+++ plato iv group notes +++

Public Notes

notes beginning march 8, 1976

file pbnotes7 printed at 5:29 pm on june 12, 1976

---------- note 0 -delay-
  03/08/76     08.40     roman       med

The aids description claims that delay allows precise
delays. Does anyone know how precise? Can I count
on a delay within 1 millisec of that stated?

---------- response 1
  03/08/76     09.16     koning       csstaff

Delays are in steps of 1/60 of a second, because they are
done by sending 'null' words to the terminal, and one
word is sent every 1/60 of a second.

---------- response 2
  03/08/76     09.19     golden       s

It is not nearly that precise. Once started, the duration
of a -delay- should be correct to the nearest 1/60 of a
second. But the exact starting time is unknown, and, if
line errors occur during a delay, the timing can be thrown
off by almost a second.

---------- response 3
  03/08/76     11.09     tenczar       s

and just be make crystal clear...-delay- if for causing
"exact" delays in terminal graphics (such as animations)...
NOT↑≤≤≤6__ for doing any timing delays within programming

---------- response 4
  03/08/76     14.34     bowery       comm

And to be even more precise: The terminal is not sent
"no-ops" (as described in X-15) with the -delay- command.
"atx" 's are sent with a value of the current -wherex- (or
is it "aty"? or has this been changed?).

---------- response 5
  03/08/76     18.38     fritz       amesrad

 It was "atx" last I knew...

---------- note 15 more lines
  03/08/76     12.41     michael       matha





AUTHOR REQUEST DEPARTMENT:

Could lesson " search" (X from the author mode page) be
modified to allow the user to specify how many lines
following the "found, line are to be shown?

_sample_ use:
There are 6 occurences of 3 critical lines in 7 lessons, all
of which need to be checked for correctness. One would
normally block search for a keyword in the first line, in
each lesson, six times, and note whether the lines are
correct. If one could show several lines in lesson "search"
one could avoid the costly block-search, reduce the searches
needed in lesson "search", if used (1/3 as many), and save
gobs of time.

The default would naturally be one line, and a small limit
(say four) to the number of lines that could be shown would
be quite reasonable.
Any hope?

---------- response 1
  03/08/76     13.24     n swenson	      mtc

Here is my vote. We have the same sorts of problems.

---------- response 2
  03/08/76     15.22     elst	      rhrc

Yes, we could us that feature as well.

2c≤/

---------- response 3
  03/08/76     16.19     maggs       law

While you are fixing up the search, how about an option
to search just for commands or to search excluding the
commands. Now when I want to search for the word "use,"
I get a lot of pa_use_ commands. Conversely, when I want
to search for a command like "set" I get everywhere in
the text of the lesson that that word is used.

---------- response 4
  03/08/76     21.04     bt williams       rhrc

This feature would be very useful to us - even more so if
one could choose to display an additional line or so _before_
_or_ after the target line containing the search object.

---------- response 5
  03/08/76     21.16     ld francis       mtc

peter,





to get rid of _some_ of those searching problems, add spaces
to fill out an 8 char search---"use_____" will exclude
"pause " etc. doesn't take care of every case you
might want, but does help for searching for commands.

---------- response 6
  03/08/76     21.44     elston       rhrc

He explicitly stated he was looking for the word "use" not
the command. Therefore your suggestion although useful in
other cases will not help in this one.

2c≤/

---------- response 7
  03/09/76     01.10     avner       s

On the contrary, the suggestion to use spaces is quite
appropriate. A search for _use_ will pick up only
instances of the word "use" (in light of recent changes,
perhaps the word "at" makes a better example of using this
procedure for distinguishing between words, commands, and
embedded letter sequences).

---------- response 8
  03/10/76     02.19     elston       rhrc

I agree that "_use_" will pick up _some_ of the occurances of
the word "use", however, it will not find the ones at the
beginning or end of lines or those followed by punctuation
without entering a long list of possible permutations of
punctuation and "use". But we digress from the issue.

2c≤/

---------- response 9
  03/10/76     12.50     michael       matha

Digressions aside, could we have a system response about the
feasability $or$ implementation of the request made in the
original note?

Michael

---------- note 24 - group-
  03/08/76     14.47     ruane       ll382

Can the -group- command be set up so the user can specify
his own delimiter to seperate the elements of the group?
I use a comma as part of a group, and it would be much
more readable if I could use something else besides

group  punct,.,.,.-    $$ etc...

Thanks.





---------- response 1
  03/08/76     15.40     berger       mflu

It would certainly be nice if either semicolons or
writec delimiters were universally acceptable.

Semicolons are a better choice in terms of space
savings, but the writec delimiter makes code more
readable because it takes one out of alternate font.

---------- note 27 fore/back
  03/08/76     15.32     bowery       comm

"As long as there is a command -backgnd-"

could there be a system reserved word telling whether to
us it or not, since the only reason -backgnd- is used
is to minimize processing time (-clock-)?

ie:

unit        foreback
branch      zbfgnd.x,1fore
backgnd
goto        q
1fore
foregnd

This could be updated periodically by one of the pseudo-
terminals depending on the amount of idle cpu time (?)

---------- response 1
  03/08/76     16.13     copland       csa

        Won't using the system defined words 'tactive', and
'ptime' give you a good enough idea of how much idle time
there should be. Also you can always run a simple test,
like a simple count loop in foregnd then the same loop in
backgnd, and compare the elapsed times for each, and then
base your choice on the results of that.

---------- response 2
  03/08/76     16.29     warner       iu

        That's not what he's asking for....
he wants a flag set so that people can know whether their
lesson is in backgnd or foregnd.
        Frankly, I think the idea is a little ridiculous...
if you need to know. it's easy enough to set a flag in the
student bank whenever you change over!

---------- response 3
  03/08/76     16.50     dave       pcp

        That also is not what he is asking for, He is
basically asking for something to tell him how much backgnd





time is available(how much is being eaten by background
jobs, background users, etc.) so he knows weather it is
worthwhile to put some hairy calcuations into backgnd
or not. Ptime and tactive can help, but they really
have almost nothing to do with the amount of background
time available. as one job can make a considerable
difference.

        I could definitely use something like this....it
would be nice to known how much backgnd time is around...

dave

---------- response 4
  03/08/76     17.16     bowery       comm

Dave's comment on the lack of correlation between -tactive-
and the amount of idle time is good. -ptime- is a farce
for this purpose. One CAN do a loop and use -clock- to
find out which mode is better but that in its self takes
significant cpu (at least 4 full timeslices) to determine.
Accessing one variable is preferable to using 4 timeslices
to aquire the same (or worse) information.

Like I said:

As long as the system supplies a -backgnd- option it seems
that it would be VERY logical to include a flag like this.

---------- response 5
  03/08/76     19.18     fay       0

Yes, yes, a thousand times YES! Those (few?) of us who
never have occasion to run a really monstrous number crunch-
er perhaps cannot appreciate the benefit to be gained from
such a reserved word. Anyone who has a heavy calculational
routine can appreciate the necessity of knowing (w/o signif-
icant additional processing) the propriety of switching to
mode backround.

---------- response 6
  03/09/76     01.22     avner       s

As was pointed out, a single job can make a big difference
in the amount of background available. We small-hour
number-crunchers are all familiar with the amazing effect
of a group of game-players moving on to a new (non-
background) lesson. A single snapshot is generally
inadequate as a basis for a decision which will commit you
for many minutes. I have much mor luck with alternate
sample loops which do computations which are needed anyway.
Based on comprisons of resulting times, dynamic shifting
can be done to let your crunching bend with the winds.

---------- response 7
  03/09/76     01.45     bowery       comm





Al,

  What does one do with things like compilers where a
logical unit might take orders of magnitude more -proctim-
than another logical unit depending on hundreds of
variables? Evaluate the prediction function and normalize
the Δ-clock- everytime?

  The pseudo terminals already do redundant tasks.....
why not let them decide, say, once every 30 seconds?
Does idle time generally change that fast?

---------- response 8
  03/09/76     09.18     avner       s

Running jobs that do compiling is going to be a lot
stickier than Monte-Carlo or data retrievel-massage
jobs. The major factor in predicting the volatitlity of
background access is the number of people in background
and the "ducty-cycle" of their demands on cpu time. A
few game players will result in massive shifts in demand
in a matter of seconds. A class full of students using
lessons which require background will wipe out all chance
of background access (for themselves as well). The result
is that short of a rather involved combination of ESP and
considered, experienced judgement it is almost impossible
to come up with a single measure that will be valid for what
is desired under current methods of allocating resources.

---------- response 9
  03/09/76     12.48     jmk       pso

Shifting into background status to test whether it is
advantageour to run in the backgnd is a disaster if there
is no background time available. I have had to wait
several minutes for just one (1) more timeslice in
the background to get a chance to get back into the
foreground. Two (2) TIPS is better than no TIPS at all!!

---------- note 40  platostory
  03/08/76     18.16     roper       siu

This is probably not going to work, but...

Is there any possibility of obtaining enough lesson space
in some account, somewhere, or even creating an account
to hold the stories that are written and rapidly disappear
from PLATO?

Admittedly, I'm personally involved with the Guand Gap,
but there are many other good stories such as the Death
of PLATO (which I have yet to see since I was off the
system this summer), The Adventures of the Red Sweater,
Adventures in Platcland, etc. which either have, or will
shortly vanish from the system.

Any chance? Even some space on disk pack helen so that





they could come back easily when the disk space crunch
is less?

Quetzal

---------- response 1
  03/08/76     23.41     jones       mcl

This may be a signifigant idea.

Plato has become an artistic medium, and it may well
be desirable to save the first (faultering) examples
of both prose, graphics and others for posterity.
But of course, leave out the awful poetry from notes
and NEWSREPORT.

(imagine visiting the Smithsonian 50 years from now
and reading the Great Guano Gap).

1.999999c≤/

---------- response 2
  03/09/76     12.03     lowbardo       ed

Here's the first contribution written specifically for
the proposed new historical, creative plato lesson.
      Roses arew red.
      Violets are blue;
      When Plato crashes,
      I am blue.
         dal

---------- response 3
  03/10/76     18.04     a wong       ed317

 I'm also in favor (for whatever that's worth) Some of
the things that have appeared in news,pad,etc are,in
my opinion, worthy of being saved for Plato freaks unborn!
How about it,powers-that-be?

mad prof

---------- response 4
  03/11/76     14.01     roper       siu

I'm afraid that the powers that be are ominously silent...

---------- note 46 responses
  03/08/76     19.44     patrick       mathw

        Why isnt there a way so that in group notes.
when someone writes a note. he can restrict the number
of responses. This would save alot of space where people
are writing resonses that have no useful purpose.

Patrick





---------- response 1
  03/09/76     10.42     berger       mfl

I don't understand - if some of the responses have no
useful purpose. then restricting the number of them
will prevent those that want to write useful responses
from writing them.

---------- note 47 new rotate
  03/08/76     19.55     sellers       arizona

Because some importan discussions about Plato IV graphics
are imminent, I have opened lesson 'rotate' for running.
The lesson is not finished or proofed, but I decided
that now was the time to open it.

I have some general ideas that I believe are important to
the future Plato graphics. The lesson conveys some of these
ideas. Please take a look at it now with the thought
of looking at it again when it is finished.

The first complete craft of the lesson will be done by
tomorrow morning and the final form of the lesson will be
ready within a week.

Thank you.
John Sellers

---------- response 1
  03/08/76     23.54     jones       mcl

3d graphics on PLATO would be a mixed blessing.

Admittedly the displays are pretty, but many of the
examples shown are too time consuming and may be good
art, but probably are not of much real use for all but
a small fraction of the PLATO applications.

Is it outside the range of propriety to ask what basis
there is for puting 3d graphics or music box control
in any language on the same footing as the basic semantic
items like branching, judging, units and basic output?

It could be argued that any command not really needed
by less than 10: of the user community be relegated to
the library of stuff in AIDS, and the work that currently
goes into adding such commands to TUTOR be devoted instead
to devising an efficient way to execute such things so
that execution is almost as good as if the command were
added to TUTOR. TUTOR is already too big.

---------- response 2
  03/09/76     01.30     avner       s

I am afraid that the 10: criterion would wipe out most of
the language. Fewer than 30 commands get as much as 1; of
total usage. But note that this is the same thing that
occurs with natural languages. The idea behind a genuine





! language! is that people add vocabulary as they need it
rather than feel obligated to learn every word in the
dictionary. As to the second point, making operation of
current commands as efficient and effective as possible
has always had higher priority than adding new commands-
given that such added efficiency is of more benefit to
the bulk of users than the new command.

---------- response 3
  03/09/76     02.10     sellers       arizona

1) 3d graphics are inevitable someday. Now is the time
to talk about it so we won't run into trouble later.

2) My displays are not meant to be fast. They are to show
the ease with which rdrawn figures can be manipulated by the
addition of one or two commands. It would be up to the
lesson writers to find how to best use the graphics.

3) As for use by minorities, vocabs was created at the
request of a single lesson writer, but we would not have
the Aids facility we have without it.

4) Graphics is basicly semantic. Choosing the commands one
can manipulate rdrawn figures with is a semantic chocie. The
true purpose of graphics is communicational in nature.
There is basicly more intuative information in 3d figures
than there are in plane figures.

---------- response 4
  03/09/76     13.19     carter       comm

TUTOR is not too big. The prime advantage of a highly
centralized system such as PLATO is the advantage of
large, comprehensive languages. TUTOR is a good start,
but is too small, not too big. The use of local
intelligence will make some of sellers suggestions
operate better, however, experience along these lines
now is valuable.

---------- response 5
  03/09/76     21.38     shirer       s

Semi-official response from part-time graphologist:

a)  We have no immediate plans to add 3-D graphics commands
    to TUTOR.

b)  By all means keep letting us know of your ideas along
    this line. They will be noted and remembered.

c) Several years agc, a user survey showed little interest
   in 3-D graphics commands, so they were not implemented.
   However, this may have changed, so I have no objection
   to people telling us what they want or don't want.
   I do not have much interest in people telling me what
   _other_ people want...I do not believe in the accuracy
   of 3d person reportage!





d) Seller's lesson is cute...might be the start of an
    interesting lesson about perspective?

---------- response 6
  03/10/76     15.46     krol       mxc

   how can you have 3-d graphics on a flat screen?
if you want a real challenge try making a plasma cube
interfaced with the i/o ports in back of the terminal
if you have a million or so to blow!

---------- note 49 J(oin)
  03/08/76     20.45     bowery       comm

Comments on yet another idea(?):

In INSPECT only mode in a lesson, allow one to type:
J<unit name> next. This would stack the person's current
block/line numbers and do a U search for unit <unit name>,
unless it had been located before in which case its location
could be store) in a table. Then when one presses "R"
the stack is popped and one is Returned to the block/line
one J(oined?) that level from.

Is there enough room left in INSPECT only in the vars or
storage for this stuff? (even 2 levels would be GREAT!)

It would make tracing down code INCREDIBLY more efficient!

Think about it next time you look through a bunch of -do-s
and/or -join-s.

---------- response 1
  03/09/76     14.17     travers       mxc

I'd love it, but I don't see why it has to be only
inspect mode. Just use 'P' for pop. (It's about time
to take off P as an alternate to C, anyway).

One vote for!

---------- response 2
  03/09/76     16.52     bowery       comm

The main reason this would be good on inspect only is
the availability of variables that are used only in
modification of code. The only problem is that perhaps
they don't know which vars those are..... in which case
they would need a J option to find out. So the lack
of a J option means that we won't be able to have one..
hahahahahaha.....

---------- note 54 compute
  03/08/76     21.49     strass       csa

Question....
      Can the length of a string used in a -compute-





be greater than 10?

                Thanks...

---------- response 1
  03/08/76     21.53     dave fuller       uimc

         "The maximum length of the string is 99 characters"

           (from aids, page 2 of the -compute- desciption)


---------- note 57 -ext-gripe
  03/08/76     22.53     kent       unl

         I have been trying to find out how the -ext-
commands works and everywhere I look, the description is
inadequate: ".,.trucates the argument to 15 bits". Is
this a truncation on the right or left; i.e., are the
high or low order bits sent?
        Nowhere does it say: not in AIDS, not in Avner's
Plato User's Memc, not in Sherwood's The TUTOR Language.
Would somebody please 1) enlighten me and 2) correct these
"deficiencies"?

---------- response 1
  03/08/76     23.09     midden       s

it sends you the lower 15 bits.
If you give it a floting number, first it rounds that to
an integer, and then sends you the lower 15 bits of that
integer.
      M.Midden
p.s. you can get an extra 16th bit, by using either an
-audio- or -ext-.

---------- response 2
  03/09/76     10.25     coates       itl

On terminals with serial nr.s above 261,both the audio and
the external are sent out the same port in serial. 20 bit
words are sent out . bit 19 1st bit 1 last followed by
parity. The last 15 bits are those sent by the audio or
ext command. Bit  16=1 for ext,=0 for audio. The format
is the same as the word format described in"The Plato
4 Terminal: Description of Operation" by Jack Stifle,
the revised edition. The timing format is described on
page 66 of the same manual.

---------- response 3
  03/09/76     10.44     ld francis       mtc

lesson "external" typically (i havent checked this case)
has information of that sort--it should be referenced by
"aids"

---------- response 4
  03/09/76     12.59     jmk       pso





Lesson !external" does appear to contain a good deal of
valuable and hard to find information. Its author is no
longer with PLATO. Informed authority has it that the
lesson is not entirely accurate/up to date or both.
When and if we can get it cleaned up and decide how to
maintain it, it may well become a part of AIDS.

---------- response 5
  03/09/76     23.48     kent       unl

         I just looked and this terminal has serial no.
1340 and _2_ ext jacks: labeled "ext out 1" and "ext out 2".
They are square and take a nine pin plug. They are next
to a much larger, 24-pin, rectangular plug which is unlabel-
ed. Will the CERL music box (gsw box) plug directly into
this terminal with no problems?

---------- note 59 nrecords
  03/08/76     23.39     b sherwood       s

Go into "nrecords" and try out the curriculum options
(on some instructor file you don't care about). Try
in in inspect mode, too! And with the various new
author/instructor options as well....

---------- response 1
  03/09/76     08.26     d zweig       iu

Entered a course via "nrecords" using the ¬INSPECT CODE, then
tried to enter CHANGE CODE, and couldn't get anywhere. That
is, couldn't get into Change Mode.

---------- response 2
  03/09/76     08.27     nagel       vu

There are some overwrites......both at the top
of the page and at the bottom.....

These happen when you look at the module listing
page. The titles at the top overwrite the course
and instructor file info. At the bottom left
the (HELP) has also been slightly overwritten.

---------- response 3
  03/09/76    08.29     nagel       vu

How do you switch into inspect only mode?

---------- response 4
  03/09/76     09.05     b sherwood       s

Type the course name, then press LAB instead of NEXT,
and you will be in inspect mode.





Switching the other direction is being redone. For now,
just go out and come in again.

---------- response 5
  03/09/76     16.03     koning       csstaff

Pressing DATA on the first page of the curriculum editor
in inspect-only mode sends you to the wrong place...

---------- response 6
  03/09/76     20.15     michael       english

Also, pressing NEXT from the instructor file info page
returns you to the wrong place.

---------- response 7
  03/10/76     09.45     roper       siu

It's nice to have random X's when entering a student's
password, but could we please have two shots at typing
it in case we goof?

          Quetzal

---------- note 62 try "sm"
  03/08/76     23.42     midden       s

For persons who like to use unadvertised features, I thought
I'd give those persons a chance. I have "almost finished"
the rewrite of the "sd" option in the editor. I seriously
hope that there are no serious bugs that will cause people
to lose work. Since I'm not offically releasing it, I hope
everyone takes care when using it. To access the option,
type "sm" (see marshall) instead of "sd" in the editior.
I look forward to few bugs, but know better than that.
I would appreciate bugs reported to me via personal notes.
                Sincerely yours.
                Marshall Midden
p.s. the text justification routine should be nicer now.

---------- response 1
  03/09/76     10.57     hody       med

this is going to be an extremely useful and versatile
set of routines...
*
because it is so complex, however, and will take so
much time to learn completely, what about retaining
the current nice and simple ID/SD as an option from
the index page?
*
even, maybe simplify it further to remove relative
draw... so it would then be for beginners who want
just lines, circles, text, and arrows...?? ..maybe
make it have the same letter commands as the new version
to avoid confusion?

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





  03/09/76     13.54     nate       iumusic

   SM can be just as simple as ID/SD. SM works much
faster, RECOGNIZES embedded ats, among other things.
Sorry, ID/SD can go anytime, thank you.

                        no offense
                        Nate

---------- response 3
  03/09/76     15.48     hody       med

no offense!-- how about some abridged instructions, then?
say two stages of HELP... like HELP first time
tells you how to put in text and simple draw/circle...
then HELP again for more information.
*
i am not thinking of all you old hands... just the
new ones.

---------- response 4
  03/09/76     17.46     snellen       medneta

AMEN to that!

---------- note 73 inst file
  03/09/76     08.34     nagel       vu

Are we ever going to be able to clean out the
dead wood from the catalogs in the new instructor
files????????? This was mentioned a long time
ago...some thought that it would be a good idea.
As it is now the only way is to destory everything
and start typing in all over again.

---------- response 1
  03/09/76     09.55     rick       mfl

I think that a proposed solution was to allow you to do
it only when the course is empty. (But nothing official
has been said since then.)

---------- note 78 charcount
  03/09/76     10.12     roper       siu

Can anyone tell us how to do an actual character count
on a packed string? What we essentially want to do is
to count the number of characters, not including shift
codes, to allow precise location of the display of the
packed string, so that we can do accurate animations,
underlining, etc.

Any ideas?
           Quetzal

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





  03/09/76     12.14     frye       mfl

I offer you the greatest of luck in finding anything rea-
scnable to perform this function -- one time, long ago, I
asked for this same function for -arrow- typeins in the
form of both a "dlong" ("display long", similar to -long-)
command and a *charcnt* (display character count) variable.
The system response at that time was "someday..." and this
request was made in early 1974!

If you're interested, I have 10-12 line calc that does
approximately the same thing -- or. if you have a blank
area on the screen, you can -showa- the string in mode
erase, then check "whrex, wherey, for the results -- re-
member, some codes move both x AND y...up AND down.

G. David Frye

---------- response 2
  03/10/76     02.03     d williams       unidel

That -showa- suggestion would work, but only with a
-write ¬oa- ... since -write- correctly increments "where"
and -showa- doesn't.

Rut I do it with the check-each-char method:

Create a word "map" and a (two-variable) word "valu" and
-segment- them into single bits (mapbit ¬+ valubit?). Put
a huge, permanet number into "valu" such that valubit (x)
is 1 if character x prints. o if not. (Thus valubit (56)=0
since 56=070 (s Shift.) Then one quick loop:

doto    60max,N‎1, ncchars $$nochars = ¬$ chars in input
calc    mapbit(N)‎valubit(Nth char of input)
60max

Now, ¬$ of chars printed out of the first 60 input chars
will be simply bitcnt(map). However, this won't keep track
of backspace, super/sub, etc., which would require further
complications.

---------- note 83 R-help1-2
  03/09/76     11.04     fries       ee

R-help1  03/05/76     190.8     fries       ee

        Back in note 218 I suggested adding the option
of deleating a line in replace mode by pressing HELP1.
This would remain consistant with the format for deletion
in the other parts of the editor. This option, along
with DATA to go between insert and replace mode, would
allow easy modification of a small piece of code.
        When I originaly suggested this, no one
objected to the format, and it would maintain consistancy
within the editor. However the was no systems comment
before. Do I hear any yea's ?





EˇE

How about a comment systems....

---------- response 1
  03/09/76     11.16     hody       med

and also on the back burner.... how about LAB to clear your
line in the insert, replace and "text" part of SD?
*
for the lentghy, lousy, line-error-ridden, long-distance
legions?

---------- response 2
  03/09/76     11.18     mike b       cornell

        Both of these ideas are excellent, especially
(I don't believe I'm saying this) Hody's. I'd like to
see these somewhere closer to the front of the stove.

M¬ ike≤≤≤≤ˇB

---------- response 3
  03/09/76     11.45     hody       med

well, mike, i am delighted you CAN on (rare) occasion
escape from personlities to issues...

*
whiel we're talking about urgent improvements to the
already excellent editor, how about that separate
warning page for entering the lesson in inspect mode?
(with option to enter change code at that point?)
["Currently you can only INSPECT this lesson, press
NEXT to inspect. or DATA to enter a CHANGE security code"]
*
if plotting time or an extra key press is a problem,
just having it for BLANK inspect codes would be
helpful and it would never be needed for entry
by LAB instead of NEXT.... yes?

---------- response 4
  03/09/76     12.09     lombardo       ed

I have, over time, discovered that hody can honestly
make a compliment sound like an insult. Did you have
an unhappy childhood, are you afraid other people will
misunderstand you intentions? SMILE (occasionally)!
dal

---------- response 5
  03/09/76     13.41     dave       infe

       Another nice option in Replace and Insert would be
a mode of exit (analogous to OUT) that prevented the
modifications from being written onto the storage-CM copy
of your block; maybe BACK1 to exit.





---------- response 6
  03/09/76     23.03     hagerman       ames

     Where is the system response, huh?
Get cracking sysprogs!

---------- response 7
  03/10/76     11.13     artman       uimatha

re: exit from Replace

How about just erasing everything you have typed, then
pressing BACK. That seems to restore the old version for
me.
hopefully helpfully,

eric artman

---------- response 8
  03/10/76     13.10     k mast       p

Someday, someone is going to get abmitious and rewrite the
editor. At which time I'm sure he will consider all those
options. Until then, the editor is just TOO big to keep
adding on things it wasn't designed for orginally.
(The original editor was designed and written in TUTOR on
Plato IV almost 4 years ago!)

---------- note 91 notes bug
  03/09/76     12.24     tricorn       hebrew

I just noticed this...if you are at the beginning of
a note index. press BACK (which does nothing) then
NEXT to go on, it simply replots without moving along
in the index.,.??

Δ

---------- response 1
  03/09/76     12.26     woolley       p

Fixed on next version.

---------- response 2
  03/09/76     12.27     tricorn       hebrew

Gee, that was fast...in fact, I just found that I
couldn't repeat in one minute after I wrote the note.

---------- response 3
  03/09/76     12.28     woolley       p

Actually, I fixed it before you wrote the note.





---------- response 4
  03/09/76     12.45     weaver       matha

On the same (notes bug) subject:
There is a bug in the notes editor.If you press I to insert,
then erase the I and then press - to move the screen, the -
is plotted and the display does not move. That is at
least inconsistent with the regular editor and probably
not a planned feature.

---------- response 5
  03/09/76     13.08     rick       mfl

The notes editor uses a standard arrow; the contortions
necessary to make the arrow accept on keypress, try to
judge it, and allow more keys if no match, are truly
horrendous. It is next to impossible (if not totally so)
to make the notes editor work like the TUTOR editor, which
uses a -pause- loop instead. This same phenomenon is
also responsible for the sometimes seemingly-erratic
behavior of the "author mode" arrow.

---------- response 6
  03/09/76     17.18     k mast       p

Right. We hope to eventually 'fix' the arrow processing
machinery to start 'clean' if all the charcters have been
erased.

---------- response 7
  03/10/76     10.48     friedman       csa

Meantime, just press NEXT after the "-" if necessary to get
it acted on.

---------- note 95 common com
  03/09/76     12.34     lombardo       ed

WOULD YOU BELIEVE DEPARTMENT:
     a way to inspect a commd located in another lesson
but used as the common for the lesson you are in, without
leaving the current lesson. Effectively it would be like
having the common located in very lesson it is used for.
This would prevent constant acking out of and into lessons
just to inspect the common.
     Is this a common problem? (excuse me!)
dal

---------- response 1
  03/09/76     12.48     clark       lawyer

this same type of idea has been mentioned for being
able to go from one lesson to another that is being
-use-d...one vote for!

---------- note 105 use-edit2





  03/09/76     14.06     anderer       ee

********
use-edit  03/03/76     01.01     anderer       ee

   Another from the 'wouldn't it be nice department':

         When you are editing a lesson that -use-es
another, hitting, for example, 'U' from the block
directory page would take you to the -use-d lesson if the
change codes matched. Otherwise it would take you to
the 'Press DATA to try,...' page.
         This would be an imporvement when you're editing...
Rather than hitting BACK, typing the other lesson name,
hitting next....Just hit 'U', and there you are.

         (And just think if it remembered what lesson
you came from! Just shuttle back and forth...)

********
Any comments? How about a systems one?

---------- response 1
  03/09/76     15.01     midden       s

Seems to me that you expect a lot in only 6 days.

---------- response 2
  03/09/76     16.28     perry       uw

  I _think_ all that was desired was a "yea" or "nay" or
"don't hold your breath."

(He still doesn't have one of the above.)

---------- response 3
  03/09/76     18.23     anderer       ee

That was what was desired.....and still is.

---------- response 4
  03/09/76     18.54     midden       s

Will you settle for:

      We see the request. Is this actually was is
      requested, or is there a more basic need that
      should be solved instead? Jim Bowerey's request
      is much more general, and would still allow this
      here simple addition to be done. As with most
      requests, don't hold your breath, and when these
      notes get lost in the archives and someone else
      requests the same thing two years from now, you'll
      be able to say, "I remember something like that."

Marshall Midden





p.s. there is also another way to look at no system response
to a note: it wasn't worth answering.

---------- response 5
  03/09/76     19.14     anderer       ee

    Thanks for the response.

(What might make a non-frivilous request 'not worth
answering?')

---------- response 6
  03/09/76     20.42     midden       s

1) One has nothing to say.
2) it is out of comprehension.
3) you have never had need for such a thing, and don't
    know if it actually is useful, or someone doing something
    that probably should be done another way.
4) you don't want to get involved.
5) you don't want to offend the person writing the note.
   but have no idea how to say that is a dumb idea.
6) a bad day has put you in a bad mood and you don't want
   to say anything that could be misinterpreted.
7) the note is so long it looks like a waste of time to
    read it even.

etc...

unit      choose
ranud     n1,maxnum
jumpout   n1;q;x;notes;hop;dogfight;airfight;aids;debug;crash;
          problem;edit;window

---------- response 7
  03/09/76     21.59     anderer       ee

    Sorry, no offense intended. Just curious.

---------- response 8
  03/10/76     10.22     judy       pso

I think the Marshall's first choice is the more common
one. Sometimes the "s" or "p" people simply have nothing
to say about an idea.

Rest assured. that your ideas ARE read by systems!

---------- response 9
  03/10/76     10.51     friedman       csa

Yes, but it's greatly appreciated if they'd at least say





something like "read and will be considered". Otherwise,
one really feels ignored...and I don't think I'm prone
to being paranoid!

---------- response 10
  03/10/76     14.54     bowery       comm

I tend to agree with Friedman.

---------- response 11
  03/10/76     19.27     blomme       s

  Yeah...but what if the real feeling is "read and
WON'T be considered"? That is probably more likely
in many cases--and it actually is not because ideas
are not useful or apply only t) a very limited number
of users (as is often the case), but just that there
are so many projects to be don; that priorites (and
personal interests) dictate a ,read and ignore" response
to many things.

---------- response 12
  03/11/76     08.39     berger       mfl

I remember something like this...

Two or three years ago, when there was a lot of resentment
against the systems staff by the user community, the systems
staff stated a policy that saic that all user requests would
be acknowledged by a systems response in notes.

Even if you are going to ignore the request for whatever
reason you have, PLEASE let us know that you read it.

Especially since there are so many new notes in here every
day, it is very easy to miss some, and the user has no way
of knowing that anyone on the system staff saw his request.

---------- note 106 base allot
  03/09/76     14.09     j thiher       amesrad

Has there been some changes in the base allocation
at any sites? I don't know exactly what our allocation
use to be but I'm quite sure it is different that what
it was last week. I am not upset or anything but I
just thought that base allocations were rather fixed.
jim

---------- response 1
  03/11/76     10.15     golden       s

Base allotments change as terminals are added to or deleted
from logical sites.

---------- note 108 canceled





  03/09/76     14.47     golden       s

The previously announced seminar by Fernand Marty has
been postponed and will be rescheduled.

There will be no PLATO Seminar on Wednesday March 10.

---------- note 111 -termecs2-
  03/09/76     15.58     putch       arizona

      hate to bring this up main, but the response to my
previous note regarding -term ecs- seemed favorable, and
there was no answer to clark/lawyer's call for a Systems
Response. Since he also indicated that this issue _has_ come
up several times(also with favorable response) before I came
≤the system, would it be possible to get a reasonably
definite comment from them on this question?

---------- response 1
  03/09/76     16.06     j thiher       amesrad

For the sake of simplicity lets avoid any more system
terms. I know that term 'ecs' would be nice when you
are interested but I question whether it is really
necessary. After all, I as an author am not really
concerned what the current ECS usage is as much as I am
concerned about the general state of the memory i.e. full,
3/4 full etc. These things can be gleaned easily by noting
the time of day. If I am really interested in knowing
whether or not I have enough ecs to condense a particular
segment of code, is it really that hard to press 'E'
at the author mode page?
jim

---------- response 2
  03/09/76     22.08     warner       iu

        If you really need TERM-ecs in a lesson, why not
add it yourself with -termop-?

---------- response 3
  03/10/76     00.29     blomme       s

     I think it is unlikely that this will be done.

---------- note 114 foregetname
  03/09/76     16.16     meers       wright

Just noticed that course names are no longer forgotten
AS LONG as you have the correct security code.
thanks

---------- note 115 ,, or , ?
  03/09/76     16.18     perry       uw





  The following line of code always sets n2 to 0, a result
expected only when n1<4 is true.

calcs n1<4,n2‎0,

To do what I expected required an extra comma.

calcs n1<4,n2‎0,,

  Is there a logical reason why the first line should not
be condensed the way I expected?

---------- response 1
  03/09/76     16.42     avner       s

In order to avoid problems of having too many separaters,
the last comma on a line is treated as if it were the end
of that line (to allow people to use either comma or not
at the end of continued lines, as they wish, without
causing grief). Al Avner

---------- response 2
  03/10/76     00.31     blomme       s

   This means you have only one option to choose "among"
and hence regardless of the decision, the same thing is
done.

---------- response 3
  03/10/76     10.27     judy       pso

Therefore, to turn something "off" in the final
position of a conditional string, you always need
a specific "off" indicator:

writec n1¬,negvalue¬,zerovalue¬,¬, $$ extra ¬,
goto   n1,here, there,x        $$ "x"
calcc  n1, b‎14,r‎b,.          $$ extra comma

---------- note 117 charsets
  03/09/76     16.24     slayton       arizona

In trying to copy charset Keys of lesson introduce, three
times in a row I tried to copy the entire charset in the
charset editor, and three times in a row it failed to copy
char 71....!?!?!?!?!?!?!

DS

---------- response 1
  03/10/76     10.02     mcneil a       uicc

Could it be that your r/w memory (charset) card in the
terminal is flakey? You could tell by looking at option 1
in the charset editor. If there is a char71 listed but not
showing then it's probably the terminal. Otherwise I'd





advise creating a new charset block with a name ¬='Keys'
and then copying charset Keys into it.

---------- response 2
  03/10/76     14.58     judy       pso

I have tried copying several character sets and playing
around with them.

The slot which is skipped always seems to be the slot
which  SHOULD be the 64th slot loaded.

---------- response 3
  03/10/76     19.31     struven t       csstaff

I wrote a note in the 'old' help notes stating the rpoblem
with charsets not copying certain characters. I moved from
terminal to terminal and not on; was the entire set copied
completly. It always skipped or _one_ letter between 64
and 85.
        I don't think that this is a problem due to term-
inal malfunction...or, if it is, it is common to all the
terminals that I tried (8 to 10).
Terry

---------- response 4
  03/10/76     21.32     blomme       s

   Thanks for the information/reminder--sorry this wasn't
fixed a long time ago. A special (incorrect) check was
being made for the 64th character and it was being skipped.
This check has been eliminated and everything now seems
to work correctly. The check has probably an artifact
of a different method used long go for the copying and
was intended to check for absolute character number 64
which is normally set to a space. (As it was, it actually
was checking for the 64th character in the set, not the
character associated with the character slot 64, so the
character lost depended upon the charset.)

---------- note 120 back notes
  03/09/76     16.37     parker       arizona

I was just looking for an item in Public notes that was
written just mid-last week. However, someone reinitialized
the notes file before I could but there. I was thinking
that perhaps it would be possible to hold released notes
for about 2 weeks back. This would allow references to
some past notes. The way it exists now, after re-
initialzation, old notes are re-eased never to appear again.
What does everybody think of that?

Mike Parker

---------- response 1
  03/09/76     16.42     woolley       p





I don't understand. Old public notes are still available
in an archive file. Type the approximate date you want
on the public notes index.

---------- response 2
  03/09/76     21.15     parker       arizona

Maybe that the notes machinery should advertize the fact
that previous notes are available. This would certainly
be a benefit to all users who have their notes archived.

Mike Parker

---------- response 3
  03/10/76     00.38     blomme       s

   It is only "advertised" in the help section; as a
matter of general policy, we have been trying to make more
information available in the various editors through the
HELP key and take up less time and display space generating
incomplete comments and information on key operations on
the main displays.
    (I acknowledge that several annoying "partial
information" displays still exist and that the information
provided via HELP is still far from complete in many
places--nonetheless, the way to look for information is
to try pressing the HELP key...and reports on our failings
there are welcom.)

---------- note 123 gnote bugs
  03/09/76     16.46     kimble       ustaf

A bug and some suggestions for gnotes:

But: When a gnotes file is completely reinitialized, with
     no archival copy and no notes kept in current file.
     the note index page lists the last note that was
     formerly in the file, with its old number, but will
     not (reasonably enought) let one access that note ¬$.
     This bogus listing disabears as soon as a new note
     is written into the file (as note ¬$0).

Suggestions (wild complaints) also have to do with the
reinitialization procedures, and are all related to the way
"deleted" notes are handled.

I realize that when a note is deleted, it is just various
pointers that are changed...no additional space becomes
available since the note is still there...it just can't be
seen (with some possible exceptions). This is quite a
reasonable way to handle things....BUT
(continued in response)

---------- response 1
  03/09/76     16.55     woolley       p

The bug has been fixed and will work again when notes is
recondensed, Until then, everything will be fine if you





just write a note in the file. (Empty files confuse it.)

---------- response 2
  03/09/76     16.55     kimble       ustaf

While this is acceptable for the current notes file, does
it have to carry over into reinitialization operations?

1. Suppose I want to reinitialize a gnotes file. Fine.
   Let's say that there are notes up to ¬$163, and that
   there are several outdated notes, and several that I
   would like to retain. Let's also assume that some
   (perhaps 10 or 15) of the outdated notes are inter-
   sperced with the notes I'd like to keep around.

   It would be nice if I could just delete notes 132-148,
   say, and then specify that notes from 101 on (hopefully
   excluding notes 132-148) should be retained in the
   reinitialized file.

   Unfortunately, the dead wood (deleted notes) are also
   carried into the new file, taking up a significant bit
   of space, though they have been "deleted" and are not
   visible.
(continued in next response)

 ---------- response 3
  03/09/76     17.01     kimble       ustaf

Got your response in there before me...

Anyway, wouldn't it be possible for the reinitialization
routine to check the flags to see if a note has been
deleted before retaining it in the "new" file, and if so,
not keep it around taking up space?

2. The related question has to do with archival files...
I don't know how you handle that. but it would be nice if
it was not required that the chival file have the same
length as the current file. since there way well not be
as many notes in it, both due to deleted notes (as above)
and also since some notes may be retained in current file.

---------- response 4
  03/09/76     17.05     woolley       p

It would take too much time (like maybe half an hour for
a big file during prime time) to clean up deleted notes
during re-initialization. Instead, I'm planning on making
that a separate option. It would be available soon.

---------- response 5
  03/09/76     17.17     kimble       ustaf

thanks!

---------- note 125 ckpt rtvar
  03/09/76     16.54     sherman       mcl





Are router student variables automatically checkpointed
along with regular student variables, or is some special
command or tag needed to do this?
         L. Sherman

---------- response 1
  03/09/76     19.06     chabay       s

Router variables are automatically checkpointed.

---------- note 130 show ?
  03/09/76     17.16     bonetti       ced

S there a way to specify the number of digits to show
after the decimal point? The only command that I can find
to do that is -showt- which doesn't suit my purposes since
it writes the numbers right justified.

For example:

I would like to have something like this:
......
write Your score is -0s, sc(re,6.2-1
....
so that if score = 12.345 the the student will see
"Your score is 12.35".
However if I use showt, four extra spaces get stuck in
between "is! and "12.35, 0 I have
"Your score is 12.35".
This looks pretty sloppy. A1 I overlooking something or
is TUTOR dedicient in this area?

---------- response 1
  03/09/76     17.21     alan       reading


-showz- might be what you want

 ---------- response 2
    03/09/76     17.24     bonetti       ced

Nope, I tried -showz- too and that didn't work either.
Any mor suggestions?

---------- response 3
  03/09/76     18.54     chabay       s

-showe- will let you specify exactly the number of
significant figures to show. Probably this isn't what
you want here, though. If you can't settle on a reasonable
compromise for the -showt- (like 3.2 in your case?), you
can calculate the number of places in front of the
decimal point to show (take the log of the number 10
be show, or whatever..) and do a-ot,score, (n1+0.2)¬1.

---------- response 4





  03/09/76     21.41     clark       lawyer

Try ¬0s,(int(score≠100))≡100,6¬1
¬W = number of digits to display
That just worked for me...

---------- response 5
  03/10/76     09.09     marty smith       mtc

well i*ll be darned, all this time and i never knew
about that! just goes to show, you learn something
new every day.

---------- response 6
  03/10/76     09.37     bonetti       ced

I like Clark's idea and I think I'll use that.

Thanks.

---------- note 135 new notes
  03/09/76     17.27     berger       mfl

It gets harder an harder to read the day's notes...

Can some thought be given to dividing public notes
into different categories?

---------- response 1
  03/09/76     19.13     al       mfl

But "general" already exit.

---------- response 2
  03/09/76     22.16     j thiher       amesrad

I heartily agree with berger. I don't know why but it
seems that it take longer every day to wade through
all the public notes. I'm not sure that the previous
system of HELP and GENERAL was better but at least then
you had some idea of what notes you wanted to read.
                 jim

---------- response 3
  03/10/76     01.38     mike       cornell

Me too. I never used to want to read help notes, but now I
am effectively forced to decide if I will bother reading
each note. Perhaps a poll to determine a system of
catagories would be in order? I would suggest:
Requests
Bugs?
Help!
General





I am sure many other people will think of better
classifications, buts lets do it soon.

          Mike Huybensz

---------- response 4
  03/10/76     08.32     mont       csa

I agree. Perhaps after writing a note, the writer would
get a page asking 'What type of note? Type _a_ for
programming problem; _b_ for request for new feature;
_c_ for general note; etc.' Then a reader would just look at
the type(s) he wants to look at. Provision could be made
for putting a note in more than 1 list ( hopefully without
making extra copies ), and for adding new categories as
time goes on, Comments?

Richard Montanelli

---------- response 5
  03/10/76     09.39     sylvia       pdg

         Depending on how you count them about half of
         the notes stored since these notes were reintialized
         are in the "wouldn't it be nice" category. I
         seldom find these notes of general interest and
         I'd sure like to see them put in a separate
         category.
                sylvia carlson

---------- response 6
  03/10/76     10.20     lombardo       ed

I agree, I think public notes is far to encompassing, more
specific catagories are needed:
        help: programming + system problems/questions
        announcements: new lessons, lectures, etc.
        wouldn't it be nices: system oriented ideas
        miscellaneous: general pap that's leftover, but of
             interest to some people

dal

---------- response 7
  03/10/76     13.31     frye       mfl

Believe me, people, this has all been gone over before. In
fact, a "working version" existed at one time but never made
it to "public" status -- I can't remember why... Is it
REALLY that hard? It doesn't take me that long.

G. David

---------- response 8
  03/10/76     13.40     woolley       p

There's a difference between just programming something...
and programming something useful, useable, and efficient





that will please everybody.

---------- response 9
  03/10/76     14.12     clark       lawyer

I disagree! Think of all the notes that were written
complaining that !this note belongs somwhere else..."
Just glancing over the first lines of a note tells you
whether you want to read the note. It doesn't save that
much time or energy to make jp more classifications.

---------- response 10
  03/10/76     14.51     frye	      mfl

Just in case I was misunderstood by dave, I meant that I
didn't find it hard to read notes once or twice a day. I
wouldn't want to be stuck with the job of programming it!

G. David

---------- response 11
  03/10/76     14.57     golden       s

Our experience with just two categories and the inability of
the user community to interalize the distinction leads me
to expect that mode categories will mean more problems. You
will end up reading all notes or missing notes you wanted
to read which were mis-file. And you will have a flurry of
notes discussing the appropriateness of various notes. I
much prefer the single public notes file.

---------- response 12
  03/10/76     15.33     k mast       p

I too agree with just one notesfile! I also find I dont
spend that much time reading notes. merely because I have
learned to look at the 1st few lines and decid how
interested I am in the subject.
One thing that would make life easier, and is planned,
is to store the last time you accessed Public Notes. Then
by pressing a key you could taken to the 1st note which
you havent yet read.
Something else that isn planned, but has been discussed, is
to use the above method to show you all notes that have new
responses since the last time you read notes.

---------- response 13
  03/11/76     00.36     keith s       matha

If you're tired of notes, work on the CDC system! You'll
PACE back to the Urbana system! One General Note,
perhaps a couple Help Notes per day.

---------- response 14
  03/11/76     08.44     berger       mfl





I'm all for user feedback , and if the notes machinery
would automatically take me to the first note that I
haven't read yet, it would be a great help.

---------- note 137 autoshift
  03/09/76     17.56     weible       german

I have constructed a random sentence generator in order
to test students on their recognition of syntactical
structures. Everything words just fine, except I can't
figure out an elegant and simple way of capitalizing the
first word in each sentence. What I need is a screen
display equivalent of the -put- command which would
cause the first letter of every sentence to be shifted.
I suppose something might be done using -pack- and -packc-,
but I hope someone out there can think of a tidier solution.
Unfortunately, trying to -white- shifted ‎ by itself,
doesn't work (shift‎ does slow up as a hidden shift,
but its effect doesn't carry over to the next -write-
statement). Ideas, anyone?

---------- response 1
  03/09/76     18.16     feugen       mfl

Perhaps I have overlooked something in the question,
(but how but this? ):
    -calc- a variable (let's say n1) to 'o70', load the
first word of the generated sentence into a buffer which
immediately follows the aurementioned variable, then
-write- ¬0a,n1,18¬1 (rest of the sentence).
    Such a structure will work relatively simply if the
first word of your sentence never exceeds 10 characters,
and only a bit harder if you could have more than 10.

---------- response 2
  03/09/76     22.14     warner       iu

   Look up the -move- command. The basic structure
in this case is to
move  thing,1,thing,2,length
move  'C',1,thing,1,1
where 'thing' is the think to be -showa-ed to 'length'.
      The first move shifts 'thing' over one character.
Since 'C' consists of SHIFT followed by 'c', the second
move will take only the shift code and glom it onto the
front of 'thing'.

----------- response 3
  03/10/76     08.43     berger       mfl

I'm not certain, but I think that Weible's problem is that
he is using "write" or "writec" commands for his sentences,
and doesn't have the strings packed.

---------- response 4
  03/10/76     10.12     mcneil a       uicc





Mike is right, It's all ritec's. My idea is to convert them
all to pack's, move the packs into the 2nd char of a buffer
that has a o70 in the first char on the first word then
is cleared to o00.

---------- response 5
  03/10/76     13.32     rd rose blatt       hum

one of the elementary reading lessons (i saw it thru
sample) capitalizes the students name after he/she
types it in, i.e.;
        type your name here > margie
        Thank you Margie: now lets....
i can't remember the lesson name, but maybe you can
check into it.

dick rosenblatt

---------- response 6
  03/10/76     13.47     nate       iumusic

   Capitalizing first charactor of student's response
is extremely easy to do, and have noticed that on my
own programs, and Clark's. Also, I don't think that
technique is what's needed here (I was wondering if you
might want to store all the sentences in a dataset, and
search for a period as a seperator between sentences. From
there, it would be easy enough to find out which char to
put an o70 in front of it and -move- the rest of the
junk (also easy to find out charector length of sentence
for the -move- tag)).

---------- response 7
  03/10/76     16.02     bowery       comm

Mcneil's note is the right one... I would have responded
but your thing about "elegance" made me reconsider.
-packc-s and moves are not exactly elegant. Here is some
code that might work.

unit   packsent
calc   string‎o7000 0001 0000 0000 0000
       char‎2 $$ char position to -move- to
doto   1,word‎1,numword
do     packc $$ all the .writec-s converted to -packc-s
*packc returns word in pbuff and length in length
move   pbuff,1,string,char, length
calc   char‎char+length
1
at     wherever
showa  string.char

Its kinda schematic but it would work.

---------- note 157 output err
  03/09/76     23.48     bob       reading

i have a datafile with the same -outputl- repeated over





and over again for about two blocks -- about 150 entries.

the -outputl- is done in the router upon return from a
lesson. a second -outputl- is done at the same time; but
this outputl turned up only once. in other words, it looks
like one of the -outputl-s got into a loop -- there's
nothing in the lesson to cause this.

the teacher reports that the student got an execution error
which is not listed in the datafile.

this all happened at 9:10 this morning. the datafile is
-readata5-, but i'll have to clean out in the morning.

---------- response 1
  03/10/76     01.23     white       p

I've spent about an hour looking at your datafile and
router, and am not convinced that your lesson couldn't
have put out all those outputls.

1) after entry return1 you appear to output a trace
   outputl without outputing anything else, so if
   the student ever gets back to return1 without
   outputing anything else, he can loop and fill the
   datafile.
2) from unit return1 the student can apparently be
   jumped out to a lesson whose name you compute.
   I assume the lesson he goes to might return him
   to entry return1.
3) the outputls were done over about a 5 minute period,
   allowing guite a bit of processing before coming
   back to return1.

I've spend about a half acur looking at the execution
routines for outputl, and im not convinced that they
don't loop either.

---------- response 2
  03/10/76     07.52     bob       reading

i appreciate you looking so hard; i hope it wasnt my routine
which got into the loop. but i'll check more into that
possibility. unfortunately, it did happen after retuning
from one of the lessons that can't vouch for -- it is
possible that the lesson does something funny that the
router isn't expecting.

---------- response 3
  03/10/76     13.22     bob y       reading

well, i've finally been file to replicate the error and
its all mine. i dont quite understand why one of the
outputl's was not done, it I'm sure that that's my
problem too.

thanks for taking your time; sorry it was a wild goose
chase.





---------- response 4
  03/10/76     16.04     bowery       comm

larry doesn7T mind.. hahaha

---------- note 184 error
  03/10/76     13.11     lacoe       mtc

Weird error"!
At 1305 I tried to edit shp97 with the following results.
Author mode page typed shp97
pressed NEXT
message appeared

One moment Please--loading character set!

error in loading character set--
Lesson name: nu
Block name:

No block name specified
Press SHIFT-NEXT to continue

No block name specified
Press SHIFT-NEXT to continue

This was repeatable even after condensing.

---------- response 1
  03/10/76     13.27     ken       conn

        Your problem can be corrected by filling
out the information on the page found by pressing
DATA1 from the block directory. The charset
lesson and block name must be filled out if the
charset is to be loaded. Some more information can
be found by pressing HELP on that page.

---------- response 2
  03/10/76     13.29     stewait       arizona

        I got the exact ;same error in one of my lessons.
It appeared just after the Shift-DATA option was added to
the editor. Apprently some garbage got converted to the
lesson name for loading a charset from. Does any sys. prog.
know exactly what caused this????

---------- response 3
  03/10/76     13.32     marty smith       mtc

i just talked with bob, and he said the last time
he edited his lesson was month ago, and today was
the first time he*s been in it since then, soooooo,
there had to be some garbage in the directory for
this error to have occurred.

---------- response 4
  03/10/76     13.40     lacoe       mtc





Correction. I last edited the lesson on 12/5/75.
       Bob
---------- response 5
  03/10/76     13.47     hecht       ed

I received an identical error on Friday. The lesson space
which gave that error had not been edited since August.

The text of the error message was exactly the same--charset
from lesson no, no block name given.

---------- response 6
  03/10/76     13.50     nate       iumusic

    I reported this to Dave Frankel as this happened to
me (I reported it the same way the new edit options got
put in). Yes, it was a system-problem. Just clear the
info out (what Frankel told me) and sort of ignore it.
Dave?

Nate

---------- response 7
  03/10/76     15.52     bowery       comm

Sounds like the stuff in there is an artifact of the
pre-disk conversion days when the directory was in a 
different format.

---------- response 8
  03/10/76     17.05     frankel       p

No, what that is (believe if or not) is the "day" that
your lesson data page was last edited. To make a long
story short, the word currently used to hold the charset
lesson name was, for a couple of weeks, used to record
the day on which the lesson data page was changed. Since
then, the day info has been voved, but the garbage was
left, so the new charset/micro stuff picks it up. Just
clear it out, and ... I am, naturally, sorry for the
inconvinience.

---------- note 189 i/o error
  03/10/76     13.31     ken       conn

    DISK I/O error
    144 323
    Block number cor

    Recieved at 13:30, while in a gnote file.

---------- response 1
  03/10/76     15.34     k mast       p





Thanks.

---------- note 203 more delay
  03/10/76     15.07     roman       med

We have used the -delay- command  and along with -write-,
-at-, and -showa- to produce an interval of 2.100 secs.
Each trial is within one millisecond of the average,
And you all said it couldn't be done!!

---------- response 1
  03/10/76     15.18     avner       s

I assume that you have a means of timing independent of
the system if you are getting data on sigle trials
(otherwise, you may not be getting actual display times).

---------- response 2
  03/10/76     15.22     frye       mfl

In that case, you must be one of the lucky people who never
get line errors. If you don't, then what's the surprise?
Data comes to the terminal every 1/60 second REGULARLY;
a specific display (including -delay-s, which are actually
output to the terminal)will take the exact same time
if:
   1) you're not autobroken
   2) you receive no -talk- or other messages

Again, you cannot time ps__r__o__c__e__s__s__i__n_g this way! Even with
-catchup- (which is only accurate to about .5 seconds),
actually processing time can vary widely...
G. David Frye

---------- response 3
  03/10/76     15.22     midden       s

From your note, I draw the conclusion that you are
timing a display. If you plot the same stuff on the
screen. it should take the same time. (excluding the
little write of space in the lower right hand corner
every now and then) Unless you are timing the display
externally, there is no accurate way to determine how
long it takes to plot the display. If you have found
a way, I'm sure there would be many people interested
in knowing how you did it.
M,Midden

---------- response 4
  03/10/76     15.23     frye       mfl

Somehow I get the impression that we've managed to all
respond at the same time...





---------- response 5
  03/10/76     18.46     koning       csstaff

What previous responses saying "it can't be done' meant is
that your resolution is only 1/60 of a second, not that
accuracy is that low (apart from line error and similar
problems).

---------- response 6
  03/11/76     10.48     perry       uw

From golden:

It is not nearly that precise. Once started, the duration
of a -delay- should be correct to the nearest 1/60 of a
second. But the exact starting time is unknown, and, if
line errors occur during a delay, the timing can be thrown
off by almost a second.
******

  The key words are !once started." In your case you
sent something to the terminal, which "started" and output
string. Before the string was finished being sent to
the terminal, your program called for the delay, which was
then tacked on to the string. And before the delay was
finished, your program tacked on the second string. Thus
you have a very precise timing (provided you never get
autobroken, etc. during this process).

  We have been using such a device to present and then x-out
words in a vocabulary test. Line errors seem to be the only
bug.

---------- note 217 ifthen
  03/10/76     19.20     phil mast       p

        A file 'ifthenI contains many notes
describing changes that might be made to allow
TUTOR source to be indented. It also describes
several proposed commands which will use
indenting to make programming easier and more
readable.

Anyone interested is welcome to comment on
these proposed changes.

---------- response 1
  03/11/76     12.29     friedman       csa

Great subject, and a good discussion just barely started,
but the file is full! Please, some more space?!

---------- response 2
  03/11/76     13.31     phil mast       p

Ok, its length has now been doubled.





    Phil

---------- response 3
  03/11/76     14.48     friedman       csa

Great, thanks!

---------- note 220 cond's/3
  03/10/76     20.58     struven t       csstaff

Would it be possible (feasible) to modify conditional
commands in the following way?

Instead of:
do var,uneg,x,x,x,uone,uone,uone,x,utwo,utwo,x,x,x,
   uthr,uthr,uthr,ufour,ufour,ufour,x

An equivalent form would be:
do var,uneg,x/3,uone/3,x,utwo/2,x/3,uthr/3,ufour/3,x
(some other sign may be used besides '/')

It seems that it is logical and would save space. Would
it be that hard to implement?
Terry

---------- response 1
  03/11/76     08.41     avner       s

It is certainly logical but it would save only disk space
(a very low priced commodity) and not ECS or CM space, I
have a personal preference for the current method since it
is much clearer and makes it easier to debug branches and
document code when working from printouts. (this is not
an official statement) Al Avner

---------- response 2
  03/11/76    13.52     struven t       csstaff

O.K...thanks for the response.

---------- note 223 -reedit-
  03/11/76     01.41     halvorsen       pcp

Issues don't stay in general notes too long since so
many notes push everything older than 2 days off the
first directory page. Nevertheless, I will start
something again.

From the past discussion, it seems that a -reedit-
command would be the best solution for everyone who
has the same problem about text editors and re-editing.
This would just have the effect of simply setting up
the arrow machinery so that it would seem as though
the student has already entered the text specified
by the -reedit- command. This would greatly simplify
the problem of writing simple text editors that don't
need to worry about line numbers.





Any systems response to this idea?

---------- response 1
  03/11/76     09.34     gilpin       peer

Maybe it would help if you were to put your editor in a
file where people could look at it. I have seen one brief
demo of it, and it looked VERY nice. If others felt the
same, it would provide motivation for your request to be
worked on.

---------- response 2
  03/11/76     11.23     judy       pso

Halvorsen, you know and I know and Gilpin knows that
systems ALREADY knows about the utility/need/desirablity
it up in notes, except to use notes space?

---------- response 3
  03/11/76     12.54     gilpin       peer

Before making my earlier response, I looked thru all the
notes titles in this file to find the earlier discussion(s)
of Halvorsen's request. Just now I have extended the
search back to the beginning of January. I'm pretty sure
it/they are in there somwhere, but cannot find it/them.
I'd be greatful, Judy, if you would steer me to it/them,
and especially to the systems response(s) you have in mind.

In any case, I stand behind my suggestion, Assuming that
the editor Halvorsen is talking about is the one Marty gave
me a glimpse of recently, it embodies a fresh (to me anyway)
approach to the design of message editors for untrained
user, and it is nicely done. It is worth a look by the
rest of us.

---------- response 4
  03/11/76     14.55     k mast       p

I believe Bruce Sherwood originally proposed the -reedit-
command in an earlier discussion of this topic. I expect it
to implemented. when is another question.

---------- response 5
  03/11/76     15.29     elston       rhrc

The recent discussion with the -reedit- suggestion is
fund by entering 3/7 and looking at the note titled
"press copy!. Bruce Sherwoods response is number 9.

2c≤/

---------- note 224 parts
  03/11/76     02.56     blomme       s





   The displays (and some of the operation) of the "parts"
option in the editor have been changed. Hopefully this
is SOME improvement if not yet ideal.

---------- response 1
  03/11/76     10.30     koning       csstaff

Nice... Maybe this could (should?) be done for the delete
blocks display as well.

---------- response 2
  03/11/76     10.32     celia       pso

I like this changes very much.

---------- response 3
  03/11/76     10.40     mike b       cornell

        This is much better than the old way! In fact,
it's so much faster that I hn't even have to request the
ability to move between parts while in the "partial mode".

M¬ ike≤≤≤≤ˇB

---------- response 4
  03/11/76     11.00     perry       uw

  I like it too.
  Is necessary to replot the whole screen when you are
finished and press NEXT? It would seem that erasing the
"Help available" line would be quicker and just as effective
in notifying you that you were out of the P-option.

---------- response 5
  03/11/76     12.29     gutfreund       css

I like it too. Has any thought been given to allowing one
to page forward while setting the condense flags. This
definitely make it perfect in all respects.

---------- response 6
  03/11/76     13.29     berger       cerlcc

Thanks - anything that speeds up the use of the editor is
appreciated.

---------- response 7
  03/11/76     14.52     friedman       csa

Very good. One other suggestion:↓
It seems the HELP is given using a -helpop-. Why not put a
blank-tagged -write- at the end of the helpop unit so that





it doesn't take up all that the erasing the help as soon
as you use it? That way, the help stays one the screen as
long as you want to use the feature. (Of course, this
makes it desirable to continue to replot the page when done,
or at least to erase all that stuff.)

---------- response 8
  03/11/76     20.29     blomme       s

   Some replies:
(1) Yes, probably both the block delete and block recover
  options should have a simpler original message and a more
  detailed explanation--maybe another day (month, year?)
(2) The replot of the block display is sort of related to
  the fact that the "parts" option is in a different lesson
  from the main one and the simplest, cleanest, and easiest
  type of connection is just to return to the point of
  replotting that main block display. This might be
  worked on some day.
(3) Using the number keys to select another part while
   remaining in the parts option would be nice (but again
   poses some problems since the basic block display
   generator is in another lesson).
(4) Thanks to Dr. friedman (r his suggestion re pre-
    venting the erase of the help information--this has
    now been done.

---------- response 9
  03/12/76     11.03     perry       uw

  Thanks for the response, Rick.

---------- note 225 want/wish
  03/11/76     08.05     walter       physio

Since people have been suggesting commands lately, I will
toss in my own suggestion.

I suggest a change in the structure of logical operations
such that a line of code such as

        (x=1) $or$ (x=5) $or$ (x=8)

could be written in a  more condensed form such as

        (x=1:s:8)

This form is still readable and is unlikely to be confused.

Oh, and while I am here, any chance of using just "v" when
specifying a vertical segment?

---------- response 1
  03/11/76     10.27     mike b       cornell

        The idea is interesting, but it probably wouldn't
save any condensed lessons space. I also think the current





syntax is more clearly readable. However, I _would_ like 
to see the ability to use "¬+" (ACCESS-+) and "¬I" (ACCESS-I)
as logical operators implemented. It has also been
suggested that "↑≡" and "↑≠" (the shifted ≡ and ≠, respective-
ly) should be able to serve as the logical intersection
($mask$) and the logical union ($union$) operators. These
would be available for those who prefer them to the less
compact $operator$ format. There is precedent for this, in
that the "*" is valid for multiplication and "c" for comment
lines, both, I assume, holdovers from FORTRAN.

M¬ ike≤≤≤≤ˇB

---------- response 2
  03/11/76     11.25     judy       pso

Gosh, I couldn't program without * ! It is so much
easier to read than x≠x ...

What would "¬I" mean?

---------- response 3
  03/11/76     11.39     shirer       s

  It has always been intended eventually to add logical
and bit union and intersection operators as ALTERNATIVES
(not replacements) for $operators$. It is such a low
priority item that more important jobs have continually
muscled their way in front.

 Also, sometime in the future, something may be done
to shorten the define segment and array specifications.

Probably this won't happen until they are improved in some
significant (and for now unspecified) way, so don't hold
your breath. Just tell yourself that typing -ertical-
improves your digital dexterit) and only takes up
i second (.2 second for RB) for good typers.

---------- response 4
  03/11/76     12.32     friedman       csa

Judy, ¬I would mean $or$, and ¬+ would mean $and$.

---------- response 5
  03/11/76    14.52     k mast       p

Note that the reason the $operators$ were used in the
ist place was because the original terminal did not have
such any other logical charactors

---------- response 6
  03/11/76     22.21     kent       unl

  Why not (I'm gonna get lambasted for this) implement
PL/I-like structures for the -define-s? Really a very





elegant way to define extremely complicated "data struct-
ures"!

---------- response 7
  03/15/76     13.59     bowery     comm

Perhaps the priority of "jobs" that systems people have
to do should be partially determined by the amount of time
they have spent on the "job board!....

Also an argument can be made against the ¬I symbol being
used in anything BUT pl/1 since that is the only place it
is used in technology to represent a logical operation.
Its evolution could be considered much akin to that of
$operator$ I would venture.... only TUTOR did a better
job of using what they had t) avoid computerese.
(re kent: wha?)

---------- note 229 goodby
  03/11/76     08.47     jordan       english

Jim Williamson has accepted a position with Educational
Services in Baltimore and will be leaving March 26. If
you would like to say goodby to Jim, I am having an open
house for him from 6 to 8 on Wednesday, March 17. Stop
in at 511 West Clark in Champaign. (His sign-on is
jimw/english if you prefer to say your goodbys via the
electronic medium that has s) effected our life styles.)
Pauline Jordan

---------- response 1
  03/11/76     10.26     golden       s

Good luck, Jim. If that company doesn't do well selling
PLATO services, it certainly won't be because of a failure
to attract good staff.

---------- response 2
  03/11/76     15.57     carter       comm

If anyone is interested in explaining this company, I
(and probably many others) would like to hear about it.

Incidetly, a mil spec plasma panel is on the market, in
case you want to build a terminal that will operate in
adverse environments and Xerox is using a panel in
a text-editing--photo-composition system.

---------- response 3
  03/12/76     08.27     jordan       english

Educational Services is a subsidiary of Commercial Credit,
itself a subsidiary of Control Data Corporation. Ed. Serv.
is opening 6 Learning Centers this year in N.Y., Philly,
D.C.,Baltimore, Clevelaid, and Dallas to market PLATO time
to business and industry, If you want more details you can





call me (3-7450). Pauline

---------- response 4
  03/12/76     11.19     carter       comm

Thanks, I was just wondering if anyone other than CDC
was jumping on the bandwagon.

---------- note 242 hardcopy
  03/11/76     11.47      johnson       research

A limited number of plasma display hardcopy units similar
to one that is available on the first floor of CERL are
available at reduced prices, For additional information
please contact Mr. Richar) M Kaig at Owens-Illinois
(419) 242-6543 x66416. If you have questions regarding
the performance of the unit and the attachment of the unit
to a PLATO terminal, please contact johnson of research.

---------- response 1
  03/11/76     13.09     maggs       law

Is this an attachment to a PLATO terminal or does it
include a PLATO terminal?

---------- response 2
  03/11/76     13.31     berger       cerlcc

If it is anything like the others that were available for
sale, it was necessary to slave it to an existing terminal,
but it in no way changes the normal operation of the master
terminal.

---------- note 253 routercost
  03/11/76     14.47     mcneil e       uicc

How much ecs charge is made against a site for using
a system router for one course? For several courses?
Per student in one course?

---------- response 1
  03/11/76     15.06     golden       s

As of now, the cost is 50 words per student user independent
of which course he is in. Sometime soon this will change
as we charge for the ECS used by leslists, including those
used by the system router and we charge for the system
router common. The exact charge will be determined by the
lengths of those leslists and commons. (Yes they will be
of variable length.)

---------- response 2
  03/11/76     16.49     nate       iumusic

    dumb questions time: What is going to be the advantage





of using leslists when we start getting charged for them?
Are leslists going to so of become obsolete (forgive
me for my non-knowledge, but it seemed that less ECS than
common was about the main reason to continue with leslists)?

Nate
---------- response 3
  03/11/76     17.21     marty smith       mtc

well, i for one will stop using leslists once we*re charged
for 'em. they won't be do any more use then, as i'll change
everything into common.

---------- response 4
  03/11/76     18.27     michael       client

Keep in mind that the system will not "publish" anything
that uses variable-based jumpouts. Leslists will be the
only way to go if you want it listed in the upcoming system
catalog.

---------- response 5
  03/11/76     19.46     golden       s

The present plan calls for leslists to be the vehicle which
connects the system catalog with routers--your router, my
router, any system routers. There needs to be some such
standard place and standard format for file information.
And it won't cost more in a leslist than in common if they
are variable length.

---------- response 6
  03/12/76     10.11     john risken       cdcc

Also note that in the event of transfer between systems,
all variable references to resources like common, charset,
lineset, dataset, micro, and leslist will have to be done
via a leslist reference so that the transfer machinery will
know exactly which lessons to transfer.

charset (n1),mychars

won't work since the transfer mechanism won't be able to
execute your code and find out just what n1 contains.

---------- response 7
  03/12/76     11.55     michael       english

I assume, then, that the system is going to modify commands
like dataset, micro, charset, etc. to accept leslist argu-
ments?

---------- response 8
  03/12/76     13.57     tom s       mfl





         It would be about time they did!

---------- note 255 primary-il
  03/11/76     14.48     simons       cs196

         last night, WCIA (channel 3) ran a commercial
         about using plato for the illinois primary.

         exactly, what is plato doing; making an estimate
         of who wins, or what ?

         and since i won't be in town at the time,
         how much will they be using it. i.e. the viewers
         won't have to be staring at a picture of a plasma
         panel for 30 minutes, will they ?

         and what if (hahaha) plato goes down ?

         also, will this use be given priority usage of
         plato.

---------- response 1
  03/11/76     15.15     golden       s

PLATO has been used by several television stations in
past elections and will be used by at least two next week.

PLATO stores incoming results, past results, and provides
displays. In the past the displays were taken directly off
the plasma panel. WCIA will try feeding their own video
character generator from PLATO this year.
If PLATO goes down, central Illinois won't get timely
results since both CBS and NBC affiliates are depending on
us.

In my own very biased opinion, the 1972 results displayed
via PLATO were more often synchronized with the live
announcer's voice than were the displays from the national
network.

---------- response 2
  03/11/76     15.39     jim g       reading

The PLATO displays were a lot better than those dumb paper
tally sheets with hand scrawled numbers that WCIA used for
some smalltime election results or even the big score
boards. I also noticed that the famous (?) Paul Douglas
was able to jump around from previous results to current
tallys almost at will to give a much better picture of
past and present trends.

When will David Frankel replace Mr. Roberts for the weather
report?

---------- response 3
  03/11/76    18.21     frye       mfl





Soon, I hope. That poor old man is going to collapse from
lack of air any day now...

G. David

---------- note 258 base
  03/11/76     14.54     s zweig       iusn

I want to make every unit in my lesson a base unit. (I
have terms in each unit and want to be able to get around
easily.) Is there anyway to do this without physically
placing a base command with the unit name in each and
every unit? Mayhaps with imain and a (variable) tag for
-base-??!?
Ʃteve
---------- response 1
  03/11/76     14.57     friedman       csa

In your ieu, include:
imain baser↓
Then have:
unit  baser
base

The blank tag clears any base unit; thus any main unit
is made the base unit.

---------- response 2
  03/11/76     15.04     s zweig       iusn

That oughta do it. I had some wires crossed and looked
right past the obvious. Thanx.
Ʃteve

---------- response 3
  03/11/76     16.49     p cohen       med

        Well, if no one else is going to say anything...

        There is a big difference between SETTING a base
pointer with -base <unitname>- and CLEARING the base
pointer with -base-.
        Putting -base -at the beginning of each unit
(or in an -imain- unit) will NOT set each unit as a base
unit, but will clear the base pointer so that no unit will
be a base unit. But setting -base- to the current mainu
will set the current main unit as a sort of a help sequence
unto itself. I.e., BACK will replot the unit! (This is
actually useful when using -inhibit erase-d -jump-s, so
that NEXT will not replot the screen, but BACK will--it
ends the help sequence!)
        Incidentally, to set each unit as a base unit,
there seems to be no way around including a -base- com-
mand in each and every unit. (cont'o.)

---------- response 4
  03/11/76     16.56     p cohen       med





        (cont'd. from last response)

        I think Ʃteve _really_ wanted to CLEAR the base
pointer in each unit, so I hope my long-winded explana-
tion was not out of line.

--paul

---------- response 5
  03/12/76     09.35     jmk       pso

In a main unit, if the base pointer is NOT set, then that
unit IS a base unit. If the base pointer is non-zero, it
contains the name of the base unit. The base unit is
where you go if you press -back- (if you haven't set the
BACK pointer with a -back- command) or shift -back-.
If you don't go anywhere when you press one of those keys.
then you are in a base unit.
Putting a -base- command with no tag makes it a base unit.
Thus the first response to this note (imain unit with a
base (blank) command) was correct.

---------- response 6
  03/12/76     12.41     judy       pso

unit test1
base
unit test1a
*
*
unit test2
base test2
unit test2a

We all seem to agree on the ideas, but not the semantics.
The important difference between test1 and test2 is in the
behavior of test1a and test2a. By naming "test2" in the tag
of the base, we _cause_ test2 to be the first unit of a "help"
sequence, so that a BACK keypress in test2a will return
to test2. A BACK keypress in test1a will not return to
test1.

---------- response 7
  03/12/76     16.15     friedman       csa

I believe I said "clear" the base unit in my answer, and
I think that's what Steve wanted to do.

---------- note 267 -compute-
  03/11/76     16.22     clark       lawyer

what can cause an execution error in the -compute- command
other than trying to compute a string over 99 characters?

(also exclude all of the regular index errors)

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





  03/11/76     16.29     kane       phys

A negative value for the length of the string.

---------- response 2
  03/12/76     15.53     gutfreund       css

I assume you were not playing with the pointer.

---------- note 270 -comload-
  03/11/76     16.48     seager       mtc

lesson has common of 2254; hence must use -comload-↓
need analysis of following "problem":↓
comload nc1,1,200
        nc201,601,6
        nc207,645,1200↓
        above code dumps data that is supposed to be in
        nc601 into nc1039!↓
comload nc1,1,200
        nc401,601,6
        nc407,645,1000↓
        above code dumps data that is supposed to be in
        nc601 into nc839!!!↓

---------- response 1
  03/11/76     16.58     seager       mtc

This code is used to transfer data betweeen CM and ECS.↓
For the 1st example, if a variable used to increment the
number of area summaries read via -readd area-, numdata,
>1199, the common is changed via a unit that contains:↓
comload nc1,1,200
        nc201,601,6
        nc207,1845,300↓
        (above code applicable only to the 1st example)↓
anyhoo...why should the data be offset on both cases?↓
the lesson name will be given upon request, needed for
debugging.↓
        thanks,↓
brent seager

---------- response 2
  03/12/76     10.12     mike b       cornell

        I think you have the adresses wrong way about.
The first argument of the comload is the CM address _to_
which you want the common loaded. The second argument
is the ECS address _from_ which you want the common loaded.
Unlike -block- and -transfer-, which have "from, to" syntax,
-comload- has a "to, from" syntax.

M¬ ike≤≤≤≤ˇB

---------- note 276 ¬$456error
  03/11/76     18.15     putch       arizona





********
exec error 03/05/76 22.15 koning csstaff

I just got a really strange execution error:
        Execution error in lesson - vav
        unit - hidarow
        State - regular (post-judging)
        Last command - unit (embedded write)
        0th command (...)
        Error type - (blank)
What I don't understand is (1) why I get an error in a
-unit- command (so far as I can see no one passes para-
meters to the unit), (2) what the 'embedded write' means.
Does anyone know what is going on?
↓PS. There _are_ bugs in the lesson, but I can't see how
they cause this. Since I fixed those, I can't reproduce
this error.
********
     It probably isn't really important now, but I was
wondering if anyone ever figured it out.

---------- response 1
  03/12/76     11.15     koning       csstaff

I would still be interested too.. All I have heard so far
from the several people asked is that the error handling
in Plato is not quite uniform enough and that there are
cases where an incorrect message can be given.Unfortunately
it seems that the unit with the problem does not contain
any commands for which this is the case.

---------- response 2
  03/12/76     12.49     judy       pso

It is true that some error judgments come out and
people "just don't know".

If you can make a reproducable error message like the
funny one you show, then maybe some intelligent guesses
could be made.

---------- note 279 line/rdraw
  03/11/76     18.49     pickens       iu

I am not sure how the lineset plots characters but I
suspect that it uses an origin of some sort. If this
be the case would it be possible to have an edit option
to convert a lineset character to -rdraw-and the
converse?

arcon

ummm....it would be necessary to "count" the number
of lines used in the -rdraw- and refuse if ¬$ > than





lineset can handle.

---------- response 1
  03/11/76     19.25     fortner       p

The option to convert from linechars to rdraws and vice
versa will be done someday...

---------- note 280 disc error
  03/11/76     19.12     young stephen       edpsy387

Disc I/O Error

FIW not set

Occurred when returning from help pages to block
directory page.

---------- response 1
  03/12/76     10.00     k mast       p

Thanks.

---------- note 286 halfspace
  03/12/76     00.44     platt       cornell

In mode rewrite the ACCESSΔSPACE halfspace does not appear
to erase the underlying character. I expected it to work the
same way as the regular space would. I'm I misusing this
feature or is there a system bug?

---------- response 1
  03/12/76     03.15     sellers       arizona

Check aids...
Says there (somewhere) that half spaces won't mode rewrite.
(that is how I learned about it >ie from Aids<)
Is there an accompanying whole space?
If there is you can use space, half-backspace, space.

Anyway that makes sense in that the terminal character
size is 8≠16 dots and not 4≠16 dots.

---------- response 2
  03/12/76     03.26     norm       rotc

To be specific, it says on the
first page of mode rewrite at the bottom
of the page (DATA ¬D mode)





Your helpful fishswine

---------- response 3
  03/12/76     05.04     t little       research

The reason the half(space/backspace) don't erase in
mode rewrite is that they aren't actually characters.
They are special codes to the frameater which tells
the terminal to move by 4 dots.

Todd

---------- response 4
  03/12/76     09.14     j wilson       unidel

And, incidentally, systems people have cautioned users
not to count on this current situation always being the
case.

---------- response 5
  03/12/76     11.34     warner       iu

        If you have to mode rewrite using half-spaces,
space and then half-backspace.
Example:
000000000000000↓¬↑ˆ$  half¬ space
with a half-space between the two words.
000000000000000↓¬↑ˆ$ half ¬≤space
with a space and half-backspace.

---------- note 300 auto-sign
  03/12/76     11.35     crcege       uicc

Had something happened to the Auto-signon Stuff? Today
I discovered that all of our assigned auto-signons were
blanked out. I had not done it and wonder how it could
have happened.
Thanks.
Sally Droege

---------- response 1
  03/12/76     11.49     berger       cerlcc

This has happened at the language laboratory before
too.

---------- response 2
  03/12/76     15.27     white       p

Some space had to be cleared on the autosignon lists
temporarily about a week ago. An attempt was made
to only clear author signons, but a few student signons
may have been cleared accidently. Sorry for any





problems it may have caused.

---------- note 306 rdraw bug
  03/12/76     15.17     funk       cs317

The following two blocks of code should produce identical
figures, but don't:
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::
  draw 1632-204;1632-196;1632+204;1632+196;1632-204
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::
  at 1632
  rdraw -204;-196;204;196;-204
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Why? Can it be corrected?
   (The first draws a square; the second locates the
   upper right and lower left corners at the extreme
   right of the screen.)

---------- response 1
  03/12/76     15.33     white       p

Sorry, "-204" and "196" are illegal coarse grid point
specifications. You probably only want to use fine
grid tags on rdraw commands.

---------- response 2
  03/14/76     12.43     danielson       csa

But why did the points specified with -204 and 204 plot
correctly, when the points specified with -196 and 196
did not? This happens for all arrangements of the points
which form a square, so it seems to be connected with the
value to be added or subtracted from the origin.

Ron

---------- note 310 POLL!
  03/12/76     15.56     curly       iu

  Announcing a opinion poll in lesson 'curly'.
Polls will remain open for another 12 days or until
my common fills up(after 300+ people have entered).
  This is the same announcement I put in pad and gpad.
Wouldve put it in here too..cept i was so tired i
forgot all about public notes at the time.

curly≤≤≤≤≤¬↑ˇ↑ˆ↑ˆ↑ˆcurly

---------- note 312 imode
  03/12/76     16.17     meers       wright

We had a strange problem here this afternoon, involving
an instructor signon, and in ECS shortage.. Rita signed
into her course with an instructor signon at terminal
22-11... She worked for a while, and our ECS allotment
dropped to its base.. As she attempted to transfer





from one section to another, she got an ECS allotment
message, and was bounced out of the course. Being
bounced out that drastically did NOT de-assign the course
from her signon.. As a result, I tried to edit the
course a while later and got the message that someone
at 22-11 was editing the course and that i could only
get inspect only... There was no one at 22-11, and
I was mystified for a while, but managed to figure out
what had happened.. I signed on to the terminal
with that same instructors signon, signed out, and
everything was fine once more,↓
(continued in 1st response)

---------- response 1
  03/12/76     16.29     meers       wright

    The cause of this problem seems to be with the way
that instructor signons are charged for ECS. Once an
author gets on, he may edit courses, write notes, edit
lessons, and condense lessons of 1500 words and less..
    Instructors do not get these same courtesies.
Lesson 'imode' is charged at 1500 words.. If an
instructor ALSO tries to get into her course, or
read/write a personal note, she is charged an additional
1500 words. As a result, instructors can oftentimes
get on, but if ECS is tight, cannot do anything that
requires more than the 1500 words.
    I would like to suggest that this be handled in
the same manner as authors, where 1500 words is THE charge
for anything less than c indensing a lesson.

---------- response 2
  03/13/76     11.50     b sherwood       s

Right. Will look into this. Thanks for the info!

---------- note 314 nookno
  03/12/76     16.31     neal       wright

when using 'nookno' the 'no' is erased after a no
judgement even though it has not been written there.
this destroys any screen display which appears there.
am about to try noword blank to eliminate this, but
it should be changed at the systems level.

---------- response 1
  03/12/76     17.16     clark       lawyer

There seem to be a number of these erase problems related
to the arrow command. Ever try putting an arrow in the
lower left-hand corner of the screen? It won't get erased
as most arrows...there a also quite a few other problems
related to the specs noo o...is there any chance that
these problems will be of some system programmer's
schedule?

---------- response 2
  03/12/76     20.01     blomme       s





(1) ALL problems are on some system programmer's list
  to be fixed some day. Some of the less critical ones
  may not be fixed for many months because of more pressing
  work. It is of course conceivable that some things could
  get lost in the shuffle over time occasionally.
(2) There are NO known problems with the specs nookno that
  have not been fixed. the problem "described" here
  actually involves the following elements:
  (a) specs nookno
  (b) sized writing
  (c) the pressing of the SHIFT-ERASE key (or NEXT in
   the case of a wrong answer)
   See a note of 2/24/76 where this problem was mentioned
and the note of 2/26/76 where its correction is noted.
However, it is unfortunately the case that as of this
moment, that correction has not yet made its way to
the prime time version--but we expect to update i in
the next day or two.

---------- note 315 nookno2
  03/12/76     16.40     neal       wright

noword blank solves the problem mentioned with nookno
in my previous note.

---------- response 1
  03/13/76     16.17     rader       s

did you read Rick Blomme's reply to your previous note?

---------- note 316 concepts
  03/12/76     16.56     ahasio       med

"It would be nice if" you could ignore specific words in a
concept. Presently vocabs allows you to forever ignore words
or forever consider them important. I often find that words
which are important in one answer are ignorable in another.

A simplified example would be an answer in which concept "C"
is the correct answer but including "a" and/or "b" would
also be correct though more specific than required. This
answer must be specified as:

concept c
        a c
        b c
        a b c

This would be better (especially when 3 or more concepts
were ignorable):

concept c <a,b>

---------- response 1
  03/12/76      17.12       michale     cornell





I think that this has been requested before; I believe
Paul Tenczar said that it was useful enough to implement,
but that he would have to re-implement the entire concept
machinery, which would take perhaps a year or more.

K↑ˇi↑ˇk↑ˇe  O↑ˆl↑ˆt↑ˆz

---------- response 2
  03/14/76     02.05     tenczar       s

michael is essentially correct, except that I said it
"might take a year to get around to" and not "a year
for re-programming the concept command"

---------- note 318 endings
  03/12/76     17.15     ahasic       med

   "It would be nice" if endings could contain something
similar to a phrase, For example:

endings 0,ate,ic,ic,*acid would allow
 
vocabs  acet//0 rather than

vccabs  (acetate,acetic,acetic*acid as it now stands.

In the case of phosphoenolpyruvate its utility becomes more
than obvious.

This problem and the previous one about concepts have bugged
me before, However I never bothered to write a note since I
knew that there was no chance of of anything being done before
the lesson which needed them was completed, But these prob-
lems have come back to haunt me once too often. Perhaps now
my children will never have to face these trials.

---------- response 1
  03/12/76     19.07     mike b       cornell

      Both of these requests seem reasonable and extremely
useful to me. But, alas, as per the first response to
the previous note, Think it will be awhile before they
are implemented.

M¬ ike≤≤≤≤ˇB

---------- response 2
  03/12/76     21.38     judy       pso

Just out of curiousity, how many letters per word is the
spelling judger prepared to process?

---------- response 3
  03/13/76     08.49     maggs       law





While we were at it, I would like to be able to
do the followng:

word*ending/s

---------- response 4
  03/14/76     02.13     tenczar       s

just to say that I read these suggesions and will keep
them in mind.

some perhaps random thoughts follow:
while it might seem to some that the concept, endings and
phrase algorithms are trivial...they sure haven't been
trivial for me to implement. they have slowlyl evloved
in a cyclical nature: from a simple state to a more complex
state..that when operational becomes obvious and thus
simple again...and once simple allows for further
complications to be implehented, etc.
so, as soon as I catch a breath and see that the current
state of these features are simple, I can procede on another
cycle...

---------- note 322 funnkeyu
  03/12/76     20.34     nike b       cornell

      Suggestion for a new command:

      Currently there is a weakness in the TUTOR language
that is most easily seen in this example. Suppose you
have a lesson with a large number of student-initiated
branches available, and suppose that you have an -imain-
unit which advertises something at the bottom of the screen.
Currently, I do not believe there is any really workable
way to turn of the -imain- for selected branches, but
leave it active for others. Similarly, there may be
certain initializations you want to perform in certain
branches ut not others. What is needed is aunit which
is executed _before_ the new main unit. but after the func-
tion key press. This is logicall very different from the
action of the -imain- unit.

    [Continued in first response]

---------- response 1
  03/12/76     20.43     nike b       cornell

         I suggest the -funkeyu- command with the following
syntax (suggested by michael/cornell):

        funkeyu unit, next,next1; unit2,help,back1;etc.]

with the conditional form:

        funkeyu expression, (unit1,next,next1;unit2,back1,
                help),(unit1,stop,data,help;unit2,data1.
                help1),[etc.]

        I would welcome a less complicated suggestion for





the syntax of the conditional form.

     Could we have a system response to this idea?

M¬ ike≤≤≤≤ˇB

---------- response 2
  03/12/76     21.08     warner       iu

         In my lesson 'energy4' there is an example of the
use of an imain unit that tests for the value of "key"
when it is branched into. If key=next it does different
code than if key=back.
         This code has problems, however. In certain cases
the student must return from a help sequence with the
effect of a -back- press. If the student presses -NEXT-
at an -end help- wrong branches might be taken. This means
that I must use the following kludgey code:
pause   keys=next,back
press   back
end     help
        Perhaps what is needed is some easy way to test
if a student is in a help nit, or perhaps an -ibase-
unit done whenever one enters a new base unit.

---------- response 3
  03/15/76     00.20     k mast       p

Would Warner's solution solve your problem? I much prefer it
to your proposed format!

I'm not sure what to do to solve your problem, Silas.
Perhaps you could look into trying to use the system defined
word 'ahelpn' or 'ahelp' (I dont remember which one it is)
to keep track of when a guy got help.

---------- response 4
  03/15/76     09.13     mike b       cornell
   
        I certainly could write could of that sort (this
solution occurred to me), but as warner admits, it is
kludgey. The -funkeyu- command would solve this problem
gracefully and be a logical addition to TUTOR, which has
developed a rather sophisticated ability to catch and
act upon the student's every keypress.

        If you mean his -1base- command, yes that would
solve my currnet difficulties, and I'm sure it would be
much easier to implement. It is not as flexible as my
suggestion, but I would consider it sufficient.

M¬ ike≤≤≤≤ˇB

---------- note 324 topics err
  03/12/76     20.52     mike b       cornell

Fatal Condense error in lesson "topics":





      Error in -use- command: bad block
	
      Got this at 2051 CST this date.

M¬ iky≤≤≤≤ˇB

---------- response 1
  03/13/76     01.49     blomme       s

    Thank you; the problem with the bad use block has been
eliminated, but theere are some other errors as well--
I think that some changes are in progress and I assume
it will all be fixed in the not too distant future.

---------- response 2
  03/14/76     14.19     tebby       pso

Sorry about the condense error in "topics". The error has
been corrected.

"topics" will soon not be an "ecs" monger, We are redoing
it so that it will be an "Id$" type lesson, but the error
Friday was not due to those changes, but to something else.

		Tebby

---------- note 327 unadvertis
  03/12/76     22.50     nate       iumusic

    I realize it was an unnadvertised feature, but just
the same, what happened to the option to load a micro
in the pnotes-editor?
Nate

---------- response 1
  03/14/76     11.50     s zweig       iusn

	   I'm just guessing here, but I'll bet that since
you can now jump around between public notes and personal
notes, the some folks were using 'animicro' and the like
for fancy signatures in public notes. Since this is frowned
upon, the feature is now gone. (Psst...am I right??!?)
Ʃteve

---------- response 2
  03/15/76     00.12     k mast       p

see my response to note 343

---------- note 329 Rus. lset
  03/13/76     00.24     tower       udtutor

Does anyone know where anrossion (cyrillic alphabet)





lineset for sized writing can be found??????????????

---------- response 1
  03/13/76     08.52     clark       lawyer

try lesson "lawchars"...part 2, block b ("russian")

see if it is what you want...

---------- response 2
  03/13/76     11.53     judy       pso

There is also a cyrillic lineset used by the UI Russian
courses.. contact curtin/mfl.

---------- note 331 -vocabs-
  03/13/76     03.13     luke       cornell

         When working with -vocabs- blocks,
 I get a condense error (understandably) when
 I try to put the -vocabs- block AFTER the arrow that
used it; but when I have an -endings-, I also need to
put that before my -vocabs- block!
*
       Thus, for one arrow. I have used 3 blocks.
is there any way to avodi this hassle?
(also, who wrote the -vccabs- block editor? I think it
could use some improvements badly, although the basic
workings of it I won't fault.)
	 Also. are there any lessons written on the
useage of -vocabs- ?

Luke Kaven

p.s. My problem was solved by converting the -vocabs- to
     source, and addiang the -endings- there; but now,
     when I want to use the -vocabs- editor, I have to
     convert.

---------- response 1
  03/13/76     11.14     woolley       p

Ther's no way to get around this restriction now.
Eventually, though, you will be able to reference a vocabs
block in the same sort of way as a common block, so it
won't matter here the block is located in the lesson.
In fact, it could even be in a different lesson.

---------- response 2
  03/13/76     18.12     nike b       cornell

         I agree with luck regarding the -vocabs-editor.
In its current form it seem; only marginally useful. I
realize that, with the adven of referncable blocks (as
per the system response), there will be a real incentive





to use the -vocabs- block. Are there any plans for
additions and/or changes to the editor?

M¬ ike≤≤≤≤ˇB

---------- note 337 resolved!
  03/13/76     14.06     seager       mtc

for those interested in the problem stated in note 270,
my errors revolved around two points:↓
        1. my defines were not addressing the same
           "location" in both CM and ECS common;↓
        2. more critical, an "offset" had to be -calc-ed
           when transfering the second part of my ECS
           common into CM, which was necessary, i guess,
           to address the proper location when returning
           the common to ECS. aids does not point this
           out at all!↓
i have thanked personaly the individuals that expressed
interest in my "problem"...*systems* response was just
overwhelming! perhaps i should have addresed the pso
group?

---------- response 1
  03/013/76     14.32     clark       lawyer

well...aids does not point out the specific problem
because there are hundreds of ways to program that
type of routine...perhaps it would be nice if the
AIDS write-up warned new authors a little more explicitly
about problems with -comload- (-stoload- also).

---------- response 2
  03/13/76     16.27     rader       s

It would appear that a note to a consultant (pso) would
have been useful to you.

---------- response 3
  03/13/76     18.15     mike b       cornell

         This seems like a goal place to give "psonotes"
a plug. I put in a request list night after 2100 CST
and had a response within a half hour or so. It would
seem that many programming problems would get faster
service by writing a psonote, rather than a Public Note.

M¬ ike≤≤≤≤ˇB

---------- response 4
  03/14/76     00.15     gilpin       peer

However, that is not an intenced use of "psonotes", and it
has the disadvantage that users at large do not have the
opportunity to learn from the discussion. Bill Golden





should give the definitive reply here, but perhaps the
practice of giving the title "help" or "pso" to all Public
Notes requesting help with programing problems would permit/
encourage rapid response in Public Notes.

---------- note 339 char error
  03/13/76     14.45     wagle       iu

got the following:

        ERROR in CHARSET!

        lesson=prosser,block=sh
        error word=003171515171655550325

that octal word is 'common cu' which is also the first
word of block 1-b....i got the error after using the
lab option in the block creator, copying the charset
in block 2-f , and attempting to edit the block..

good luck...

---------- note 340 POLL!
  03/13/76     15.16     wagle       iu

********
POLL!  03/12/76     15.56     curly       iu

  Announcing a opinion poll in lesson 'curly'.
Polls will remain open for another 12 days or until
my common fills up(after 300+ people have entered).
  This is the same announcement I put in pad and gpad.
Wouldve put it in here too..cept i was so tired i
forgot all about public notes at the time.

curly≤≤≤≤≤¬↑ˇ↑ˆ↑ˆ↑ˆcurly
********

C'MON people!
only 71 people have voted..
it really doesnt take long at all..
we need YOUR opinion!!!

---------- response 1
  03/15/76     08.24     mont       csa

Poll on what?

---------- response 2
  03/15/76     10.20     judy       pso

As a matter of fact, your opinion is NOT needed, because
a decision has already been made.





---------- response 3
  03/15/76     13.19     curly       iu

Which is?.....

---------- note 343 loadamicro
  03/13/76     18.06     slayton       arizona

What ever happened to pressing DATA1 from the pnotes editor?
Didn't a person used to be able to load a micro table there?
I tried it and it doesn't word now...is this some new
feature that I am ignorant off?

$DS≤≤     DS≤≤     DS≤≤     DS≤≤     DS

---------- response 1
  03/14/76     08.40     s zweig       iusn

         I'm just guessing here, but I'll bet that since
you can now jump around between public notes and personal
notes, that some folks were using 'animicro' and the like
for fancy signatures in public notes. Since this is frowned
upon, the feature is now gone. (Psst...am I right??!?)
Ʃteve

---------- response 2
  03/15/76     00.09     k mast       p

The personal notes editor was recently combined with the
group notes editor to make this easier to change. As a
result, the special option to load micros was removed.
Note however, that this option was never officially
announced, hence can be change) or deleted without notice.
If someone can justify to David Woolley and myself why it
should be put back in (for other purposes than animation!)
than such a feature will considered.

---------- note 346 m and p?
  03/13/76     18.18     clark       lawyer

what significance does the """ or "m" command have
in both the p-notes and groupnotes editors?

I noticed that they replot the screen but don't seem
to do anything...

(also, what does the "P" command do...)

---------- response 1
  03/13/76     19.51     woolley       p

p = +  (forward one line)
m = -  (back one line)





---------- note 355  augean
  03/14/76     15.27     donchin       psych

Recently i was informed on attumpting to read my notes
tht the course (psych) was full.It turned out that
the system would not allow me to read notes under these
circumstances.
The carch is that if I cant read my notes i cant
delete them either.
Seems catch22ish.How is one supposed to clean up
the locked stable?
Many Donchin.

---------- response 1
  03/14/76     20.26     woolley       p

I've tried filling a notes file completely, and everything
still seems to work. Can you give us some more details?

---------- note 357 Copy?
  03/14/76     21.43     cole       langlab

    The copy command doesn't seem to work when I
use the following code. What am I doing wrong?

1rest
at         410
write      Xˇ1,Yˇ1=
           Xˇ2,Yˇ2=
           Xˇ3.Yˇ3=
           Xˇ4,Yˇ4=
arrow      417
copy       n21,20
storen     n21
storen     n22
ok
endarrow
erase
branch     1rest

Dov Gole

---------- response 1
  03/14/76     22.01     blomme       s

   Copy and edit work on alphanumeric strings, not on
numbers (storen).

---------- response 2
  03/14/76     22.02     bateman       a

1
arrow      xxxx
copy       n1,20
storea     n1,20
storen     n21
storen     n22





ok
endarrow
erase
branch  1

the problem was that the copy command expects the variables
to contain alpha characters and the 'storen' command
puts a numeric representation of the number in the variables.

---------- note 363 okextra?
  03/15/76     00.24     snellen       medneta

The following code works ok if the answers given are
'real' and 'virtual', but 'real image' or something
similiar (which also would be a legitimate response) is
entered then the answer is treated as if it is incorrect.
*
at         908
write      1
           2
arrow      910
storea     ¬a
ok
arrow      1110
storea     ¬b
ok
writec     ¬a='real' $and$ ¬b='virtual' $or$ ¬a='virtual' $and$ ¬b='real',right,wrong
*
Is there some way to produce the effect of a specs okextra
in this code?

--JES

---------- response 1
  03/15/76     00.34     luke       cornell

    How about using the neat new -getword- command?
Or since you are not in judge state, why not try a
-find-, or a -findall-? You can do many searches,
if all you want to find is the occurence of something,
rather than a non-occurance.

Luke Kaven

---------- response 2
  03/15/76     00.35     luke       cornell

    My good god, then you could throw away that
redundant extra arrow!

Luke

---------- response 3
  03/15/76     01.28     snellen       medneta

The trouble is that, according to aids, -getword- requires
that one predict where, in the response, the correct word





is located; -find- may be a better choice, but I haven't
quite been able to figure out how to apply it.
Also, I'm curious to know, why do you consider the second
arrow to be uncessary?

--JES

---------- response 4
  03/15/76     09.41     warner       iu

        Why don't you use the numerical -concept- options?
Set up a vocabs with "real" and "virtual" attached to the
same concept but given two numbers. Then you can do a
specs okxvocar
concept real,n1‎real,virtual,n2‎virtual
and test n1 and n2 in any way you like!

---------- response 5
  03/15/76     10.29     judy       pso

Silas has a nice idea with the "concept", but if you don't
want to bother with a -vocabs- for such a short question...

unit         try
arrow        1010
answer       real <image>
answer       virtual <image>
endarrow
calc         ansr1‎anscnt
arrow        1210
specs
judge       anscnt=ansr1,nc,x
writec      anscnt=ansr1¬,You already have that answer.¬,¬,
answer      real <image>
answer      virtual <image>

---------- response 6
  03/15/76     10.32     snellen       medneta

Hmmmm, maybe that _is_ the way to do it even though it seems
awfully inefficient to have to create a vocab block if all
you want to do is handle a single question---still, this
might be a good time to learn how the -concept- technique
works. Thanks for your suggestion.
 --JES
P.S. I'll be back if I run into trouble.

---------- response 7
  03/15/76     10.55     snellen       medneta

Cops! You beat me to the punch on that last response, judy.
I'll try your idea too; thanks. --JES.

---------- response 8
  03/15/76     11.22     golden       s

You do not need a vccabs block to have a vocabulary. Just
add it to the rest of the code.





---------- response 9
  03/15/76     21.29     judy       pso

Just in case you decide to try the -vocabs- technique,
Darlene points out that Silas' example is not quite right.

define  value=n1
vocabs  mine
        (real=1,virtual=2 )
        <image>
***
arrow    1212
specs
writec   value,,,you chose real,you chose virtual
concept  real image,value‎real

---------- response 10
  03/16/76     11.53     snellen       medneta

Ok. It'll be a day or two before I have time to give it a
try, but I will certainly test it to see how it works.
My thanks to everyone for their help.
--JES

---------- note 367 beginnotes
  03/15/76     04.22     bonnie       matha

One "group" that might benefit from having it's own
groupnotes file is beginning authors.

For them, many public notes contain information that they
don't yet understand and don't want to wade through.

Many new authors may not write a question in public notes
because they know that the greater part of the public
already knows the answer.

A "beginners" notesfile (that authors could both read
and write) could be a place where new authors write their
questions and where they can read other new authors'
questions and responses. PSO people could maintain this
file, write notes of special interest to new authors, and
also transfer from Public Notes any notes that would be
of interest to new authors.

---------- response 1
  03/15/76     04.22     k mast       p

Good idea" I was just about to suggest the same thing
myself. For new authors, beginner notes would temporarily
replace Public Notes, hence getting them into the habit of
reading notes everyday.
        Can we have a PSO response please?
        Can we have a new author response please?

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





  03/15/76     09.01     luke       cornell

       Excellent!

       I'd much rather know where I can find the 'help'
notes, than have to sift through all the public notes
to see if they are a new author's request for aid.

       Although I like the new notes features, I would
like to see 'help' notes the way they were. (My only
hesitation, however, is that possibly urgent requests
for aid will go unanswered by the authors who have
the necessary experience. Perhaps 'psonotes' could
be used for this, then?)

Luke Kaven

---------- response 3
  03/15/76     09.20     railing       iu

  The idea is a good one. Even if you are not a new
author, there are still many things that some people
take foregranted that we still don't know, most likely
because we have never had need for them, My questions
are usually answered either by other authors, consultants,
or even AIDS, but a notefile of this type would be a
lot of help to new people. I learned a lot about TUTOR
from the old HELP NOTES, and even though I may not need
much of what I read in PUBLIC NOTES, I can now understand
most of it. I also agree on getting new authors into
the habbit of reading the notefiles every day, otherwise
they get bogged down in just their own areas.

Malkin

---------- response 4
  03/15/76     10.50     snellen       medneta

Prior to last fall I had never _seen_ a large computer much
less tried to write a program for one so that probably
qualifies me as a new author; virtually everything I've
learned about tutor I've gotten from Sherwood's book or
have picked up in bits and pieces from other authors.
I think the beginner notefile is much needed and would be
deeply appreciated by new authors. Not only would it spare
the necessity of coming into public notes as the villiage
idiot, it would probably increase the chances of the
response to a question being comprehensible.
As was mentioned in one of the earlier responses, much of
public notes is meaningless to the new author.
--JES

---------- response 5
  03/15/76     11.15     golden       s

I am mildly opposed because I would expect the proposed
notes file to evolve along one of two routes:





1)  It would become just what help notes was with users
rather unclear about what should go into which file;
or
2)  It would be a separate notes file and some authors
would never feel right about moving to Public Notes;
therefore, questions of greater and greater complexity
would begin to appear.

In either case, authors and CERL staff would still be
forced to read all notes to be well informed.

It seems better to me to work to overcome the reluctance
of new authors to use Public Notes. When it comes to using
an interactive computer system for instruction, everyone is
a beginner! We shall continue to learn from. each other for
a long time. I favor the unified Public Notes.

---------- response 6
  03/15/76     11.35     michael       english

I tend to disagree, Bill. New authors, especially very new
authors, tend to have a large number of quite simple prob-
lems related solely to their lack of familiatity with PLATO
conventions. "My lesson says INSPECT ONLY. Why won't it
let me edit?! "I accidentally deleted a block. HELP!!!!
I doubt most experienced authors benefit by wading through
these types of notes; mode importantly, I think new authors
are afraid to bring up such problems in PUBLIC NOTES, and I
can't see how this situation will improve.

Personally, I think the major problem with old HELP NOTES
was that it was never defined exactly what should be in it,
despite the HELP page description. A G-NOTE "novicenote"
with a descriptive title like "Novice Programmer Notes"
could potentially improve things.

---------- response 7
  03/15/76     12.36     lawrance       medokla1

        Oh, a novice note file would be so nice. We neonates
are quite reluctant to call up operators just to ask if
operators mind being called up. Trivial programming problems
can stump us for hours. IT would be pleasant to have a place
to congregate and just get familiar with the language before
stepping out into the traffic with all you guys.

---------- response 8
  03/15/76     14.40     golden       s

Don't let a programming problem "stump you for hours"!
Call a consultant; or, if you cannot reach one, leave a
public note. If that doesn't work, leave a psonote. Call
an operator, or another author, or a systems programmer.
There is always some help available.

---------- response 9
  03/15/76     20.02     ld francis       mtc





i too favor novice notes----the idea is certainly
cheap to try and should be equally easy to evaluate.
it seems that the famous "try it and see" philosophy
plato is know for should be applied here.

---------- response 10
  03/15/76     23.09     kent       unl

        _I_ haven't been very bashful in the past about using
public notes for anything and everything that's bugged me
and what happens? I get a _very_ critical pnote from a rather
well-known palto author assigating me for the naivete I
displayed. Now I'm not much of one to be intimidated and
I doubt that this sort of thing happens very often, but once
is too much for many authors; so I very much agree with the
idea.

---------- response 11
  03/16/76     12.05     snellin       medneta

Believe it or not, I once went into the old HELP notes to
ask what the purpose of the "S" (security code) option was
on the author mode page---obviously it must be a useless
option since one can change their password on the password
entry page---and whether you call it a "password" or a
"security code" it's all the same thing--right?

Seems I got a pnote (aleeit a courteous one) suggesting that
HELP notes was not the appropriate place for this sort of
question.
--JES

---------- response 12
  03/16/76     13.00     nelson       ced

It only takes one snide response to a dumb question to
inhibit the askee from asking any more questions for
quite some time.

I think the notefile for beginning authors is a great idea,
and it seems more in the Plato trandition to try it out for
a while than to think up resons why it won't work. Us
folks trying to convince skeptics that programmed
instruction is a worthwhile thing have already heard
more than enough reasons why something won't work.

Furthermode, I don't think tht people had all that much
trouble differenriating help notes from general notes in
the old style files. 95; if the notes were categorized
reasonably, and 5: disageement in qualitative analysis
is actually a respectable figure. When you are dealing
with humans instead of machines you got to make some
allowances.

---------- note 374 instructor
  03/15/76     10.00     errol       kka





One of our instructor was unable to use his
instructor signon to see the course curriculum.
Plato kept taking him in circles.
Details will ne given upon request to systems person.
Please page me or send pnote to which I can reply.

---------- response 1
  03/15/76     11.20     b sherwood       s

Fixed.

---------- note 375 cdc-cerl
  03/15/76     10.10     golde       s

A series of agreements covering PLATO hardware patents,
PLATO software, and courseware produced at the University
of Illinois have been signed by the University and Control
Data Corporation.

Details of the courseware agreement, which is the one
that most directly involves authors, will be available soon.
Institutions other than the University of Illinois may
choose to include themselves in that agreement.

With the signing of these agreements, CDC is expected to
increase their already substantial efforts to make a
commercial success of PLATO, and CERL will continue the
development of new and more sophisticated PLATO systems.

---------- response 1
  03/15/76     14.44     hody       med

congratulations-- that is perhaps the most important
note in recent times...

*
might you want to make it more visible as an inclusion
in system feature notes

---------- response 2
  03/15/76     17.05     bowery       comm

Well, it has little information in and of itself except
to say that there will be more information fourth coming.

Now THAT information should probably have a LESSON devoted
to it.

---------- note 384 exam/stat
  03/15/76     11.34     eyl       mtc

   Can anyone tell me of there are any lessons on the
system which deal with statistical analysis of exams?
I am interested primarily in utility lessons which would
give measures of test reliability and item analysis
including discrimination and difficulty indices.





Thanks in advance,
A.W. Eyl

---------- response 1
  03/15/76     15.41     jmk       pso

lesson "mcquiz" provides means of writing exams and some
analysis of results, including item difficulty and
discrimination.

---------- note 386 disk space
  03/15/76     11.49     error       kka

I've been trying several times this morning to extend
a couse by 1 part but have received the message
"insufficiant disk space available."
Can anyone explain this and offer a suggestion?  Thanks.

---------- response 1
  03/15/76     12.48     j widson       unidel

When this happened to me, Larry White explained it this
way: To extend a course, the routine looks for space
on the pack that has the most space. If it cannot find
enough contiguous parts there to copy the file into plus
what you want to add, you get the message. Either keep
trying of find a friendly systems programmer or operator
to help. It seemed even stranger to me initially--I
got the message trying to SHORTEN one.

---------- response 2
  03/15/76     12.53     white       p

That message means that Plato was unable to find a
contiguous area of disk space large enough to hold
your longer file. I would like to encourage those
account directors that are creating empty files, in
order to have space later when they need it, to
delete all those files. The ability to create and
destroy disk files was given to account directors
to make it easier for them to do their work, not for
them to hurt the Plato system by creating extra files.

Work is in progress on the system level to lessen this
problem, but till then wanted to create as few files
as we want other account directors to create.

-------------------------------------- note 390  nrecords
  03/15/76     13.12     meers       wright

        It appears that the 'type in your course code'
page of the new records editor works on a pause, and it
will not accept any of the following characters:

< >. [ ] $ : _ ' * ( ) ; : , " . ! / ? + Ʃ - Δ ≡ ↑≡ ≠ ↑≠
      OR backspace.





As a result, we can't set into some of our courses.

---------- response 1
  03/15/76     13.25     curly       iu

  While i agree that that shouldnt be so...you CAN get
into your course by entering the proper code via the
'S' option on the author mode page. Also....when
is -DATA1- to enter change code gonna work if one enters
in inspect mode? And third...I would still prefer the
'See who's running option' on the first page.

curly≤≤≤≤≤¬↑ˇ↑ˆ↑ˆ↑ˆcurly

---------- response 2
  03/15/76     14.36     stone       politics

First, they need to get rid of the multitude of other
errors. I think that they should bring back the old
course editor until thus one is truely finished and
thorougly tested, i would be glad to help test it,
as I have found several errors(>10) errors already.
A couple of which are very serious.

Scott Stone

---------- response 3
  03/15/76     17.47     b sherwood       s

Finally understood the problem with non-alphanumeric keys
in the security-code typeing. Someone had tried to point
it out to me, but I hadn't understood. Will fix. The
problem came about thru an attempt to unify security-code
typeins with the "set signin password" option. The
password option allows one letters and numbers, not
punctuation marks, to ease the problems students have in
trying to sign in (other says are thrown away). This is
indeed not appropriate for file security codes -- will fix.
As was suggested, use the (shift)-S option for now.

I do hope Stone will write me a p-note and tell me what
the majot errors are! Hard to correct them otherwise......

---------- response 4
  03/15/76     17.47     b sherwood       s

Oh yes -- DATA1. That will be cleaned up shortly, too

---------- note 392  personotes
  03/15/76     13.19     p cohn n       med

        After reading personal notes, can a "BACK to
reread your notes" be implemented? 'Twould save one
key press, two jumpouts, and three seconds of aggrava-
tion. Thanks.





               --paul

---------- response 1
  03/15/76     15.45     jmk       pso

Maybe just a way to keep from falling off the end of your
p-notes after deleting the list one!

---------- response 2
  03/15/76     15.56     k mast       p

Both of these suggestions are useful, will try to get it
changed.

---------- note 394 why ECS
  03/15/76     13.43     straayer       telcomc

      When is the UI Ui bana spring break?

      Why is site network blessed with quantities of ECS
today, while usually operating at 90-100: from 9 to 5?

      Are the above 2 questions related?

---------- response 1
  03/15/76     13.44     fisher       conn

Yes... spring break is now in, and the ecs is because
the U of I is not ( i would assume ) crowded...

    fishswine

---------- note 396 label/calc
  03/15/76     13.44     buz       phar

just noticed that a branch label initiates a calc seq. is
an unannounced feature, or anounced so long ago that i hadnt
heard of it?
ie.
*****
do       abc
calc     n‎5
branch   a<b,1out,1in,1error
1in
goto     dummy
1out     n‎6 $$‎‎****** this line works as if it was
do       def      ***** 1out
more tutor          *** calc n‎6

---------- response 1
  03/15/76     13.55     fisher       conn

         Looking in AIDS (DATA¬Dlabels) I find that it
has two lines of code saying:





1start
        n1‎n1+1

In answer to your question.. It was announced a long time
ago.. I have been using it almost since I came on the
system....

      fishwine

---------- response 2
  03/15/76     14.09     luke       cornell

        Man, I have been using that since I got my records
(one of the first things I learned)

        It's all over aids, I believe, go read about it!

        Luke

---------- response 3
  03/15/76     14.14     luke       cornell

    Yes, I just checked, its there for sure. Look under
'lables' from the "What TOOTER feature " page. It's a
couple of NEXT's deep in there.

    Luke

---------- response 4
  03/15/76     16.46     buz       phar

it says a line with a label may _contain_ a calc...i know this

what wasnt clear was the fact that a label STARTS a calc,
also...ive read over that segment on branches several times,
and it only seemed to imply that if you were in a calc seq-
uence, with labels in that sequence, any of those labels
could contain a continued calc...not that a label in the
middle of the netherlands would BEGIN a calc.

---------- response 5
  03/16/76     10.27     michael       matha

WHY ISN'T THE FACT THAT A "LABEL" _INITIATES_ A CALC
INCLUDED IN "AIDS"? A SIMPLE TEST UNIT WILL SHOW YOU THAT
IT _DOES_. MOST PEOPLE DON'T SPEND THEIR TIME LOOKING FOR
FEATURES THAT ARE UNANNOUNCED OR UNDOCUMENTED, AND _THIS_
FEATURE IS OF THE KIND THAT COULD SAVE MUCH REDUNDANT CODE.

---------- response 6
  03/16/76     11.34     mike b       cornell

         Sir.





        There's no need to shout. If you can't hear the
discussion, turn up the volume on your terminal. Capital-
ized notes:

        1.  are hard to read

        2.  make printouts of notes look very messy,

        3.  rattle the flowers on the table next to my
            terminal.

            Thank you.

---------- response 7
  03/16/76     12.28     michael       matha

I am usually a soft spoken person, but when I discover that
all along I have been writing code like the following, in
order to initiate a calc, I tend to get upset.

calc
1label some calc, branch or doto statement

How much nicer it would have been, if authors could have
been informed of this feature, instead of accidentally (or,
deliberately) stumbling upon it, or hearing of it by word of
mouth. Just because "It's been that way for years", it
does not mean that everybody knows about it.

I apologize for losing it over such a small detail, too.
Michael

---------- response 8
  03/16/76     14.20     railing       iu

   As I remember, it was the -branch-, not the label that
started the -calc-, not both of them, But if you remember
that -branch- is a -calc- type command and the lable is
part of the -branch-, it all makes sense. I agree that
it saves a lo of code and is a nice (announced) feature.

   Malkin

---------- response 9
  03/16/76     19.45     blomme       s

   Frankly, I think that code which looks like
calc
1label
   is nice and quite readable and I would be loathe
to substitute
1label
   for it. Off hand, my guess is that the former costs
you one word of disk space and no words or time when
executed.

---------- note 404 psonotes
  03/15/76     15.05     celia       pso





In response to the suggesting about a beginners' note file:

After some discussion, the consultants decided we don't see
the advantage. New authors should definitely contact us
with their problems, either consult, or in psonotes or
public notes.

The kinds of things appropriate to psonotes:
1. complaints, suggestions, etc about aids or pso services
2. things that concern system security
3. specific problems in lessons whose names you want kept
   private
4. things you're embarrassed to ask in public notes

In general, if you don't know whether to put it in psonotes,
go ahead and put it in.

        jmk, darlene, judy, tina, dave, celia, elaine, bill

---------- response 1
  03/16/76     12.58     snellen       medneta

For what it is worth---on more than one occasion I have
found pso to be like unto annoasis in a parched desert.
      --JES

---------- response 2
  03/16/76     13.10     hinton       ssu

"pso" is helpful to people who have been around a while,
too !
        ndh

---------- response 3
  03/16/76     16.15     rowlen       mflu

         And at 3:30 in the morning there aren't very
         many consultants on the system, not to mention
         times like 6:30 or so in the evening (there are
         at times, but not when I run across a goody).

         Jim

---------- note 408 pack ack!
  03/15/76     16.07     david       cerlcc

It would be nice if -pack- and -packc- both took
the location and length variables in the same order.

Or is there a reason why:

pack length,string,warts

     ≤≤≤but





packc freak,string,length,warts,moles,corns

---------- response 1
  03/15/76     16.15     warner       iu

        The -pack- command is the exception to the rule
just about everywhere else....in -search-,-compute-,
etc. the string comes first, followed by the length.
        Someday the -pack- command's variables will
be reversed, and a system conversion run.
         Then, of course, the curmudgeons will come out
and demand to know why their memories are being unfairly
taxed by this seemingly pointless reversal...

---------- response 2
  03/15/76     17.23     blomme       s

  Yes, I think it has been agreed (and mentioned before)
that the pack command should and presumably will be changed.

---------- response 3
  03/15/76     20.49     david       cerlcc

thanks...I must have missed me discussion.

---------- response 4
  03/15/76     21.43     judy       pso

I think the -pack- was originally made reversed, because
of the (expected) tendency of people to say

pack n1,n2 $$ putting the variables in numerical order

If you had
pack string,length
then for strings of more than one word, the length (n2)
would overwrite the string (n1).

It _is_ confusing to have it backwards from everything else,
and besides, most people define their variables, so that
one doesn't see the n1,n2 anyhow. So someday perhaps
we will change,

By the way.... don't try to remember all these order
conventions. It is practically painless to look them up,
and who can remember 200-odd tag formats??

¬[ I do assume that you will remember -do- and -writec-!¬]

---------- response 5
  03/16/76     18.18     dave       infe

        Which reminds me that if you want to -pack- a string
containing both a comma and a semicolon, somebody has to
fix the condensor to accept ¬, as a delimiter...

---------- note 411 nrecords





  03/15/76     16.53     david       cerlcc

I (for one) would like to see that last site signed on at,
when inspecting an individual record.

Also, if someone is currently running, where.

---------- response 1
  03/15/76     17.35     fisher       conn

                   I would also like to see the last site
the records were used at to be on the I display (from
the author mode page)
Many times when I sign on, I miss that page, possible
either because I'm in a hurry or that I came in after
the system has been down, and that line is overwritten
immediately....

               fishswine

---------- note 417 term talk3
  03/15/76     19.22     fisher       conn


********
term talk2  03/05/76     19.46     fisher       conn
********
termtalk    03/04/76     11.26     berger       mflu

When I erase a line in TERM talk in monitor mode,
I lose 5-10 characters of the next line I type.
This was not the case a few weeks ago.
********
This gets very aggravating. each time you hit next,
you cannot type for about two seconds.....

if possible.. could this be 'fixed' or something
done to speed it up?

     a slowed down sorta fast typist...
     fishswine
********
     At least get some response as to it has
been read by systems and is being worked on??????
     fishswine

---------- response 1
  03/15/76     19.44     blomme       s

     The notes have been read and if we find anything in
that area that could be contributing to this, it will
be fixed.
     On another track, I find this game of copying other notes
irritating. In public notes please say what you have to say
once (and as briefly as possible); if there has been no
system response to a matter you consider of prime importance
and in need of immediate attention, then point out the note





--I'd guess that several system staff members read every
note, but it is often hard to say anything much more than
"Oh, really?" or "Note read" or something equally inane
to some of the notes.

---------- response 2
  03/15/76     19.51     frankel       p

I know what the problem is, but I can't fix it until I
find out who broke it. Soon, hopefully!

---------- response 3
  03/15/76     19.57     fisher       conn

                    Thank you Mr. Frankel, for telling us that
it is being worked on, and I'm sorry if I aggravated
you, Mr. Blomme.

     fishswine

---------- response 4
  03/16/76     19.46     blomme       s

    We've tried one "fix" but it didn't seem to have
much effect.

---------- note 418 no pointer
  03/15/76     19.36     bowery       comm

How many times have you seen code like this:

calc    pnt‎0
compute x,xchar,xlen,pnt
calc    pnt‎0
compute y,ychar,ylen,pnt
calc    pnt‎0
*
* etc.. ?

Could a -compute- command with to pointer tag just be
assumed to be doing a "one time only" compilation and
avoid all the stuff associated with repeated execution?

---------- response 1
  03/15/76     20.52     david       cerlcc

just out of curiosity...

does anyone have records on how often, if ever, the
buffer for -compute- compiled code fills up?

---------- response 2
  03/15/76     22.05     judy       pso





1) you are allowed to keep more than one pointer, so you can
   switch from one -compute- to another without constantly
   resetting the pointers.

2) I do not know how often the -compute- buffer fills up.

   The string which is stored in "pointer" -- and which
   has absolutely no meaning for you or for me -- contains
   information about where in mermory the compiled code is
   kept and some "checking" bits which must match up with
   a portion of the compiled code: itself. Therefore if the
   compiled code is over-written, the -compute- will
   automatically recompile itself, even though "pointer"
   was not zeroed by the program.

---------- response 3
  03/16/76     09.47     b sherwood	      s

It is sufficiently difficult to explain the need for and
the use of the pointer for repented executions. Requiring
the pointer puts a tiny extra birden on those people who
really don't intend to re-execite the code, but keeps
things crystal clear for those who might. It seems to me
that if the 4th argument were optional many people would
not use it when they really did need it. The present
format is designed to be educational, in that it forces a
certain amount of attention to be paid to how the thing
actually works.

In other words, it takes a certain amount of sophistication
to use a -compute-, and even more to distinguish those
cases where the pointer is superfluous.

---------- response 4
  03/16/76     10.13     david       cerlcc

It just seems like unnecessary effort for me to make
the system send my compiled code off to be stored in
ECS, when I know I'll never need it again.

---------- response 5
  03/16/76     15.30     sellers       arizona

How about allowing the pointer in the -compute- command be a
constant 0?

compute a,b,c,o

This would be the one time case that bowery wants.
As it is now one gets a condense error on the above form.

---------- note 420 voc.ed.
  03/15/76     19.46     boas       unidel

	I am interested in filling lessons and curriculum





materials on floriculture and inamental horticulture.
       I have gone through ics, library, and catalog
and had no success.
       Can anyone out there help me out with information
on who I might contact or whan material is on the system?
			Thanks
			    Ev

----------response 1
  03/16/76     08.17     luke	      cornell

	Contact tebby / pso via p-note. She is
really nice, and will be happy to tell you anything
about lesson areas. curriculums, authors, projects,
etc.

Luke Kaven

---------- note 423 ignore+ops
  03/15/76     19.58     blomme       s

    A slight change has been make in the handling of
a helpop after a judge ignore. It has been the case that
althought the last writing on the screen is NOT erased when
a judge ignore is done, that writing IS erased when a
helpop branch is then taken. This has been changed so
that the helpop branch no long causes this erase to
occur. This seems more logical (in some sense), since
after the execution of a judge ignore, entry of answer
is begun afresh and so there should be no writing on the
screen to be associated with the currnet answer. If this
will cause anyone problems, please let us know. We can
change this back if there is something that has been missed
--but the new way seems more convenient for many uses.
   A related issue is whether or not the judge ignore
actually should erase the last answer-related writing.
but this has not been changed at this 
---------- note 436 mathnotes
  03/16/76     11.41     dibello       cerlcc

A new groupnotes file called 7mathnotes' has been set up...
it is for anyone concerned with the PLATO mathematics
materials for Community Colles and Adult Education....
any comments regarding design operation of individual
lessons, implementation of the ourseware or matters of
information are appropriate for 'mathnotes'.....
access is unrestricted........Lou DiBell
Coordinator, Community
College Math Project

---------- note 440 helpflag
  03/16/76     12.51     snellen       medneta

There seems to be a lot of interest in reviving some sort of
distinction between requests for help and "other" notes.
Would something like the following be workable (and useful)?





After writing a note for the public notes file and entering
a keyword, the writer is asked:
Is this a request for help?
The writer then responds with ..y" or "n"; if the response is
"y" then an -D or * is displayed next to the keyword on the
note selection page.
At least this way one could spot the "help" notes before
pageing through everything.

--JES

---------- response 1
  03/16/76     13.46     luke       cornell

       This is an idea, but I believe that most of
the objections that have come from "older" authors
were that they had to sift through all the notes
as they were going through them, to determine
whether or not the note was a request for help.
       To have this flag as you suggest, the author
would either have to "remember" the number of the
note which was undesireable to him, or he would have
to look for it (if put there) inside the body of
the note itself; but then he has already been
forced to read the note somewhat, because he is
still looking for this flag. (another bit to set)
       I believe that things can safely enough be
left the way they are.

       Although always a one for change, I will
go along with pso's decision not to furthur
specialze our public notes-file system

luke Kaven

---------- response 2
  03/16/76     15.45     jan       park

instead of a flag set by the notes file ...why not just
label the notes yourself as you put them in?
then the writer of a note can signal to the rest of the
notes readers that his note, at least, is a request for
help.
start a convention of preceeding all the help-type
requests with a certain letter, followed with a descriptive
word

we can do this ourselves, with a little bit of effort in
the area of clear, concise communication with our plato
collegues

---------- response 3
  03/16/76     16.02     lh nawrocki       mtc

jan/park's idea sounds workable and simple. Why not
just state "HELP" or "ASSIST" at beginning (don't need
anymore new abbreviations)?





---------- note 444 Δ sys var
  03/16/76     13.15     struven t       csstaff

Does the new mark-up (Δ) have a system defined
variable, or does it use 'orden'?

arrow	1708
force	micro $$ cslibrary, fortman; uses only fortran chars
specs	bumpshift
putd 	*,* X *
answer	read X id X iscore
no
writec	order=-1,,+you have left something out.
endarrow

The above code judges for the response of 'READ, ID,ISCORE'.
If only READ,ISCORE is entered, the markup looks like
READ,≤↑ˇΔ↑ˆISCORE and the message 'You have left something out.'
is printed. But if READ,,iD is entered. the markup looks
like READ,IDˇΔ....but no message is printed! If there is
no specific variable now, is there one planned?

---------- response 1
  03/16/76     13.38     michael       english

Use system variable *entire* to test for Δ markup.

---------- response 2
  03/16/76     13.56  struven t       csstaff

Thanks. I missed it when I ment through AIDS.

---------- response 3
  03/16/76     14.24     darlene       pso

The new markup material will be in AIDS soon.
darlene

---------- note 445 eyeproblem
  03/16/76     13.21     mike	b       cornell

there is a fellow here at Cornell who would like
to author lessons for PLATO, but he is visually handicapped.
He cannot read the plasma panel without magnification, but
simple lenses will not solve his problem. He needs a
magnified _real_ image, not a _virtual_ image (which is all
a simple lens will provide).

Has there been any work done with magnifying the
image of the panel and projecting it on a screen? If so,
is the necessary equipment commercially available? Any
help would be greatly appreciated.

M¬ ike≤≤≤≤ˇB

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





  03/16/76     14.29     b sherwood       s

You can photograph the screen with a TV camera, then
show the image on a television screen. Use a zoom lens
to enable the magnification. you need a fast lens (f2 or
better). Since most cameras are not very sensitive in
the red, you will probably want to amplify the signal
coming out of the camera, then use the contrast control
on the monitor to make a high-contrast display. The
resolution is not high enough to see the whole plasma
panel display well. but with the zoom you can look at
sections of the panel.

Of course you can also use projection video to put the
image on a large screen.

---------- response 2
  03/16/76     15.59     lh nawrocki       mtc

Gee, don't know the technology, but lately have been
seeing ads for screen projection devices to attach
to TV sets (and TV/projection combination sets) which
increases viewing size by a fautor of 3 or more. Might
do the job, but haven't checkd out such gadgets myself.
Anybody else know mcre about this?

----------response 3
  03/17/76     08.02     maggs       law

Make you own text↑ˆ↑ˆ↑ˆ↑ˆ↑ˆ↑ˆ↑ˆ↑ˆ↑ˆ↑ˆ<7↑ˇ↑ˇ↑ˇ↑ˇ↑ˇ↑ˇ↑ˇ↑ˇ↑ˇ↑ˇ≤

editor which sends out a unique signal
for each character that the user types, out
through an external cutlet on the terminal either to an
audio device or some sort of black box beeper. Thus
everything typed in could be echoed in morse code for
instance. This would give the necessary feedback. A
second service program could head programs (after they
were converted to common) ove the same type of audio
device.

---------- response 4
  03/17/76     11.41     judy       pso

John Risken (john risken/cdoc) and his cronies (perhaps Bob
Yeager (bob y/reading) and John Eisenberg (david/hebrew),
I'm not sure who) already have a Morse code device which
takes a typed-in string and makes the proper beeps.

If you are going to go to the bother of a special text
editor, it would be easy (relative to all the other work)
to make a displayer which would display the stuff in
big letters (!typeface" characters, for example).

How much magnification does me need? Can he read -size-d
writing, or does it need to have the wider "strokes" which
are not provided by -size-?

By the way, you do not need an expensive T camera, a very





simple one will probably work best. You can probably check
one out from your "instructo ,resource center! (every campus
seems to have one) and try it out.

---------- response 5
  03/17/76    14.19     mike b       cornell

Thanks to all for the suggestions.

The trouble with beepers is that you have to know
morse.	TV is a possibility, but the set up of terminals
here would make this technique difficult in practice.

We are investigating the possibility of an editor
which uses sized writing. There are obvious difficulties
with this solution. however (betting condensable code is
awkward, for one thing; the EDIT key won't work for
another).

The best solution would still seem to be a magni-
fying device that projects an image onto a screen (thus
the image would be real). I suppose the plasma panel may
not be bright enough for this to be practical, however.

			M¬ ike≤≤≤≤ˇB

---------- response 6
  03/18/76     00.40     daleske       amesrad

Another problem is the amount of information that is
passable at morse speeds and that which can be passed
easily enough for a person to make a "picture" of the
screen in his/her mind...

John

---------- note 447  newcourses
  03/16/76     13.46     errol       kka

Is there any way to change a name in the roster of
a course from the page which lists the names
I didn't notice this feature in the new course set-up
on the page which shows all the students. It would
on very helpful to have this and other options
(e.g. deleting a name), on the same page so you can
see what you are trying to d.

---------- response 1
  03/16/76     13.53     luke       cornell

I would also like the option to see a record
from the roster display page by pressing the number
from there.
	The new formatting is pleasant, though, I
like the whole thing. (especially the new author
options!)
	
Luke	Kaven





---------- response 2
  03/16/76     23.31     stone       pfw

Changing an authors name was intentionally left out
of the editor, the options at the roster page were
left out mostly because the options differ for
most individuals.

Scott Stone

---------- note 454 data1, pls
  03/16/76     14.34     friedman       csa

On the lesson data page, is there any chance that the BACK1
key will return to being something else (preferably DATA1)?
If I had a nickel for every time I've hit DATA1 to enter a
security code in the last couple of weeks, I'd already be
rich. This is really an anjoying change, esepcially since
DATA1 has now been adopted (and rightfully so) by several
other editors, both system and non-system (e.g., datafile
inspection, course records,...).↓
PLEASE!

---------- response 1
  03/16/76     15.12     b sherwood       s

Records editor has now been converted, and text editor
is next. (And data editor, and...)

You will be able to press BACK, then next, and you will
see a page warning you that you are entering with
insepct privileges only, and inviting you to type a
change code.

To see the idea, inspect a course, then leave the course,
press NEXT, and you will see what we are working towards.
(There is one difference -- I make BACK1 be the exit from
courses rather than BACK, just because there are so many
menus in editing courses.)

---------- response 2
  03/16/76     16.51     friedman       csa

Thanks for the response, Bruce, but I don't quite see it.
Using new course records as an example doesn't help
completely, because no course has a blank inspect code.
I tried what you said, but i still don't see how it will
work when entering a lesson having a blank inspect code.↓
I'm bugged by this thing, though. All this change because
a new function had to added to the editor, and the only key
to put it on was DATA1? What bugs me is not so much the
need to retrain my fingers, fut the fact that assorted
non-system editors, including at least two of mine, have
adopted the up-to-now-common system usage of DATA1. We
now have to choose between being different than the system
editors, or reprogramming. 'Taint right!





---------- response 3
  03/16/76     17.41     b sherwood       s

Hold on....in a couple days you will see something nice.
which will help both novice and experienced programmers.

---------- response 4
  03/16/76     17.50     alan m       nuc

Can we have a clue as to what is something nice?

---------- response 5
  03/16/76     19.54     b sherwood       s

Yes, read the first response to this chain.

---------- note 455 election
  03/16/76     15.09     anderer       ee

    Since PLATO is being used to tabulate the election
stats tonight. is there any way we interested parties
can keep current on the results thru the system? In other
words, any way to see the latest data and projections?

---------- response 1
  03/16/76     16.25     frankel       p

Turn on your TV set.

---------- response 2
  03/16/76     16.30     anderer       ee

I had hoped for a useful response.

---------- response 3
  03/16/76     16.31     alan m       nuc

Can you please tell me what station it will be on in the
Chicago area?

---------- response 4
  03/16/76     16.49     golden       pso

To the best of my knowledge it will only be on three
central Illinois sttions and their repeater stations in
nearby cities.

---------- response 5
  03/16/76     17.49     alan m       nuc

Since it is unavailable thru TV up here, where can we locate





it on Plato?

I am not tryping to be obnoxious but I will be on or near
Plato for most of the evening, I do not own a transistor
radio, and I am interested in the election results.

---------- response 6
  03/16/76     17.56     k mast       p

I dont think it will be available on plato, as the lessons
written to display it are not secure enough to allow general
access. Sorry.

---------- response 7
  03/16/76     18.24     schramm       cs317

Will the telecast be from CERL? If so, will it be
possible for those of us who will be at cerl to drop by
the hookup site and get information on the election?
I'd hate to have to wait to read about it in the
NEWSREPORT.
J¬≤R ≤S

---------- response 8
  03/16/76     18.37     frankel       p

No. PLATO terminals are located at the three stations using
the system. At the moment, serving hundreds of thousands
of viewers is much more important than giving a few PLATO
users direct access. Maybe in November.

---------- response 9
  03/16/76     18.48     alan m       nuc

Can I request then that we have hourly updates, either dis-
played thru a note on the A.M. page or a note here, with no
replies, so it can be constantly updated.

---------- note 457 ¬$of notes
  03/16/76     15.23     stone       research

	are there any numbers (or even estimates) on the
number of notes written each day? also, how much disk space
is currently allocated to notes? i would like the
breakdown by notes type (p, public, group) if possible.
maureen stone

---------- response 1
  03/16/76     15.43     golden       s

Personal notes are written at the rate of about 1000 per day
and each is read or at least looked at 5 times. You can
count public notes as easily as I. I don't know any numbers
for groupnotes.

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





  03/16/76     16.26     woolley       p

We're getting about 225-230 group notes written per day.
This includes public notes and all other gnotes files.

There are approximately 90 group notes files (excluding
archives.) I don't know the average size.

---------- response 3
  03/16/76     16.56     k mast       p

There are currently slightly over 200 Personal Notes files,
I have no idea what the average size is, either. There have
been approximately 400,000 Personal Notes written since its
inception in August, 1974. The average pnote is about 30-35
words in length.

---------- response 4
  03/16/76     17.41     stone       research

          thank you all for the quick response.
                   maureen

---------- note 476 routerp
  03/16/76     17.57     dave       infe

         Why is it possible to turn off the 'router' flag
for a lesson (make a router a nonrouter) but not to turn it
on (except when the lesson is created)?
        Just wondering...

---------- note 484 records
  03/16/76     20.22     b sherwood       s

At the moment, you must press shift-BACK or shift-STOP
to exit from editing a course. I put in BACK1 with the
idea that there are so many menus in the course editor
that you might like to be able to hit BACK without going
all the way out. But maybe it would be nicer to have
a simple BACK to get out from the main index page. What
do you think?

On a related topic, I think it is true at the moment that
shift-STOP doesn't necessarily take you all the way out,
but rather takes you only part-way out from some parts
of the course editor. I will try to locate these spots
and fix them up.

---------- response 1
  03/16/76     21.35     nate       iumusic

   Frankly, I would rather shift-stop NOT take me out
totally from places where I don't feel like waiting
for certain things to finish (ie. the people-running
page and especially the statistics page). Anybody else
feel this way?





Nate

---------- response 2
  03/17/76     13.10     meers       wright

         As long as we can just press NEXT to get back into
a course, having BACK take you all the way out should be
no hassle..

---------- response 3
  03/17/76     17.11     fay       o

BACK1 is my choice.

---------- note 487 edit execu
  03/16/76     21/43     clark       lawyer

strange error....

execution error in lesson edit
unit replace3
last command copy (6th)

error occurs whenever you try to R>50 in an empty block...
of course it doesn't really matter....

---------- response 1
  03/16/76     21.51     a harkrader       medneta

Must've taken a while to find that bug...

---------- response 2
  03/17/76     15.39     clark       lawyer

not really...I was just showing a new author how if you
want to insert or replace the last line in the block, you
could just R2222 (or some other large number) and PLATO
would bring it down to the last line in the block....

*** end of notes ***