execution errors in the readset
command be eliminated? I am referring to the particular
case that I am presently fighting, of 0 or -0 (";;;;;;;;;;")
It is sort of a pain having to discover/allow for this, and
I'm not sure how many other permutations of the same problem
could be around, undocumented.
        Perhaps at the same time some returns could be added
to readd and company to eliminate the threat of execution
errors in this species of commands, and make the handling↓consistent with datasets.↓
2)      Is there anyone out there who relies on the afore-
mentioned facts to survive?

3)      Would it be possible, someday, to be able to specify
what type of file you are after during a readset? I mean:
readset course;(name),(pass),(nstu)
readset datafile;(name),(pass)↓
        Seller's priority: 3/100

---------- response 1
  12/05     08.43     berger       mfl

"readset" currently tells you whether you have attached
a course or datafile, if that helps at all.

---------- response 2
  12/05     11.05     alan       reading

It doesn't help someone who is reading your code. There is
considerable difference between a course and a datafile
(I think). Something should be done about this situation
and Dave's proposed format sounds like a good one.

The execution error problem is just plain ridiculous.

---------- note 212 pnotes
  12/04     21.30     plato pnotemaster       p

To all our pnote patrons:

        We at the PLATO PERSONAL NOTES SERVICE are happy to





announce that unlike our major competitor, there will be
no rate increase after Christmas. In fact, we see no plans
for a rate increase at any time in the future.

        We encourage everyone to send their Christmas
Personal Notes early this year, avoid the last minute rush!
Though we are currently delivering notes 172,800 times
faster than our nearest competitor, we will refuse to
promise delivery in time for Christmas for any notes sent
after December 20, 1975.

        Thanks for your support during the past year, our
1st full year of service. We are currently handling over
1200 notes per day.

Your Plato Pnotemaster

---------- response 1
  12/04     22.08     whansen       csa

It is not at all clear your rates are lower.
The federal goverment subsidises both PLATO and the
brand x service. I pay for both with my tax dollars
(which I suppose, come mainly from the taxpayers of
Illinois, myself included, ...).  It is probable
that the government subsidy pe_r unit message ≤_ is
lower for brand x. (I still prefer PLATO. Keep up
the good work.)

---------- response 2
  12/04     22.58     brand       matha

brand - matha, thank you.. NOT brand - x.....

( and i dont have any services, yet )

---------- response 3
  12/05     05.32     santa       claus

Hmmmm....
How much are your parcel-post rates?
(oh, that's right...no deliveries December 24)
Listen...keep up the good work and you may find
a new processor in your stocking this year!

---------- response 4
  12/05     14.53     gilpin       peer

Congratulations to the Plato's Pnotemaster. I at least
would be interested in more complete statistics, such as
the growth of pnote use over time, both absolutely and in
relation to ¬$ of authors (or author-hours), etc. Pnotes
are an outstanding example of a highly useful feature,
not originally foreseen, which has emerged during Plato's
evolution.

As far as comparisons with the Other Outfit are concerned,
it is notable that Plato pnote service keeps getting better





and more convenient, rather than worse.

---------- response 5
  12/05     19.35     marty smith       mtc

Dear Santa,

Are you really there? The guys here at work say that
you don't exist; and then made fun of me when I said
that you did too!

Now I can't prove it, 'cause Uncle Plato told me that
you aren't there either.

Is there REALLY a Santa Claus?

Your-been-a-good-boy-all-year-SCHMUCK!

---------- response 6
  12/06     09.58     ken       conn

       Yes, Marty, there is a Santa Claus...

---------- response 7
  12/06     10.19     friedman       wright

hahahahah,,,,,,,,but wont the ole pla(to) be down from
December 20 to January 10 ruining my whole plato based
Christmas vactaion. HUHUHUHUHUUHUH????
HYMIE

---------- response 8
  12/06     21.23     carter       comm

Re: response 4.
  p-notes "were originally foreseen". I can supply an
extensive bibliography. Simply because Plato has chosen
to make only very basic use of its potential as a
communications medium doesn't mean such uses have not
been conceptualized.

---------- note 226 load ¬6
  12/05     09.27     elston       rhrc

I've been wondering if there has been any progress on
loading arrows with program generated text. I know this
was thrown around a long time ago but I haven't heard
anything lately.

Specifically, what I want to do is show line(s)
of text on the panel and act on them as if they had
just been typed in by the user, so that the edit and erase
keys could be used on them.

Of course, loada or close lets me do just the judging of





such text, but I also need the editing facilities which at
present require some rather strange pause loops or a -press-
COPY1.

One other thing that I really could use is the ability to
assign touch keys to a micro table.

Any possibilities of these things being implemented?

---------- response 1
  12/05     14.06     judy       pso

You want a keypress of "a" (for instance) to be equivalent
to the touching of some point, by defining "a" in the
microtable?

---------- response 2
  12/05     15.42     elston       rhrc

Well, what I would like is a micro table similar to the
standard micro tables with from 10 to 40 characters assigned
to each touch square(this would require 1 to 4 disk blocks).

For instance, if touch square 0 (lower left corner) were
touched with a force micro in effect (or a square associated
with the micro key had just been touched) then the character
string or function key would be entered at the arrow.

It would be helpful if the touch-micro editor could use the
touch panel in generating the key to assign a string.
 A numeric selection would also be necessary and a nice
numeric touch grid on demand would help there.

---------- response 3
  12/05     16.00     elston       rhrc

While I'm at it. Is there any way to get at the strings
stored in the standard micro table in a running lesson
other than at an arrow.

For instance, could there be a command -micro- or -showm-
or some such which would read out or display the contents
of a given micro key. Example:

at     1010
showm  1

would essentially -showa- the contents of the "a" slot in
the micro table.

Or, perhaps, a systems array (i.e., zmicro(xx)) which would
contain the contents of the micro table. This would be
read only, of course.

Any chance?

---------- response 4
  12/08     09.40     judy       pso





If you want to show the "1" position from the standard
micro table, it is infinitely simpler to just say
-showa ¬a-. (-showa "access,a")
How would you use such a feature?

In source code, micros do not work by table look-ups, but
by direct substitution. I.e., when using your own micro-
table to edit and you press MICRO,e and see "elephant",
your source code actually has in it the letters "elephant"
and _not_ "MICRO,e".

Your idea of a -touch- "microtable" is interesting.
To get a touch grid, just press TERM and type in "grid".

---------- response 5
  12/08     10.49     elston       rhrc

I assume you mean -showa '¬a'- with quotes.

As to my intended use of a -showm- (or some such) command,
I wish to use it in the following manner with my own micro
tables (touch keyed or not). Some part of my programs now
show various lists of words and phrases on the panel and
allow the selection by touch of any combination of those
terms. The total combined lists number around 20,000
entries and is growing. At the moment I make use of large
commons and datasets to store these terms and their list
descriptions. The commons take up a great deal of ECS and
the storage needed to store data from datasets is taken up
with other tables.
So if I could store some of the most often used lists in
micro tables and show them on the panel with -showm- I could
save a fair amount of ECS and CPU in these lessons.
I would settle for 2500 words of storage or executable
commons.

---------- response 6
  12/08     11.08     elston       rhrc

As an example of actual use it would be some thing like this
this:

unit  terms
doto  1,i‎1,numtrms   $$ number of terms to show
at    start+i≠100     $$ position of term(s) on panel
showm trm(strtl+i),20 $$ this would show the (strtl+i)↑ˆth↑ˇ
*                       micro slot (e.g., 'elephantoid')
1

I'll probably find some other way to solve my problems and
the conversion and upkeep of the appropriate micro tables
would be prohibitively time consuming now any way, but I
still would like to see these features _someday_....

The numeric grid I was thinking about would have numbers
from 0-255 (or 0400-0777 or some subset of either of these
sets).





---------- note 228 disconnect
  12/05     10.39     burr       m

I will attempt to answer some of the questions concerning
telephone disconects. The savings will vary with the user
and as yet I have not obtained a grand total and may not go
to the trouble to obtain same. Only those line which orig-
inate in Urbana and have the other end outside the state of
Illinois(interstate lines) will be effected. Most individ-
ual telephone companys have tariff restrictions, which we
cannot meet, regarding minimun time on temporary disconnect.
        It is true that the possiblity exists for the user
to come back on different facilities than they had original-
ly however in actual practice this is unlikely. The normal
procedure for a disconnect of this short a duration would
be to break the line in Champaign only. The problem will be
that if a line goes down during the interval we wont know
until Jan 10th in order to start repairs.

---------- note 229 more disco
  12/05     10.46     burr       m

More on disconnects. Regarding special reduced rate lines.
My negotiations resulted in assurances that any loss of
telpak would not be the result of our temporary disconnect
but just normal priority bumping. I will emphasis again,
as I always have to GSA users, that use of these special
billing sevices has a high risk factor over which we have no
control. Things are better than they used to be but, as we
are dealing with the federal government, they will never
approach perfection. (see the response by Straayer to the
original golden note)

---------- note 231 recursion
  12/05     11.32     osborne       iucs

Hey, kids...... I'm back after an eight month absence
and work once again has been started on my recursive
interpreter but the lesson name has been changed from l
to languagel. Take a look and see what recursion is all
about.
    Osborne, iucs

---------- response 1
  12/05     11.35     stone       pfw

who is osborne?

---------- response 2
  12/06     01.11     stephanidis       iumed

I give up, too.

---------- response 3
  12/08     20.47     meers       wright





osborne is the guy who is writing a lesson on recursion
recursion recursion recursion recursion recursion recursion

---------- response 4
  12/10     16.36     dave       infe
  
        So if you want to keep fooling around with it (which
is your prerogative) please take the inspect code off so we
can get SOME idea what's up!!
        Thanks!

---------- note 236 crash
  12/05     14.01     little buddy       matha

option 2 in lesson crash takes you to the author mode page.

LB

---------- response 1
  12/06     01.12     avner       s

Thanks, fixed now. (at last!)

---------- note 239 "M" opt
  12/05     14.32     volpe       mfl

I think the "modify" option should be better explained
than it is. When you enter it, all it does is ask you
what you want changed. It says NOTHING about the SHIFT-M
part of the modification. Any programmer who hasn't worked
with the "M" option will be totally lost when he tries.
I know, I was.
tom volpe

---------- response 1
  12/05     17.35     snellen       medneta

I AGREE!! I have looked at it, but I still don't know how
it works.
JES

---------- response 2
  12/06     04.37     david       cerlcc

Perhaps you should press -HELP- while viewing a block,
and check out the directives for modifying contents of
a block.

But there should be a help page in the "m" option as well.

---------- response 3
  12/06     13.46     snellen       medneta

As a matter of fact, I _have_ pressed HELP while viewing a





block, and I _have_ seen the description of the m and M op-
tions. That is precisely the problem--these options are
_described_, but they are not _explained_, I _know_ these
options are used to modify character strings, but _how_ are
they used and in what context?

JES.

---------- response 4
  12/06     15.19     david       cerlcc

Clued in by pnote.

---------- note 240 queue
  12/05     14.45     volpe       mfl

To add another note:

Today, why am I the last one or next to last one in the
condense queue? It never fails. It took me 10 minutes
just to condense a lesson. EVERYONE was jumping in ahead
of me. HELP!

Please explain the whole condense queue farce to me, will
you? thanks, tom volpe

---------- response 1
  12/05     14.53     k mast       p

Try looking in lesson 'aids', under 'queue', As for being
a farce, I'd like to see you get comparable service at any
one of a number of other computer facilities!

---------- response 2
  12/05     15.37     volpe       mfl

Is "cpu time spent condensing" also refered to as TIPS?

---------- response 3
  12/05     15.47     k mast       p

No. TIPS refers to CPU time spent executing a lesson.
I know of no way to find out what your CPU time spent
condensing.

---------- response 4
  12/05     16.27     b sherwood       s

But the effort spent by the condensor on your lessons
is roughly proportional to the size of the lessons
condensed. In order to get good service, you can help
by condensing only those parts of the lesson which you
are in the process of checking out. There is roughly
a factor of 10 between condensing a 2-block routine and
a 20-block whole lesson, even tho you might execute only
the 2 blocks. Of course, some programs can't be worked





on in little pieces in this way, in which case you will
get slow service compared to people who do not request
so much condensing. Note that the effect is cumulative:
ask for little all day, and during the afternoon you will
get good service. We do hope within the next 6 months to
provide additional condensing power with new computer
equipment.

---------- response 5
  12/05     17.08     judy       pso

Often authors may condense several blocks, just to do
the necessary "setup" to test a single unit or iteration.

It will save you condense time, and may save you the
nuisance of working through 14 preliminary questions,
if you will write a little "setup" unit just for testing:

unit setup $$ set necessary parameters
calc length‎5 $$ choose arbitrary values
     steps‎3
     tryno‎8
     increm‎.1
     end‎14
etc.
jump testit $$ the unit of current interest
*
I can usually get by with 1 block for defines, 1 block
for "setup", plus 1-4 blocks of the real program.

---------- response 6
  12/07     14.38     david wallace       mathg

  I know exactly what you mean. I'll be 1 out of 1
one minute, and the next minute about 10 people have
gotten ahead of me!!! Do some people have precedense
when it comes to condensing?
-David Wallace

---------- response 7
  12/08     09.42     judy       pso

Can't you read? People who have less condensing charged
against them than you do have priority over you.

---------- response 8
  12/08     09.56     mike b       cornell

       Students, of course, also have priority.

       That's what we're here for, remember?

---------- note 243 disk i/o
  12/05     15.02     michael       english

How are the 10 disk accesses allowed in a finish unit
computed now that record boundaries overlap disk blocks?





I.e., is there any difference in transferring 10-64 word
records than there would be in transferring 2-320 word
records?

Reason I ask is that it would be more logical for my pur-
poses to use the smaller size, but not at the cost of great-
ly increased disk accessing demands.

---------- response 1
  12/05     18.29     blomme       s

   The count is per command, not the number of accesses
the system may actually require (which is invisible to
the user and possibly subject to change).

---------- note 253 NOTES
  12/05     19.07     ian       conn

Is it intended that when a note is sent to a person
using a system term, he will not see his NOTES flag?

I doubt it,
Ian

---------- response 1
  12/05     19.27     blomme       s

   While not desirable, this is a problem that has
long been known (and commented upon)...perhaps some
way of solving the problem can be found, but don't
hold your breath.

---------- note 260 nc copy?
  12/06     07.35     kawell       cerl

   What is the reason for the -copy- command only working
with Student Variables? It makes a common editor kinda
kludgey. Also, the execution error doesn't say anything
as to what the trouble is.

Len

---------- response 1
  12/06     09.36     warner       iu

The reason for the -copy- command only working with
student variables is that while you are typing at an
arrow, your lesson (and its common) are not loaded
into CM from ECS. Only your student variables are
loaded, saving transfer time.
I don't think this will be changed in the near future.

---------- response 2
  12/07     10.44     b sherwood       s





Execution error? You get a condense error! (And the
aids description of the -copy- command says you must
use student variables.)

---------- response 3
  12/07     12.41     white       p

Bruce, you only get a condense error if it is an
explicit *nc* variable reference... for example,
copy nc(n1),20
is not a condense error.

---------- note 261 small site
  12/06     09.32     roper       siu

What do people at other small sites (1-3 terminals) use
their terminals for and what are their priorities?

Obviously a school with only one terminal cannot do all
of the things that U of I can with its large (enormous)
number of terminals. We are now trying to figure out
what should be done with this terminal here, and would
be interested to know what other people in the same fix
are doing.

Thanks,
Quetzal

---------- response 1
  12/06     23.26     mike       cornell

Answered via pnote.

---------- note 265 disclaimer
  12/06     12.34     berger       mfl

Apparently, someone with a warped sense of humor is
calling UICC, and identifying themselves as "Mr.
Champaign", "Mr. Urbana", and "Mike Berger".

If this happens to anyone else, I am NOT making
these calls, and I don't know who is responsible
for them.

Incidentally, UICC will not accept any collect calls.

Mike Berger

---------- note 267 condenchek
  12/06     14.11     andrew       hebrew

I suggest that you implement the following:

When a condense request is submitted, read the directory
of the lesson and check space available in each block.
If there were 322 words available in each block (and
whatever is expected in block "a"), then you could





save the wait in the condense queue, as well as the
charge for condense.

Also, it would help if you could check file type on
receiving a condense request. For example, I can now
submit a request to condense an instructor file from
the author mode page. The system tries to condense
it, in spite of the fact that it can't be condensed!

---------- response 1
  12/06     15.08     michael       english

I feel little pity for someone who tries to condense
empty files and illegal file types.....

---------- response 2
  12/06     15.25     berger       mfl

However, when the system runs slowly, and I end up
condensing some random lesson because 8 keys were
dropped at the author mode arrow, it would help a
lot if I didn't have to condense empty lessons.

---------- response 3
  12/06     18.59     blomme       s

   Instructor files are of the same type as TUTOR lessons;
all separate file types ARE determined immediately and are
not condensed (and do NOT require disk accesses to do so).

---------- note 276 S-notes!
  12/07     08.50     clark       lawyer

Lesson "easyread" provides a way by which you may store
permanenly any General, Help, or P-note; and then have
an easy way to page through the stored notes!
To save any note, press shift-S while reading the note.
Here is how to use the lesson:

!IS" all notes you want stored in source blocks. The source
blocks should all have the same name. It doesn't really
matter how many notes you put in each block. Convert these
blocks to common (press DATA1 while editing the block).
Then, set up your lesson to "use" easyread (i.e., change
option "f" on the DATA page of the lesson to read
"easyread"). Create another source block in your lesson
and write the following code:

common yourlesson,yourcommon,length $$ length<1500 for now
use notes

Then just condense. Directions are also in block "a" of
the lesson. (cont. in 1st response)

---------- response 1
  12/07     08.52     clark       lawyer





Send any comment or questions to clark/lawyer. I hope
that this makes it easy enough for authors to be able
to store permanently and then read, any note that they
think will be helpful to them.

J. Clark Kelso

---------- response 2
  12/08     02.57     mike       cornell

While experimenting with this lesson, I had several problems
editing the 'notefile' I set up, with dissapearing changes
when I converted back and forth from common to source.
Perhaps if there was a command that would allow one to read
into variables whatever was in the Save Buffer of an author
(that it was permissable to use, because it probably exists)
then many lessons could benefit, besides this one and the
parrello/agg char generator. It would seem harmless, right?
And probably be VERY easy to implement....

Mike Huybensz

---------- response 3
  12/08     16.03     clark       lawyer

I had the same problem editing the common. As it turns
out, you must first abort the ecs copy of common
before converting it to source to add notes, etc. I will
add this warning to the help pages.

See help note ¬$201. Apparently authors aren't going to
be given access to the save buffer for awhile. Not much
I can do about it...

J. Clark Kelso

---------- response 4
  12/08     16.37     woolley       p

There is really nothing magical about the save buffer.
It is simply some extra ECS storage like you can get
in your own lesson with a -storage- command.

---------- response 5
  12/08     16.39     clark       lawyer

Then what is the problem with giving authors the power
to at least copy what is in the buffer to other variables??

---------- note 281 whose?
  12/07     23.27     s zweig       iumed

        I think this would solve the problem of "who owns
this lesson" when there is an inspect code on the lesson.

Insert an option on the data page which would allow the
desired info..i.e. name and course of last editor, or





perhaps the name on the owner line of the data info ..to be
displayed on the 'Sorry-- wrong code' display reached
by trying to edit without inspect or change codes.
The author would have the option of having this displayed.

sz

---------- response 1
  12/08     09.00     berger       mfl

Such a thing was requested in 1974 - the old general
note, dated 07/30/74, is still there (218).

I agree that such a thing is needed, and wonder how
many people really object to letting other see their
data pages.

Frankly, most of the people that have inspect codes
that I've talked to lately admit that they are
covering up code that embarasses them rather than
hiding things for security reasons.

---------- response 2
  12/08     10.49     hody       med

yeah-- silly isn't it? anything that works is fine!
*
why should clumsy code be a reason for embarrassment?!

---------- response 3
  12/08     13.16     obrien       uicc

I've been watching this go by for a year now without
comment but this is getting ridiculous. I fully agree that
this is one of the biggest "holes" I can think of in the
area of general system services.

---------- response 4
  12/08     19.31     wegner       phar

it would be kind of nice to know whether it is not being
added to the system for political reasons or software
complications. Sometimes i feel all these security options
are used more for the systems protection rather than the
authors protection. I havent heard too many authors seriously
objecting to certian bits of imformation either ie last
edited by etc being made public.
2 cents

---------- response 5
  12/08     20.29     church       iumed

      Please respond. system personnel...





---------- response 6
  12/09     00.45     dowling       ames

        See current general notes 21 ¬+ 114 for related and
further discussion of this matter.
        I feel that the concensus of the discussants is in
favor of making such info available, at least unless the
file owner specifically requests otherwise. ¬↑ˇW¬≤↑ˇC¬≤↑ˇD

---------- note 282 quiz
  12/08     02.04     rich       cs196

Multiple choice quiz-
When PLATO comes up after the holiday it will:

       a) Be more effective than ever
       b) Have problems caused by links between CPUs
       c) Have problems not linked with the CPUs
       d) Be just the same as before

---------- response 1
  12/08     02.21     midden       p

c,d,e.

---------- response 2
  12/08     03.00     mike       cornell

Does (e) stand for "Be permanently infested with Gremlins"?

Mike Huybensz

---------- response 3
  12/08     09.28     golden       s

In all probability, the additional cpu will not have
been connected, perhaps not even installed, and certainly
not in use. Thus, any problems we have when we come back
up will be due to having been off rather than to changes
in the computer configuration. We are moving equipment
around, removing a wall, and changing power and cooling
facilities. The computer changes come later.

---------- response 4
  12/08     11.00     hinton       ssu

Answer:

one of the above

ndh

---------- response 5
  12/08     11.21     volpe       mfl

Is there any chance of any lessons being "damaged" during
the down time? In other words, should I request a print-





out of all my lessons?
tv

---------- response 6
  12/08     11.31     golden       s

The library will be as safe as it always is.

---------- response 7
  12/08     13.02     frankel       p

Actually, it will probably be safer than normal.

---------- response 8
  12/08     13.40     railing       iu

  Midden's (e) probably ment...All of the above.
It was the one thing that was left off the questions
that I seem to find on most of my tests of that type.

Malkin

---------- response 9
  12/08     16.32     rich       cs196

        There was no (e) because the first and last
answers conflict....

---------- response 10
  12/12     19.47     lombardo       ed

e it will go right back down again

---------- note 307 passedargs
  12/08     20.39     fritz       ames

  I'm not even sure this is feasible, but here goes...

  I would like to propose a change ot the way variables
are passed. With the present method of passing the value,
several temporary variables are required (which you may
not really be able to afford!). Also, in the common case
where you want the called routine to affect the passed
values, you must either do it yourself, ie

do   change(abc,xyz)
calc abc‎I
     xyz‎J





where unit change is something like

unit change(I,J)
calc I‎I↑ˆJ≠arctan(exp(sin(I)))↑ˇ $$ or similar monstrosity
* whatever else you want to do to I and J

--- cont

---------- response 1
  12/08     20.40     fritz       ames

or, alternately, you must perform some unnatural act like

do change(varloc(abc),varloc(xyz))

with unit change altered to

unit change (I,J)
calc n(I)‎n(I)↑ˆn(J)≠arctan etc. etc.↑ˇ

which still requires some temporary variables.
  My suggestion: could an addition be made to the passing
machinery to allow the IDENTITY of a variable to be passed?
Ie, in the case above, unit change could be called by
a statement like

do change(<abc>,<xyz>), or
do change(¬oabc¬1,¬oxyz¬1), or some such.

-- cont

---------- response 2
  12/08     20.40     fritz       ames

Then, in the -do-ne unit, the "temporary variables" could
be referred to as

unit change(¬0I¬1,¬0J¬1)

where I and J are NOT actual variables, but dummy arguments,
which are treated eXACTLY as if the passed variables were
inserted at each occourance of the dummies (analagous to
a -defined-d function). If it would make it easier, these
"dummies" could be defined along with your other variables,
but with a special identifier instead of a variable.
  Obviously, checks would have to be made on the -do- (or
-jump- or whatever), since this could only apply to
individual variables. (-do-ing unit change with passed
values of, say, 1 and 2 would run into trouble when
it tried to assign a value to "1"!)

  Is something like this at all possible?

---------- response 3
  12/08     21.03     white       p

Yes, it is possible, but would require the system
to hold the dummies, rather than you holding them
in your own variables. This would require N words





per user on the system to allow N dummies, whether
that user was ever going to use a lesson that needed
them or not. Thus, there are some serious drawbacks
to implementing such a feature. Even so, it may
still be done. However, don't expect any decisions
on it for some length of time.

Also, it would be rather difficult to pass them
as in the define command, but it is possible.
It is more likely, however, that addresses would
be passed, similar to your passing varloc(variable).

---------- response 4
  12/08     21.13     fritz       ames

  Um...
I am stupid. Why does the dummy have to be carried by
each user? It seems to me that the most that would be
required would be the identifier for the dummy, which
could be, if nothing else, assigned some pseudo-constant
in the define. No actual variable would be needed anywhere;
it would more or less be a replacement of the identifier
with the variable (although I'm not at all sure something
like this could be implemented into Tutor, since it would
have to be dynamic).

---------- response 5
  12/08     22.01     midden       p

example:
do      f(a,b)
...
unit    f(I,J)
do      g(I,J)
calc    I‎I+J
...
unit    g(X,Y)
calc    I‎X+Y
*

This should show why one must require reserving each user
with XX many locations for this. After all the size of
each persons student bank is not variable in length! gads
what a mess if it was.
this seems to suggest that either address replace of
arguments or not remembering multiple levels. And also what
if one did just a -goto  g-?

*sigh* and some people say tutor is too simple a language.

---------- response 6
  12/08     22.28     fritz       ames

  Hmmm. I begin to see what you mean. If these things
were called to many levels, you wouldn't be attaching a
variable, you would be attaching a dummy which was attached
to a dummy which was ..... which was attached to a variable.
But... I would think that if the initial concept were
possible, this nesting would be too - IF (which I forgot





to mention before) these dummies were LOCAL to their
unit. Then the problem with -calc I‎X+Y- wouldn't
exist, since I simply wouldn't be defined for the
second unit. How full of holes is this idea?
  As for the case with no passed args: I agree -- yuch.
How about setting them to 0 if there is no corresponding
passed argument?

---------- response 7
  12/09     07.57     frankel       p

Like Larry said, it may be a while.

---------- response 8
  12/09     09.55     b sherwood       s

Local arguments for subroutines will probably be implemented
some day, but they are hard to handle in our environment,
where little status can be saved. But note that judging
is now interruptable, thanks to dynamic allocation of
extra storage when temporarily in that state, So there are
precedents. Some day......

---------- response 9
  12/09     13.06     fritz       ames

Thanks for the explanations. I realize it would take a LOT
of time and work to implement something like this -- just
thought I'd see if I could get some gears a'grindin.

---------- note 319 hardcopy
  12/09     08.58     reid-green       etsu

We have succeeded in hooking up our Plato terminal to a
Calcomp 936 plotter through an H-P minicomputer. This
allows us to get plotted hardcopies which are of good
quality: can be used in publications, etc. If there are
no legal problems we might be interested in providing this
as a service to Plato users if there is enough interest.
       If you might be interested, send me a personal
note: reid-green of etsu

---------- response 1
  12/09     10.36     stone       research

      would like to see such a system, where are you all
located?

maureen

---------- response 2
  12/09     15.14     walter       physio

Perhaps you could include some more details in a note in
lesson -peripheral-. That lesson has the purpose of being
a source of information on use of peripheral equipment with
a PLATO terminal. Probably one of the most important





questions is how much does your equipment cost? Like
Maureen, I would also like to see your apparatus. Are you
in Urbana?

David Walter

---------- note 322 editmessge
  12/09     12.44     curtin       mfl

Please rewrite the new message the author sees when she
goes into a lesson to edit. The 'or to edit etc' should
go down _with_ the message about what to do if you want
to edit. Thanks
C. Curtin

---------- response 1
  12/09     14.48     celia       pso

I agree. Also, it would be sensible if pressing DATA on
THAT display took you to the lesson in student mode. If
that can't be implemented, then the message should tell
you to press BACK, then DATA.

---------- response 2
  12/09     15.54     p cohen       med

        Hey, it does work that way! A DATA press goes
directly into the lesson in student mode.
--paul

---------- response 3
  12/09     16.18     celia       pso

Yes, it does! My first test confused me because it was
a lesson that accessed another lesson, and I ended up in
an unexpected (but correct!) place.

---------- response 4
  12/09     21.22     b sherwood       s

Ok.....message appearance changed as of next reload.
Thanks for the suggestion, Connie!

---------- note 323 touchc
  12/09     12.55     bird-dog       cha

Following the discussions of the limitations of the
touch software. I hope that there is some work going
on to extend the power of the language to utilize the
panel. I am also hopeful that there is going to be
some conditional forms of the touch feature, like
     touchc n2¬,1001,2,1¬,223,1,2¬,
Unlike the touch string now which simply judges all
turned on areas as correct inputs.
   Can we have a systems response so we can begin
thinking of which things will be avaliable and when
those things will be ready.





---------- response 1
  12/09     15.59     p cohen       med

bird-dog,
≥≥ N.B.: The "¬," symbol, known officially as the
"writec delimiter", can be used ONLY with the -writec-
command and no other!
--paul

---------- response 2
  12/09     16.10     larry north       p

Something like a touchc is being considered. We might try
to include touch specifications in the match command or
create new command like touchc.

For now you can use a specs followed by a series of touch
commands and get the information you want from the system
variable anscnt.

---------- note 326 p-solving
  12/09     13.42     novak       css

Could someone please direct me to any lessons dealing with
problem solving? I need a good set of varied problems to
use in a small experiment that I am conducting next month.
Please respond here, I don't want pnotes right now.

---------- response 1
  12/09     18.36     travers       mxc

If you mean algebra-type problems, try solve1.

---------- response 2
  12/10     00.31     dave fuller       uimc

        No pnotes?
        Hmm... fill up a permanent disk file instead of
recyclable pnotes... I guess.

---------- response 3
  12/10     08.48     lewis       unidel

I think NO p-notes is a great idea for this sort
of question. As everyone knows this type of
note crops up a lot and when responses are sent
dircetly it leaves us all hanging.

---------- response 4
  12/10     12.56     nelson       ced

There are some problem-solving simulations accessible from
-mclcat- which is the medical network router. Look at the





first 7 or so lessons in the clinical section, choice 8 on
the main index page.

---------- note 328 HP45 's
  12/09     14.58     stratton       pew

   For all HP45 owners interested in using the digital
stopwatch on your calculator, I have the details in my
lesson....'pfwlogic'. Execute and "READ ALL ABOUT IT!"
Direct all questions to stratton of pfw. If you read it,
make sure you read the section on accuracy!!!

---------- response 1
  12/12     13.38     bennett       phys

By golly--it works! Thanks for sharing this info.

---------- note 330 datapgetnx
  12/09     15.28     ken       conn

       Since I have seen no comment regarding this,
       I feel that I should make one.

         THANK YOU, system staff, for the new, complete,
       concise display shown before an incomplete
       data page to a just-created lesson.

       Thanks again.

---------- response 1
  12/09     16.35     midden       p

You're welcome, and my thanks to the many people that have
revised it and or suggested it.
M.Midden

---------- note 335 new SD
  12/09     16.19     ahasic       med

    I have some suggestions for the new SD. It would be nice
if there was some way to skip over all the non-displayable
commands with one key press. As it stands now one has to
wait for a message to plot, press next, and then wait for
the message to erase for each line.

    I have seen the new SD and instead of the problem being
remedied it is worse. The message is longer and thus takes
longer to plot and erase. Also the functions of NEXT and
BACK have been replaced with NEXT1 and NEXT respectively.
Not only is this confusing to old users but it is easy to
accidentally press NEXT when you want NEXT1. Then you have
to go back and start all over.

    Why not leave NEXT and BACK as they were for the non-
displayable line message and use NEXT1 to automaticly skip
all the remaining non-displayable lines.





---------- response 1
  12/09     17.21     midden       p

Thanks for the suggestion. This has prompted a new system
features note. If you wish to know the options that exist
(as the help pages aren't written *sigh*), leave me a
personal note, and I'll explain to you how to _test_ it.
M.Midden

---------- note 338 merry xmas
  12/09     16.57     merry xmas from       pfw

*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*
*()                                                       ()*
*()                                                       ()*
*()       Merry Christmas and Happy Holidays              ()*
*()                                                       ()*
*()       From all the gang at PFW                        ()*
*()     Purdue University, Fort Wayne Indiana             ()*
*()                                                       ()*
*()  bauermeister,bittner,copeland,ditton,folk            ()*
*()  linnemeier,metzger,modesitt,nichols,peer,postel      ()*
*()  seslar,stone,stratton,sunderman,want,williams        ()*
*()                                                       ()*
*()                                                       ()*
*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*()*

---------- response 1
  12/09     18.16     johnston       ssu

So that's what PFW means. Well, best wishes.

---------- response 2
  12/10     10.05     marty smith       mtc

↑ˆ¬I≤↑ˇ¬I≤↑ˇ¬/≤/≤≤-¬≤-¬≤-¬≤-¬≤-≤≤↑ˇ¬I≤↑ˇ¬I≤↑ˇ¬I

Merry Christmas

*¬≤≤↑ˇ**¬≤≤≤↑ˇ***¬≤≤≤≤↑ˇ****¬≤≤≤≤≤↑ˇ*****¬≤≤≤≤≤≤↑ˇ******¬≤≤≤≤≤≤↑ˇ***** ¬≤≤≤≤≤≤≤↑ˇ******
≤≤≤↑ˆ↑ˆ*******¬≤≤≤≤≤≤≤≤↑ˇ********¬≤≤≤≤≤≤≤≤≤↑ˇ*********¬≤≤≤≤≤≤≤≤≤↑ˇ********¬≤≤≤≤≤≤≤≤≤↑ˇ*******
¬≤≤≤≤≤↑ˆ↑ˆ**********¬≤≤≤↑ˆ** ↑ˇ¬¬≤≤≤≤≤≤≤≤≤≤≤↑ˇ***********≤¬≤≤≤≤≤≤≤≤≤≤≤↑ˇ************≤¬≤≤≤≤≤≤
↑ˆ↑ˆ*≤↑ˇ*≤↑ˇ*

And Best Wishes for the New Year





from Marty

---------- response 3
  12/11     08.42     mailman       ustaf

Hey, that's pretty.

---------- response 4
  12/11     17.40     badnarik       iu

       Hey, I like that too!
Season's Greetings to One and All!
       A VERY MERRY CHRISTMAS and a HAPPY NEW YEAR.

---------- response 5
  12/12     19.51     lombardo       ed

Yes, a very Merry Christmas and Happy New Year, especially
to all those kind people in PLATOLAND who have spent many
frustrating hours trying to explain things to me. You are
all doing a great job, especially the ever faithful pso
people!
dal

---------- note 342 double sup
  12/09     18.54     chabay       s

Students typing answers involving superscripts and
subscripts often accidentally hit the SUPER or SUB keys
twice without realizing it. This is an easy mistake to
make, since nothing actually happens when either of these
keys is pressed, but this typing error is difficult for
students to spot.

One possible system-level solution to this problem would be
to ignore all but the first of a series of non-locking
super or subscripts typed at an arrow, It would not change
the handling of locking sup/sub, so it would still be
possible to raise characters by more than one level.

We're interested in reactions to this idea. (Would it
affect current lessons?)

---------- response 1
  12/09     20.12     berger       mfl

I use 2 consecutive SUP's for display purposes: the
displays may be typed in at an arrow in the editor and
it would be disturbing if the SUP key didn't function
logically.

I wouldn't mind it if the answer judger ignored them
(in order to make judging chemical formulae easier
without having to do kludges like:





putd /ˆˆ/ˆ/ $$ sup sup to sup

as we are constantly being advised to do), but it should,
of course, -storea- properly to avoid unpleasant surprises.

---------- response 2
  12/09     20.17     al       mfl

Why doesn't the answer judging (and calculating) machinery
treat two consecutive -SUPS- as one, anyway? That would
be a bit more logical.

---------- response 3
  12/09     21.26     b sherwood       s

But you could still move up two levels by hitting shift-sup
twice, couldn't you?

---------- response 4
  12/10     04.17     b parrello       med

I think we need a bit of clarification here.

Is the proposal to:

        * change the operation of SUPER and SUB so that
          no matter how many unlocking SUPERs and SUBs
          the student types in a row he only goes up five
          dots?

        * alter answer judging so that sequences of
          unlocking SUPERs and SUBs are treated as single
          SUPERs and SUBs?

        * alter the operation of the arrow so that the
          aforementioned sequences of SUPERs and SUBs are
          all compressed into single SUPERs and SUBs before
          judging ever begins?

---------- response 5
  12/10     04.27     b parrello       med

Basically, if this change were made, would it mean that
the (_)≤≤ˆˆ_≤c will now displays as (ˆ_)_≤_≤c, or does it simply mean
that it will no longer be possible to type in (ˆˆ_)_≤_≤c any more
without using locking SUPERs?

---------- response 6
  12/10     08.28     fay       o

I would be much more in favor of including the handling
of your special case, Ruth, in the proposed canonical
judging machinery that I saw proposed sometime ago.

It would make much more sense for the judging machinery to





be able to look at the answer, spot a superficial addition-
al shift code, and be able to decide that, mathemetically
or algebraically, it didn't affect the answer.

My reasoning is that in spite of repeated efforts of in-
structors to educate their students on the nuances of
PLATO's interpretation of key -super- key as opposed to
key -shift-super- key -shift-sub-, many of them never quite
grasp the idea and have to be helped along at otherwise
simple answers.

---------- response 7
  12/10     08.34     fay       o

I don't anticipate any better success in teaching the
student user to differentiate between super-super and
shift super shift super.

---------- response 8
  12/10     11.05     b sherwood       s

But when is it appropriate for the student to type
shift-super shift-super? When do we ask a student to
go up two levels deliberately? With Ruth's suggestion,
two supers in a row would result in the second being
thrown out. Authors or other experienced users could still
get the special effect of going up two levels, simply by
deliberately hitting shift-super twice. Students who
accidentally hit two supers in a row wouldn't get in
trouble. Or are you worried about students deliberately
hitting two shift-supers in a row (in which case canonical
machinery would save the day, but Ruth's wouldn't)?

---------- response 9
  12/10     12.39     al       mfl

You forget that SUPER1 is a 12 bit entity while SUPER is
only 6. This means that there will be an extra 12 bits
to store if you press SUPER1 SUPER1 to get up 10 dots.
Now, if you come up with some proposal for letting us
store 12 characters in a word, rather than 10...

---------- response 10
  12/10     13.02     frye       mfl

I hope that the discussed change refers to Bruce Parrello's
"*" ¬$2 -- let actual judging ignore excess sup/subs -- since
actual removal of those sup/subs from -storea-able code
would result in mass hysteria amongst the "news" freaks...

Dave

---------- response 11
  12/10     14.59     michael       english

It does seem unreasonable to alter the basic properties
of -storea-, that of accepting any typeable input string





without translation of the input. Why cannot this be done
at the system answer-judging stage instead? It seems a
rather drastic change to make just so lessons utilizing
formula judging can have it a little easier for themselves.

---------- response 12
  12/10     19.33     stan smith       chem

Actually the suggestion goes beyond making it a little
easier. A very large fraction of answers judged -NO- are
the result of multiple -sup- and -sub-. Because multiple
-sup-/-sub- caused so much trouble, PLATO III system
software did ignore the additional inputs. What is being
suggested is that we return to the system which was found
by extensive use over a period of a couple years to work
instead of the current system which has, by similar
experiments, been found not to be satisfactory.

---------- response 13
  12/10     19.43     blomme       s

   Good grief! The change has to do with solving a
REAL problem with student input; the question is whether
anyone anywhere has answers that REQUIRE students to
enter two (or more) consecutive unshifted supers or subs,
not whether a theoretical "flexibility" is being lost.
   The "canonical" form is intended to make equivalent
screen displays have an identical form when the character
string is passed on for judging. Please note that the
appearance on the screen is DIFFERENT if 2 superscripts
are used rather than 1; to keep the meaning and functioning
of "canonical" clean it is not appropriate to make these
strings identical.
   Rather than the current puts or a new specs tag to deal
with this, we want to go to the heart of the problem and
prevent such accidental entries while still permitting the
multiple level 'scripting for those who intend/need it.

---------- response 14
  12/11     09.29     friedman       csa

Seems to me the cleanest fix to the problem is to have
the compiler, or whatever it is that interprets the
_algebraic_ value of an input, simply ignore consecutive
SUPs, SUBs, shifted SUPs, _and_ shifted SUBs, since they
cannot have any algebraic meaning anyway.↓
As to their difference in appearance, that is part of the
point! None of these four "characters" should be ignored
for screen display purposes, because two of them DO appear
different than one; they should only be ignored in their
_algebraic_ interpretation. But for true student arrow input
of an algebraic expression, the different appearance is
too subtle to bother the student with.

---------- response 15
  12/11     13.24     judy       pso

Yes, I have lessons which REQUIRE the students to type





two consecutive -sup-s. Since this is done using a
MICRO and -force micro-, it is not a typing problem for
the students.

Yes, I COULD change to use shift-sup-sup in my MICRO table.
BUT...
I am using a 20 character micro, and it happens that
there is space to add the three extra characters which
would be required for the shift-sup-sup scheme. However
for an author using 10-character micro's, the addition
of three more characters to his string is significant, and
might make it impossible to make the preferred MICRO.

If the change required "shift-super-shift-super" rather
than "super-super" or "shift-super-super", that adds 6
characters. I don't believe I could do it.

I don't care how the "sup-sup" is judged. My judging could
be modified to accommodate either scheme.

---------- response 16
  12/11     13.33     judy        pso

In case you are puzzled by the "3 extra characters" of
the previous note:

My current MICRO string is
        SUP SUP a (0676701) which is 3 characters

The proposed scheme would require, since I must return
to my "base line"
        SHIFT SUP SUP a SHIFT SUB which is 6 characters
or
   SHIFT SUP SHIFT SUP a SHIFT SUB SHIFT SUB
        which is 9 characters

If I want a character from the alternate font set, I
would require under the 2nd scheme:
FONT SHIFT SUP SHIFT SUP a SHIFT SUB SHIFT SUB ACCESS FONT
which is 12 characters and forces me to use a
20-character micro.

---------- response 17
  12/12     13.12     shirer       s

   Although I can't imagine why I would ever want a student
to type in SUP SUP in a response, I believe that

1) it should still display two supers above line

2) the -calc- machinery should ignore extra SUPs or SUBs
occurrring one after the other

c) the -calc- machinery should not change its handling
of SHIFT-SUP.

Hows that for a SUPer SUBmission?

---------- note 347 save note
  12/09     21.33     dowling       ames





        Would it be possible to line notes to the editor in
a manner similar to that used for aids? (Possible key to
activate: N). The saving and inserting of saved notes would
be greatly facilitated by being able to move directly from
the source listing page to notes, and after saving a note,
to be able to return to the lesson without having to re-type
the lesson name. The present exit after saving a note would
be fine if the lesson name could be in n1. Pressing NEXT
would then give the block-listing, and SHIFT-NEXT would give
the last edited block.
        Your consideration of this proposal is appreciated.
Thanks. ¬↑ˇW¬≤↑ˇC¬≤↑ˇD

---------- response 1
  12/10     07.55     k mast       p

       Sounds like a good idea to me.

---------- note 355 data page
  12/10     09.28     jim g       reading

Could we _please_ have a SYSTEMS response to general notes
numbers 21,114 and 282 here. The content of those notes
dealt with being able to see some _portion_ of the data page
of a file without having the inspect code or the file having
a blank inspect code. The logic behind the request is that
its implementation would facilitate author¬A¬Dauthor
communication on problems relating to a file.

Please do NOT reply to this note unless you are a system or
system policy making person!

---------- response 1
  12/10     11.10     b sherwood       s

There is a general feeling that file information is
private. One might say that an author could/should set
a bit saying that he is happy for people to look at
the basic file identification info. I might say that
such positive action can also be obtained by the author
writing his sign-on as well as his name on the title
page of the lesson.

---------- response 2
  12/10     11.13     avner       s

This is one of those areas in which there are conflicts of
goals and interests which make quick decisions difficult.
As a working policy, it has been held in the past that
lessons which are truely ready for wide use should have
current contact information on the student-mode title
display. On other lessons, if the author doesn't want
people to use a lesson yet--it is no one's business who
said lesson belongs to or who should be contacted about it.
Lesson identification will clearly have to be a matter of
policy on "published" lessons. Al Avner





---------- response 3
  12/10     11.35     jim g       reading

Thank you for the definitive answers.

---------- response 4
  12/10     12.37     golden       s

To carry the argument further,(and probably beyond the point
of reasonableness), you are not supposed to look at a lesson
unless you have permission from the author! Therefore, in
theory, you don't need to ask who the author is.

---------- response 5
  12/10     19.41     gilpin       peer

This seems to me to be a very peculiar situation. Four times
in this batch of notes, different authors (from different
courses) have requested open access to information about the
author of a lesson. The same request has been made often in
the past. There is virtual unanimity in favor among authors
responding. But systems people say: "No, that information
should be private." No real reasons are given. It's like
when the FDA said: "No, you can't but fish flour...you
wouldn't like it." PLATO is so open in some respects that
this attitude in this case is strange. I say, if an author
doesn't want you to see his lesson, he can dump out anyone
who shouldn't be there. Most authors don't care. Nor, appar-
ently, do most object to their authorship being open. They
should not have to take explicit steps to make it so. I
think open access should be tried.

---------- response 6
  12/11     09.35     friedman       csa

I have asked for this question to be on the agenda of
the UI Plato User's Committee; I think it needs a thorough
airing!

---------- response 7
  12/12     20.00     lombardo       ed

BRAVO for open lessons! It shouldn't be that lessons
are closed unless approved by the author, but rather
they are open unless the author chooses to close them.
It is for the former reason that data pages should be
accessable so individuals will know who to contact to
get permission to view a lesson. If you close the lesson
and the information as to who you can contact to open the
lesson, then I say scrap the lesson too. There are many
lessons I have viewed that I recommend to others so a
good lesson becomes known, open or not, and individuals
will try to view them. Certainly every author should
have the right to do with his lesson what he wants but
let's take a positive approach rather than a negative one.
The wealth of information on PLATO should be made available





to the many users of the system.

---------- response 8
  12/12     20.49     warrens       uw

Well now,
    I have added to the pro-open discussion of lessons.

However I have a personal case which occurred about a week
ago, in which I "destroyed" anyone going into my lesson
who did not have my sign-on, and intentionally put as much
security on the lesson as I could..... the reason was
that I was running an experiment and things just could not
be changed between my sessions. I doubt anyone would
have stumbled onto the lesson, or done anything to it if
they had, but I took the precautions. This type of thing
is why I suggest openness in the hands of the authors,
possibly by the process I gave in my response to gn.114.

I still vote for access to author responsible for lesson.

---------- note 356 erase
  12/10     09.34     bird-dog       cha

Will size i writing and erasing ever be repaired so
that I can erase line 32 without changing the erase
to a size 0?
   The following code will not erase line 32:
size    1
at      3110
write   Warning: The surgeon general has determined
        that cigarette smoking is dangerous to your health.
at      3110
erase   54,2
*It will erase if the erase is changed to size 0

Will you guys fix this?

---------- response 1
  12/10     11.18     banks       mfl

Why do you need to use a size 1? Size 0 will work
just as well (and faster) !

               Fred Banks

---------- response 2
  12/10     12.42     al       mfl

For emphasis, Fred.

---------- response 3
  12/10     13.09     bird-dog       cha

I was using size because I want the writing to
appear more slowly. The erase problem is the same
for any of the size specs tho, regardless of wheather





it is size 1 or another size.
   Please stick to the subject, that of not erasing.

---------- note 362 delet aids
  12/10     11.29     jim       pso

after the first of the year, the references to the system
defined variable "error" will be removed from AIDS.
NOTE: "zreturn" has replaced the function of "error" in
        TUTOR. all values of "zreturn" and commands which
        set "zreturn" have been in AIDS for a few months.

in a few days, the documentation on the old system router
(named "router") will be replaced by information on the
new system router (named "mrouter").
NOTE: after this academic term, the "router" version will
       _no_ longer be available.

for information on the value of "error" or the old "router",
please contact me after these are removed from AIDS.

                    jim ghesquiere

---------- response 1
  12/10     11.54     keith s       matha

Since converting lessons which use the system variable
error is improbable (correct?), would it be possible
for the system to notify authors whose lessons use error
before the variable is made obsolete?

---------- response 2
  12/10     19.55     blomme       s

   Jim only announced the removal of DOCUMENTATION
regarding "error" returns; we have no present intentions
(that I know of) to stop "error" from being set. If and
when we do decide to stop setting it, a conversion would
indeed be very nearly impossible and we would have to
do it by fiat--giving a warning a few months in advance
(and maybe finding all lessons with such references and
providing their owners with special notification).
   You will certainly be doing both yourself and the
system a favor, though, if in your "spare time" (which we
all have so much of) you fixed all your lessons to
eliminate the use of the "error" returns.

---------- note 373 term time
  12/10     13.24     ian       conn

ok guys, here we go with another system term suggestion:

        TERM 'time' and/or TERM 'date'

right, i know that you can get this from the main notes





option page, the author mode page, and the block
directory page....BUT i am plain lazy.

what say?
ian

---------- response 1
  12/10     14.30     midden       p

A general feeling is that things like this should be
considered to a great extent before putting them in.
They are  written in tutor, and must be one of the first
lessons condensed when the system comes up.
M.Midden

---------- response 2
  12/10     14.45     temmerman       com

yean,i know it's trouble for the programmers.
        But, the would be especially helpful for the 
student signons on terminals where no clock is handy.I
know it's a matter of convenience, but it could help in a
stdent's choice of a lesson.

---------- response 3
  12/10     19.34     ian       conn

thanx for the support, and i see your point, marshall.

heck, it doesn't seem to me like the tutor code for
it would be all that extensive.

---------- response 4
  12/10     20.07     fritz       ames

  Can't it be put in.  say, syslib?  It's condensed anyway,
isn't  it?

---------- response 5
  12/10     20.35     midden       p

So, the condense gets longer. The load of the system gets
a little more frustrating.  Yes, that would be were it went,
if/when someone gets convinced that its important enough
to worry about.

---------- response 6
  12/11     10.20      rader    s

Marshall's point was that indeed it would go into syslib,
indeed it would get condensed every time--and obviously
one such feature doesn't take long to condense.  The
problem is that every user (n>>1000) has his own idea of
some neat feature that should be added.  The historical
accumulation of neat features is why it now takes so
long for us to get up after the first "press next to begin".





We really must limit additions to those that are really
needed and useful to a broad class of users.

Maybe this suggestion will be one such.  But we really
don't want to add every feature suggested overnight.

---------- note 374  calc7;refs
  12/10     14.08     bennett       phys

Would the author of lesson "calc', please identify itself
(himself, herself?).  I would like to include this lesson
in an index available to physics students. Leave me a pnote.
   To authors in general--your favorite lessons would
be used more if people knew how to contact you. A reference
name on the title page of lessons would be helpful since
users without inspect codes cannot inspect lesson data
pages to get this information.

---------- response 1
  12/10     15.51     bennett       phys

Thank you--author has been identified.

---------- note 377  down  time
  12/10     14.53     reece       pea

Is there any information yet on down time during the
Christmas vacation period?

---------- response 1
  12/10     15.49     fay       o

According to the bulleting board ("B" from the author
mode page)  the system will go down Sat., Dec. 20 at noon
and will be down until 8 AM on Sat.,  Jan. 10.

---------- note 379  the 3 effs
  12/10     15.05     socrates       faa

Waddya get from the FHA!.....A loan,  a loan,  a low cost
loan.

Waddya get from the FFA?......A cow,  a cow,  a real  live
cow.

Waddya get from the FAA?......Best wishes for a Merry
Christmas and real short condense queues for "76"

Pete Kochis  Tom  Schley
Jack Veazey  Jim  MCKenzie 
Craig Burson Frank Fitzgerald
           Bill Vlaming

---------- response 1
  12/10     20.29     blomme       s





 In-"f"-able!

---------- response 2
  12/11     08.55     jan       park

 ¬I≤ˇˇ¬I
 ¬/≤/≤≤---
 ˆ¬I≤ˇ¬I
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

     ...and they went cut among the toiling
        
        multitudes to spread (a)  good cheer...

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *


thanks for the thought...now go control some airplanes

---------- note 382  stop1  pswd
  12/10     16.15     michael     english

It would be so nice if the password routine on the signon
sequence accepted STOP1 as well as NEXT. Now STOP1 kicks
you back to the beginning.

---------- response 1
  12/10     16.25     dave       infe

        Yeah.

---------- response 2
  12/10     18.10     stone       pew

i must say that i am used to  :

stop1,name,stop1,course,stop1,password,next

making stop1 available would be very helpful,
and heartily welcomed...if anyone on systems staff
has a few extra minutes(?) please make this change..

Scott Stone

---------- response 3
  12/11     03.13     simons       cs317

second the motion.....  

---------- response 4
  12/12     20.04     lombardo       ed





agreed

---------- note 385 inhib from
  12/10     16.42     scheibelhut       css

Consider a student who signs into the router "csrouter".
The router jumps him out to lesson "a". Lesson "a" jumps
him out to lesson "b". Now consider the effect of the
following units in lesson "b":
*
unit    A        unit B           unit    C
inhibit from     inhibit from     inhibit from
jumpout csrouter jumpout (router) jumpout q
*                *                *
unit    D        unit    E        unit   F
inhibit from     inhibit from     finish fini
press   next     jumpout router   press  stop1
end     lesson   *                pause
*
unit    fini
inhibit from
*
Continued in response...

---------- response 1
  12/10     16.45     scheibelhut       css

unit    A        unit     B       unit    C
inhibit from     inhibit from     inhibit from
jumpout csrouter jumpout (router) jumpout q
*                 *               *
unit    D      unit     E         unit   F
inhibit from   inhibit   from     finish  fini
press   next   jumpout   router   press   stop1
end     lesson *                  pause
*
unit    fini
inhibit from
*
When the student returns to csrouter through units "A" or
"B" the student is "from" lesson "a". Exiting through units
"C", "D", "E", or "F" results in the student being "from"
lesson "b". I cannot believe things were planned this way.
How about making -inhibit from- work for all exits from a
lesson?

David Scheibelhut

---------- response 2
  12/11     09.48     friedman       csa

As the author of csrouter, I am most interested in this
problem, and encouraged Scheibelhut to make the experiment
he has reported.↓
Basically, it would seem that -inhibit from- is ignored
for a -route-d exit from a lesson.↓
What say the systems staff, please?

---------- note 390 cstart





  12/10     19.40     kane       phys

unit   test
calc   n1‎1
cstop
cstart
       n1‎2
show   n1

In this example the value of n1 shown is 1.  Lines with 
blank command fields immediately following a -cstart- are
apparently treated as a continued -cstart-.  I think it
would be helpful if the condensor at least flagged these
lines as errors (assuming it can't treat them as
continuations of the last command in effect before the
-cstop-).

---------- response 1
  12/10     19.57     blomme       s

   There are some problems here that indeed should
be fixed if possible;  thanks for the reminder.

---------- note 396  catalog
  12/10     20.16     blomme       s

   The comments/suggestions/criticisms regarding lesson
data displays have now gone on so long and to such
absurd extremes, that perhaps some perspective is needed.
   Indeed, there IS a basic problem hidden amongst those
comments, and that is the problem of finding out what
is available on the system on different topics, the
exact content of apparently relevant lessons, who authored
them, and providing access to such lessons if they
prove appropriate.  Currently such information is spread
amongst sample,  topics, catalog, and authors (and there
is no way simply to "select  a lesson for examination
or for usage by students).  What is needed is a single
lesson that provides access to a large cross-referenced
catalog in much they way that the aids lessons now do
for information on the system and the TUTOR language.
... (continued)

---------- response 1
  12/10     20.27     blomme       s

   This has long been recognized as something that
must be done and indeed quite a lot of thought and effort
has already gone into the problem. Diverting ANY time
from this major undertaking to fool with access to
lesson data info on lessons which have inspect codes
ins really inappropriate.  Trying to access information
regarding a lesson because somehow you happen to know
that the name used by the system to find it is "abczilch4"
is really primitive and ineffective.  We want access to
be much more open and much simpler for everyone--and we
are sorry that such a facility is not yet available, but





the real problem here IS being thought about and worked on.

---------- response 2
  12/11     01.02     gilpin       peer

You seem to be directly contradicting what Bruce, Al, and
Bill had to say in responses to GN356. Probably that is
because you are referring to the Catalog for Published
Lessons, while the various user requests have been for ac-
cess to the authors of any and all user lessons whatever.
If that's so, then the two kinds of access are complement-
ary, and somewhat separate issues. Users are simply report-
ing that they keep finding themselves bumping into a gratui-
tous-seeming barrier, and they would like to have the bar-
rier removed. To me, that seems quite a reasonable and
valid request.
        Even if I've misunderstood you, and the Grand
Centralized Solution (effort = 97) will completely displace
the modest decentralized solution (effort = 1.3), it would
not seem appropriate to insist on waiting for the GCS unless
the GCS is going to arrive rather soon.

---------- response 3
  12/11     09.07     tebby       pso

Please be patient waiting for the "super" programs which
give you all the information about lessons on the system.
We are progressing. "catalog" and "authors" are being
combined with better information retrieval capabilities,
all to be accessed in the student mode as one more step
towards the "ultimate". Dave Fumento has been working on
this program since early summer and it will be available
early in 1976. This is not the same thing as the "catalog
of published lessons" which we are hopeful will be forth-
coming sometime next year too. Meanwhile keep on sending
me "pnotes" for information and I will promptly get it
for you if I can.
Tebby Lyman

---------- response 4
  12/11     09.52     friedman       csa

I have asked for the question of lesson data page privacy
to be on the agenda of the UI Plato User's Committee.
I think it needs a thorough airing.

---------- response 5
  12/11     12.54     jones       mcl

I think that the Northwestern Univ. Hypertext/Tutor system
was developed as a solution to this problem. Is there any
possibility of using any code and or concepts from that
system?

---------- response 6
  12/12     08.58     mont       csa





What about allowing an author to set a flag on the data
page which will allow others to see his data page, without
knowing the inspect code?

---------- note 397 ?termop?
  12/10     20.19     dave       infe1

       -termop- cries from the depths for existence!

---------- response 1
  12/11     07.56     john r       reading

nested termops?

---------- response 2
  12/11     12.47     jones       mcl

system terms seem to be 'termops', so some kind of
mechanism already exists along the requested direction,
though it is probably a collection of special cases
right now.

---------- note 401 line-opt
  12/10     21.19     hagerman       ames

     It would seem to me to be a good idea to have a lineset
inspect option from the author mode page.

---------- response 1
  12/10     23.05     fortner       p

it will be in in a few days (after finals)....

---------- note 402 platointro
  12/10     21.34     siegel       peer

Some instructors may be interested in lesson "pp63".
This lesson is an introduction to using PLATO. It is
a modification of "eslhelp1" written by Tom Volpe and
John David Eisenberg. The original lesson was designed
for foreign students. "pp63" was designed for a general
adult population.

Marty Siegel

---------- response 1
  12/11     10.02     siegel       peer

The modified lesson is now in file: helplesson

Requests to use this lesson should be directed to
Roberta Stock (roberta/mfl).

---------- note 408 varian use





  12/11     08.51     fay       o

***NOTICE***

We are trying to optimize the use of the varian copier
in the few remaining days before the scheduled down time
(Dec. 20th). In order to do so, we are asking LARGE users
(¬>30 min. per session) of the varian to please come in to
the operator's office (rm. 166 CERL) and sign up for a
scheduled time. Please come in to sign up only between
8 AM and 5 PM weekdays.

Between now and the 20th, the policy on use of the varian
is as follows: An unscheduled user may use the copier
during a scheduled time (the op on duty can tell you if
someone is scheduled at a particular time), but if the
scheduled user appears, the unscheduled user must vacate
the terminal.
***Continued....LAB***

---------- response 1
  12/11     09.10     fay       o

We discourage users needing only one or two copies from
signing up for a scheduled time. We would appreciate it
very much if such users would merely attempt to take care
of their copying needs early (like the next few days).
Unfortunately, the large users (projects and certain admin-
istrative users within CERL) cannot copy many things
"early" by the very nature of their materials.

We regret this situation and will do our best to attempt to
accommodate all users who need varian copies.

PLEASE: copy early, avoid the last minute rush. There
may not be much we can do for you if you wait until next
Thurs. or Fri. to copy materials you absolutely need before
we go down.

We are planning on having an on-line scheduling system
late in January to avoid future scheduling problems.

The operating staff.

----------- note 418 data1 code
  12/11     10.40     warner       iu

        Is there any reason why, when in INSPECT ONLY,
pressing DATA1 will not allow you to enter your security
code if you are in a block? I realize that during editing
DATA1 changes the type of source blocks, but since this
function is disabled in INSPECT ONLY, is there any reason
why DATA1 should not be switched to send you to the
security code page?

---------- response 1
  12/11     11.17     fay       o





I, for one, would prefer to keep the function of DATA1
consistent between the inspect and change mode. Even though
the key is currently disabled while inspecting the contents
of a particular block, giving it a "dual function" at that
point in the editor would be confusing, I think.

Is it that difficult to back out to the directory page to
enter your security code?

---------- response 2
  12/11     11.47     friedman       csa

Well, it IS damned annoying to go into a lesson, perhaps
with NEXT1 (where you don't see the INSPECT ONLY message),
do some searching to find where you want to change
something, then discover you can't; so you back out of the
block, use DATA1 to enter security code, and lo and behold,
the editor re-initializes itself to display part 1 and to
forget any searching directives you might have typed in.
So you now do your search all over again....↓
It would help if the editor did not re-initialize when it
changes from inspect to editing mode.

---------- response 3
  12/11     12.22     hody       med

it is also extremely confusing for new authors when
they are in a block and can not change anything and
dunno that they are inspect only... would it be possible
to print ** INSPECT ONLY ** at the right hand field
in the top line of the line display?
*
i realize it is extra overhead and plotting time....
so another possibility would be to have the system wait
until an editing directive is received, then if in
inspect only mode, write ** Sorry, you are only inspecting
** or some such at the top of the screen...

---------- note 421 xmas gift 
  12/11     12.05     warner       iu

Merry Christmas to all from warner/iu!
↓↓ And with this gift goes a Christmas present to
all course directors everywhere...
a quick statistics reader that allows you to input the
name of any given course and have the statistics of its
students printed out for Varian-copying...similar to
the statistics collector in course records, but capable
of collecting all statistics in one place. Coming soon,
a formatter which collects  the statistics into common
for printing in special format!
↓↓ Open this gift in lesson "readr"!

---------- response 1
  12/11     13.05     fay       o

Looks good, Silas.





I might add, for the edification of users in general, that 
Silas called me and asked if it would be possible for an
operator to sit at the varian and go thru the records for
users in each IU course, getting copies of the student
stats. I suggested we would be happy to, but it would be
much more efficient to write a simple lessons, using
readset and readr stats to get the info into a condensed
form without the necessity of entering each student record
in each course.

Bravo!

---------- response 2
  12/11     15.51     berger       mfl

I wrote a similar lesson a couple of months ago:
instructions to "use" lesson "deleted" with your
own common are in lesson "routines", block "deleted".

"deleted" has more display options than "readr", and
I am working on an option to make "terminet" printouts.

Incidentally, there is no inspect code on "deleted"
so you can see that I don't store course codewords.

Silas, I suggest that you do the same, for paranoid 
course directors (and what IS a run code?)

---------- response 3
  12/11     23.35     warner       iu

     Now lesson "readr" has been extended...
Typing in a dataset name and block limits will allow you
to format multiple course statistics in a dataset for
printing using the
*list special
option. Ignoring the dataset name causes "readr" to
simply generate displays for Varian-copying.
     Also, for paranoids, I have followed the
suggestion in the previous response. Please do not
laugh at my subroutinized code.

---------- note 433 deleted
  12/11     15.12     golden       s 

Note about new game deleted at the request of its author.

---------- response 1
  12/11     15.13     waldrop       ed317

Thanks!!!

Kent Waldrop

---------- response 2
  12/12     09.03     mont       csa





Thanks for letting us know why a note disappears. It
helps keep finals time paranoia at a lower level.

---------- note 437 R-option
  12/11     16.13     wood       medsiuc

Just for my own interest, what does the 'R' option
do on the author mode page?
Ray Wood 

---------- response 1
  12/11     16.14     little t       research 

It writes the ECS copy of a common back to disk.

Todd

---------- note 440 step 30
  12/11     18.28     dave       infe

     A current feature of the _step_ mechanism doesn't seemoptimal.
     If you do, say, 'step 30,' and somewhere within the
scope of the 30 steps there's a pause, the 'step count'
isn't zerod, so after your keypress the remaining quota of
steps is taken. The same applies to judging commands.
     I would have expected 'step n' to stop after n
steps, or when a keypress is required, whichever comes
first; and would much rather have the extra control this
would provide than save the few keypresses to enter another
'step' directive.

---------- response 1
  12/11     19.55     fisher       conn

I would love it if step worked that way!!!!!!!!!!!
Many times, I have stepped through a long procedure
in which I knew what was happening, just so I could
find out what PLATO was doing with my keypress..........

PLEASE! SOMEONE! Put this feature in!!!!!!!!!

your needy fishswine 

---------- response 2
  12/11     21.29     warner       iu

I disagree.
It seems to me that if you had the step-count feature
enabled, it should carry you through the number of steps
you asked for. Disabling the step-count could lead to
frustration if you have the opposite effect...a -pause-
loop which you have to pass through.
Suggested compromise:





How about, when step mode "waits for key" while there
are still more steps to pass, the message gives a little
more information, such as the number of steps since the
count was initiated?

---------- response 3
  12/11     23.19     dave       infe

     Fine; better yet another option, Go, which acts as
I request.

---------- note 441 datapage
  12/11     19.10     wegner       phar

there seems to be a conflict of interests between the
policy makers and the authors of this system(unless the
authors objecting to making the data page public have kept
their objections private which makes sense?). It seems
the policy makers keep on wanting to connect the data page
with published lessons and the authors want to connect the
data page with easier communication with authors and student
i feel easier communication among authors is really importan
and shouldnt be held back by published lessons. there are
stages of developement between the writing of a lesson and
the finished(published) product. During this time easy 
communication is rather useful. i am rather surprised there
isnt a lesson where authors can announce their lesson as
being ready for testing and have it stored like general note
but the key word would be a pair of words ie suggestive 
title and lesson name.
WEG 

---------- response 1
  12/11     19.13     wegner       phar

ps. i sure am glad that i dont have to ask permission of the 
author each time i want to look at a new lesson. i wouldnt
get give demos at a very fast rate or explore the system
at a very good rate either.
WEG

---------- response 2
  12/11     19.22     golden       s

The system staff are very concerned about easy
communications among authors and regularly develop new
features to improve those communications. We are working on
a way to send comments about a lesson to its author even
if you don't know who the author is.

We are also concerned about privacy and protection of 
copyright possibilities and similar problems. Finally,
we are busy, and don't always give the same priority to 
tasks that you do.

P.S. Who said you should look at a new lesson without asking 
permission?





---------- response 3
  12/11     19.26     berger       mfl

I agree with Mr. Golden. Even if you COULD see the lesson
data page without the inspect code (and I would like to see
it implemented), that doesn't give you a general invitation
to use my lessons without permission.

---------- response 4
  12/11     19.27     wegner       phar

if every author had to ask permission of the author to look
at his lesson every authors pnote file would be full. And
i would feel sorry for people like stan smith who has
dozens of lessons and would most likely get several requests
to the point that giving permission would be an outright
hassel.

---------- response 5
  12/11     20.35     john r       reading 

Nevertheless, although it is currently easy to use lessons
without an author's permission, you should not assume that
you have any _right_ to do so. Part of Golden's point is
that implementing what you request could imply to future
users and future courts that privacy and ownership are not
inherent features of PLATO authoring.

As for authors speaking up, I'm the head of a group of
people who have developed 1600 lessons spaces
full of material. When we put inspect codes on lessons,
it's because we want them private. I vote against your
idea.

---------- response 6
  12/11     20.46     golden       s 

Despite the impracticality which you describe, the lessons
I am working on are like the papers on my desk-- you have
no right to look at them even if an opportunity presents
itself unless and until I either publish in some sense or
invite you to inspect them.

Even a lesson without inspect code should not be entered!
You wouldn't think it right to enter my home uninvited just 
because you found the door unlocked.

---------- response 7
  12/11     21.08     wegner       phar 

i think have been misread here. there is a BIG difference
between looking at a lesson as a student and using or look
ing at the lesson as an author. I dont think i would like 
people using some of my code either. but i would like to
make sure that as many people as possible know my lesson 
exists so i can get feedback on it and i would like them 
to know how to get a hold of me as easily as possible with





a mininum amount of hassel. Basically what good is a product
if no one knows about it and has trouble finding out who
wrote it. who knows they might even want to have that author
write some lessons for them. Seems to me that easy access
to certian very specific types of imformation is important.
ie a lesson of a certain type exist and the name of who wrotee
it. by the way how many of you have explicity asked
permission of an author to run his lesson and have done it
for every new lesson you encountered.

---------- response 8
  12/12     07.58     john r       reading

The question is NOT what is convenient nor is it how YOU
would treat YOUR lessons. The question is what policies
afford each author the greatest authority over access to
his materials. With the current scheme, one can allow
exactly as much access (for authors and students) as one
wishes by advertising through notes or pnotes.
With the schemes you propose, your own very lax
standards become the default, and nothing more secure is
possible.

By the way, I assume this whole discussion began because
you couldn't find out who wrote a lesson. Has it yet
occurred to you that that author may not want to be
bothered by you?

---------- response 9
  12/12     09.56     hinton       ssu

I would like to say that I have never had problems or en-
countered major delays in getting the author's name and
course from tebby/pso, or in getting responses from the
author. The only time I was not able to get permission
from an author in time to use his lesson in a demo occurred
when I myself did not make the request soon enough. I am in
favor of access to lessons when it is reasonable, and I
would note that fils such as PAD and others are often used
to announce new lessons. If, however, I have a lesson which
is not ready to be demo'd, or do not want demo'd for some
reason, why should someone else expect that he has a RIGHT
to demo, inspect, run in student mode, or otherwise inter-
fere with my work ?

---------- response 10
  12/12     11.00     deiss       phar
To Paul Wegner:

!Do you think you can take over the universe and improve it?
I do not believe it can be done." (Tao Te Ching)

---------- response 11
  12/12     11.04     deiss       phar

To the others:





"The more laws and restrictions there are,
The poorer people become.
The sharper men's weapons,
The more trouble in the land.
The more ingenious and clever men are,
The more strange things happen.
The more rules and regulations,
The more thieves and robbers."

(Tao Te Ching, Lao Tsu, 600 B.C.)

---------- response 12
  12/12     11.18     michael       english 

The analogy between a PLATO lesson and papers on a desk is
a bit stretched. If lessons are really considered by the
system in that light, why not enforce the "no browsing"
rule by creating a new codeword on the lesson data page
which must be blank for students to run or matched by an
author to be requested from the AUTHOR MODE page?

Instead...

'Twould seem simplicity itself to take a prior suggestion:

Let the author specify whether the data page will be open to 
anyone, perhaps by toggling a bit via DATA from the lesson
data page itself. The default would be as it now stands,
no access. If this is a programming problem, say so; if
it's a policy matter, somehow I just don't see it.....

Personally, I don't like to include my signon on the title 
page of a lesson; it is not my name, but a PNOTE to "Gary
Michael" will be rejected.

---------- response 13
  12/12     11.31     mike b       cornell

      It seems to me that a little common sense and
courtesy is all that's required here. There is no need to
obtain the permission of a author to use an announced les-
son, as far as I can see. It is quite another matter to
discover a lesson by typing random characters at the AUTHOR
MODE arrow. and execute it. I know that, in the lesson I
am currently developing, a person executing without my per-
mission could change stored information that would make my
debugging efforts much more difficult. Therefore, I think
the system policy should stress that you do not use unan-
nounced lessons without obtaining the permission of the 
author. As for the DATA page, I can see no reasonable ob-
jection to implementing the suggested feature--an author
selectable display to be given to people who want to get in
touch with the author. After all, thing would be easier for
tebby.

---------- response 14
  12/12     11.56     friedman       csa

I don't understand the heat that this discussion always





brings out!↓
Personally, I agree with what Gray Michael said a couple
of replies back.↓
This discussion brings out different explanations for the
present situation from different system people. Golden
makes it a privacy policy matter; Blomme makes it a 
practical matter of how to allocate system programmers'
time. Both are presumably true. Both are subject to
discussion. Something like Michael proposes would seem
to me to meet Golden's objections, and I doubt it could
require an awful lot of system staff time to implement.

---------- response 15
  12/12     12.21     siegel       peer

Gray Michael's suggestion is a good one.

It is sometimes undesirable to have title pages
with author information. Most students would be confused
with the message: "contact siegel/peer...."
Many lessons are not of the "stand alone" variety (as is
a lesson for a university course) and therefore, a title
page is often not wanted.
****
If I am debugging a lesson and want to keep other people
out I simply -jumpout q- other people.
I use similar code to keep people from certain courses out.
****
The above holds for when you know the name of a lesson.
Having a general catalog to search for lessons of interest
is a complicated problem. But Michael's idea is a nice
interim solution.

---------- response 16
  12/12     15.42     nate       iumusic

   to continue on with seagel's point: Ever hear of
signon-restricting a developemental lesson while it's in
the developement stage??? I do that with all my just-
started projects. (however, I do put a message saying
that it is not ready yet, and sign it, sometimes with
both name and signon). What's so hard about that,
and making the data-page optional.

Nate

---------- response 17
  12/12     17.10     mike b       cornell

  In answer to you question, see golden/s's reply to
the following note.

---------- response 18
  12/12     20.13     lombardo       ed

   I really fail to see the analogy about walking into
someones house if the door is unlocked. Insurance companies
sure take a dim view of that attitude and so would, I would





bet, a legal copyright hearing. The point is not one of
idealism, but rather one of bare fact. There are several
provisions for excluding people from your lessons and they
are readily available to ALL authors. Should an author not
take the steps necessary to do this then he has essentially
agreed to let anyone run his lesson as a student. I cannot
believe that anyone can seriously believe that an idealistic
point of view is sufficient to solve this problem. You can
not quote points of ethics, morality, etc to the world and
hope for results. You must deal with hard facts and the
fact is that without an inspect code it is open to the
public.

---------- response 19
  12/12     22.20     hanoka       ed

  I agree with wegner, michael, and lombardo.
  Also, I vote that we have our choice via a
  toggle bit to have them see or not, but the
  priority of such a job would be very low.

  mike

---------- note 457 no look?
  12/12     09.52     roper       siu

In the past, I had believed that if a lesson could be
condensed, it was permissible for an author to take a
look at what he could. I get the impression from the last
note that this is not so.
This surprises me. If an author really wants to lock other
people out of his lesson, all he needs are an inspect code,
a pause, a term, and a press stop1. It will work quite
effectively.
Really if I am trying to explore the capabilities of the
system, must I send every author of every lesson that I
would like to access (just so I can see what it does) a
p-note asking permission?

---------- response 1
  12/12     09.53     roper       siu

I believe, that if an author does not want other authors
looking at his lesson, he is responsible for protecting it
(via inspect code, terms, etc.). Should he not protect it
in these ways, any undesired author use of the lesson is
a result of his negligence in not protecting it. And
invitation to enter, failing to protect a bicycle at SIU
by locking it will result in confiscation--by the police.
Analogies are interesting but not always applicable.
   
    Quetzal

---------- response 2
  12/12     11.02     golden       s





Just can't accept the moral code you present. Locks are
used to protect me from the immoral minority. The majority 
respect my privacy and my property out of courtesy. They
do it beause it is right and because they expect similar
treatment. When that ceases to be true, no lock, weapon, or
police force will be adequate to project us.

---------- response 3
  12/12     11.33     mike b       cornell

  It couldn't have been said better, and nothing more
need be said.

    M¬ ike≤≤≤≤ˇB

---------- response 4
  12/12     13.13     s zweig       iumed

  It has been said, however, that locks tend to
keep honest people honest.
   
sz

---------- response 5
  12/12     13.19     al       mfl

May I please have the signon of the author(s) of lesson
edit so that I may ask them if I may use their lesson?
After all this time I would hate to find out that I was
doing something wrong.

Alan

---------- response 6
  12/12     13.24     durkin       siu

Good point.

---------- response 7
  12/12     13.39     golden       pso

No, it is not a good point! There are hundreds of lessons
for which permission to use has been granted to any all
users. Among them are all the lessons in sample, all the
system lessons which have been publically announced (but
only those) and many private lessons which have been
announced.

A great many lessons have been opened only if you use an
author-provided multiple record.

Everything else should be assumed private and not viewed
without explicit invitation.

---------- response 8
  12/12     14.46     al       mfl





Could we please have the number of the old system features
note that announced *edit*?

---------- response 9
  12/12     15.32     golden       s

You have it backwards! The burden is on you to be sure you
have permission to use a lesson you wish to use. Now do
you remember all the times you have been invited to use the
editor and its various features?

---------- response 10
  12/12     16.11     al       mfl

I hate to belabor the point, but in your last note you just
said that it was MY responsibility to get permission to use
a lesson, and then just turned around and asked me if I
remember all the times I was invited to use it.

In high school, we all used everyone else's programs without
asking. If the author wanted to keep other people out, he
could then protect his lesson from other's useage.

The DEC-10 has an even better system. Each file has a
protect code associated with it allowing control over what
other's may do to the file.

On either system, a file is there for everone's use unless
otherwise stated.

Alan

---------- response 11
  12/12     16.44     charlie       mathw

I agree with alan and others that one should be able
to use (in student mode) lessons by default. And given
the ideologies of PLATO (my interperatation) this is
the way it should be. Attributes are changable, so
if you do not want your lesson in the public domain,
than insert the simple code.

---------- response 12
  12/12     19.36     david       hebrew

Re: This whole megillah concerning datapages,
"public announcement," and cetera

  *
 ***
*****
ˆ¬I
↑ˆ↑ˆ↑ˆ¬I

May your hearts be filled with love and joy
This Holiday Season





---------- response 13
  12/12     19.51     mike b       cornell

  Hmmmm..... How did THAT bit of good cheer get in
here?!?

  Isn't it reassuring to see that, no matter how
tough the fiber of the argument (debate, for you gentle
types), something like this can always seep in?

  Merry Christmas to all, and to all a good night!

  M¬ ike≤≤≤≤ˇB

---------- response 14
  12/12     20.31     lombardo       ed

Bill;
   While I will certainly agree with your philosophy
regarding the state of the world when you HAVE to lock
up everything, and concur we will be in a hell of a mess
when it happens, please let me point out that because
this whole discussion comes up time and time again, and
everytime it drags out numerous responses, it should be
obvious that not all people BELIEVE the same things. What
you say is true but if another individual doesn't agree
and feels just the opposite then he will not necessarily
be willing to play by your rules. The world is in a hell
of a mess, but let's do whatever is necessary so that
everybody plays by the same rules rather than a few
gentleman standing around getting trounced by the Mongol
Hordes.

---------- response 15
  12/12     22.27     hanoka       ed

  I agree that unless the lesson is guarded against
  unwanted entry, anyone should be able to look at
  it. Also, from what I've been hearing, the
  majority agrees with me, and the majority doesnt
  ask everytime to go into a lesson.
  AWAY WITH OLD≤ MORALS AND VALUES!!!!!
  Accept the hard, cruel world as it is.

  mike

---------- response 16
  12/13     08.43     stan smith       chem

Unless the lesson is in "sample" I always ask the author
if I may look at, show, or use a lesson even when such 
permission has been granted for that particular lesson
in the past.

Simply counting responses to NOTES does NOT give a





good impression of how the system is or should be used.

---------- response 17
  12/13     09.43     deiss       phar

But then now do we get a "good" inpression? I agree, but I
also contend that leaving these decisions to a select few
at U of I (without consulting the rest of us) is also going
to result in erroneous impressions. Would you go along
with a site director poll or something?

The biggest problems of our way of life are often traceable
to the fact that someone somewhere was not allowed to have
an input to a decision. There can only be majority rule
if everyone has a voice. For many of us that voice is
NOTES.

---------- response 18
  12/13     10.11     al       mfl

But this whole argument started over the fact that we
CAN'T ask the author for permission to use his/her lesson
because it is the SYSTEM'S policy that you shouldn't use
a lesson until you ask the author and that you shouldn't
be able to find out the author except by executing the
lesson and reading it on the title page. Still no one has
come up with an argument as to why we shouldn't be able
to inspect the data page to find WHO TO ASK!

By the way, notes is almost full.

*** end of notes ***