+++ plato iv group notes +++ Public Notes notes beginning feb 1, 1976 file pbnotes3 printed at 1:11 pm on june 12, 1976 ---------- note 19 logic cir 02/02/76 09.17 terry unl i am interested in obtaining information about any lessons on design and simulation of logical circuits. if your can help ---> pnote: terry of unl thanx ---------- response 1 02/02/76 09.39 want pfw answered via p-note ---------- response 2 02/02/76 12.47 friedman csa Check the lessons under "Logical Design" in lesson "cslessons". ---------- response 3 02/02/76 15.49 t obrien politics Try lesson 'spice' or lesson 'circuit'. These lesson names void where author prohibited. Terry O'Brien ---------- response 4 02/04/76 09.25 hadden css You might try logicff and (later) logicseq, but these are not completed yet. (Should be done this semester.) --geo ---------- note 22 condense q 02/02/76 09.56 fries ee Could someone explane why the ESC is wiped out when the system is brought down intensionaly, causing the long condense line (am I right there?). But when it crashes due to an error, everything is still in ECS after it is brought up. E↑ˇE ---------- response 1 02/02/76 10.16 golden s You are interpreting the data incorrectly. Whenever the system goes down, ECS copies of lessons are lost. But if the system which is then loaded is the same as the one which had been in, binary copies of most lessons will exist on a disk and delay is on the order of one second per lesson. If a new version is loaded. which is what usually happens during non-prime time, all lessons need to be recondensed. ---------- note 24 no create 02/02/76 10.04 woolley p Since the question has come up a few times already, I guess I should re-state it: Account directors can not yet create group notes files. They will be able to do so in a few days when things have settled down some. ---------- note 30 mast notes 02/02/76 11.10 schreiner csa I think the fine print in notesnotes notes (???) merits some attention: Where is information on Kim Mast's student notes notes available? I recall requests for uniform system-supported notes facilities for years... Since you are apparently now finally getting fed up with everybody's private notes facility, how about announcing (if not discussing...) the entire replacement?? I would really hate to see us jump (-out) into groupnotes only to find out that what we really always wanted to do was jump (-in???) to mastnotes. Or whatever! ---------- note 31 d option 02/02/76 11.15 schreiner csa in the logic department: How about making 'd' on the note index page let you read and write personal notes? ---------- response 1 02/02/76 12.19 oberpriller arizona right, it would then be consistent with options b and c. dave ---------- response 2 02/02/76 12.31 golden s Good idea. ---------- note 32 template 02/02/76 11.19 judy pso We have noticed that the "template" option of course records is not working properly. Until David and Ruth come in and check it, take CARE when creating student records. You will probably have to set the options for EACH student individually. ---------- response 1 02/02/76 12.50 nagel vu I saw that a while back and reported the problem but.......... As near as I could tell at the time it was only the curriculum template that didn't work properly.... the most importat one. ---------- response 2 02/02/76 13.48 b sherwood s I'm puzzled. Several people have observed this, but I just now tried templating curriculum options, and the new record correctly went to the expected module. Can someone send me a specific example as a pnote, please? ---------- response 3 02/02/76 15.00 b sherwood s Hmmmm.....David and I have found the behavior indeed erratic. The same operations in different courses produce different results! We will continue to search. In the meantime, if anyone finds a consistent pattern to the failure, please tell us. We assume at the moment that failures occur only with mrouter curriculum templating. ---------- note 33 c=pnotes 02/02/76 12.06 stone research how about switching the position of group notes and p-notes in the index. i am SO used to pressing N, c when i want to read my notes. maureen ---------- response 1 02/02/76 12.11 ken conn I think that this has already been brought up in "notesnotes", with responses. ---------- response 2 02/02/76 14.30 stone pfw I think that it is much better the way it is now.. relatively easy to change a conditioned pattern.. Scott T. Stone ---------- note 40 indenting 02/02/76 12.52 jim pso some time ago the subject of "indenting" (micro i) lines in the editor was brought up. i would like to see such a feature implemented on PLATO because it would make teaching TUTOR much easier. consider the following: * write some question... arrow 1010 specs bumpshift $$ indent symbol (¬D) is MICRO i; TRY IT!! answer (idea,apple) write That is the legend... wrong eve write Try again, for... no writec ntries¬,¬,¬,etc. endarrow wouldn't it be easier to teach the concept of regular ¬+ judging commands if we said: after the arrow, judging com- mands begin at the left margin. if that judging command is matched, the indented (regular) command which follows is executed. this would clear up many fuzzy heads. comments? ---------- response 1 02/02/76 13.43 jan park since i don't have to teach many new authors how to use tutor, this feature would mean little to me, other than: 1. a few extra key presses to set the margin 2. loss of the margin space (assuming a micro i will use up a word of space in the block(?)) your suggested change would probably neither hinder nor help the existing programmer/authors whose responsibilities do not include training new authors...so, if it makes your task alot easier, i see no reason to delay the change. ---------- response 2 02/02/76 14.11 fritz ames I like it! It makes the complex oddities of judging much clearer and easier to read. How long has this been allowed? ---------- response 3 02/02/76 14.23 ken conn If these ¬Ds would be used a great deal, as they would probably be if most people originally learned with them, would it not be wiser to select a character which would not look !rather strange"(for lack of being able to write out what it really appears as) when printed? Or, if CERL is going to be getting a Diablo printer soon, is this going to be of any worry? ---------- response 4 02/02/76 14.49 judy pso I thought the idea was to make the additional indents optional. Thus, one could choose to indent any group of lines (commands) in order to improve legibility. Wouldn't just TAB TAB be the simplest way to do it? ---------- response 5 02/02/76 18.13 tenczar s to answer Judy, the problem is that some TUTOR commands interpret a blank command following them as a continuation of the previous command and some don't...so, consider the following (in light a non-visible indenting scheme): arrow 1010 answer cat write yes, you have the correct answer George is the last line above a part of the -write- command or a new -answer- command!!! ---------- response 6 02/02/76 18.46 clark lawyer Is this in AIDS somewhere? I could not find it.. ---------- response 7 02/02/76 23.03 alan reading An "invisible" character (visible only to the condensor and possibly the line printer) to delimit indented statements would be preferable to ¬D. ---------- response 8 02/03/76 08.21 broadus css Not true! I think you would want to know in the afore mentioned case that the answer was not part of the write. I like it the way it is and hope that it will stay. One more step towards structured programming! Bob Broadus ---------- response 9 02/03/76 08.41 judy pso to clark/lawyer (and others): This feature is NOT in aids.....this is a discussion of _ideas_. ---------- response 10 02/03/76 09.21 friedman csa Indentation was announced some time ago, and has been a working feature. Should it not therefore be in Aids? ---------- response 11 02/03/76 12.49 alan reading Though the case sited in Tenczar's response illustrated his point, it is unlikely one would find such a construct in actual code. The H option can answer any question you might have about a particular line. Is the following so incomprehensible? arrow 1510 answer lalala write hoo hoo hoo answer googoo write Absolutely! no write Absolutely not! pretend those write statements all begin with a ¬D ---------- response 12 02/03/76 15.17 tenczar s not to overkill, but: did you know that... answer lalala write hoo hoo hoo and answer lalala write hoo hoo hoo are currently identical in function now!!! That is, the answer command "uses" a blank tag as the "mark" to tell that this next line is additional words for the on-going answer command! This is _THE_ problem I have been trying to point out with an invisible indentor. ---------- note 42 infratouch 02/02/76 13.44 b sherwood s Beware of strong infrared light sources hitting your touch panel! At certain times of the day my touch panel didn't work. This was finally traced to reflected sunlight off a neighboring building, which jammed the phototransistor detectors in the touch panel. ---------- response 1 02/02/76 14.52 judy pso This was _very_ angle-dependent, since the reflecting source (a bright strip) was relatively small. Once the cause was understood, we could see the effect by shoving the corner of the table back and forth a couple of inches! Very odd... ---------- response 2 02/02/76 15.53 jan park lucky people working in places with WINDOWS!!!!! ---------- response 3 02/02/76 16.54 bruce iehl mtc I think we should start a new lesson: PLATO's believe it or not!! ---------- response 4 02/03/76 08.32 avner s It is already on line--called "PUBLIC NOTES", I believe.... ---------- response 5 02/03/76 12.09 lombardo ed I'm at the UIUC campus, what's a window? dal ---------- response 6 02/03/76 17.26 judy pso A -window- is a command which allows you to restrict the portion of the plasma panel on which line-drawn displays will be shown. Line-drawn displays _outside_ of the -window- will not be drawn. ---------- note 44 margins 02/02/76 14.07 olson ced I once saw a lesson that would right justify text in a lesson so that the margins would all be equal. Does anyone know the lesson name or remember seeing it? Thanks.... DˇO ---------- response 1 02/02/76 14.08 koning csstaff Lesson jdewk will do that (also contains instructions), and so will the new version of ID/SD that Marshall Midden (middn/p) is writing. ---------- response 2 02/02/76 14.47 mike b cornell The new version of SD is called SM, and it has some very nifty features, including text centering, text justifying, char load, lineset load, and so on. The only problem is that it is not documented yet. Good luck. M¬ ike≤≤≤≤ˇB ---------- response 3 02/02/76 14.51 berger mfl And I suggest that you get permission from midden - p, the author, before using it, since it is not announced. ---------- response 4 02/02/76 15.12 mike cornell Gee, the folks here sorta remember it being announced in new system feature notes..... Mike Huybensz ---------- response 5 02/02/76 16.16 midden p The "sm" in the editor was never announced, but I did say that it was coming, and if you wanted to use, just ask me. Thanks for announcing it though, saves me the trouble... also I need comments from users before it finally gets released, and after the rewrite of the help pages. (yes, some help pages exist!) Marshall Midden Note: circles and circlebs are really "ncircle" and "ncircb" in "sm", in the event that this will soon be necessary after a coming conversion. ---------- note 49 -atc- 02/02/76 14.31 stratton pfw As a resourceful author on the system, and seeing that everyone else makes requests for new commands, I am very interested in using conditional -at- commands, perhaps you inventful systems people could come up with an -atc- command? Anyone else thinking what I am? atc n3;1231;2435;256,256;912; etc. (fine and course) ---------- response 1 02/02/76 14.35 stone pfw Oh no..... while we are at it.. how about a conditional -unit- command and a conditional -randu- command.... *snicker* ---------- response 2 02/02/76 14.38 nate iua202 hey, that's just executable commands...how about a conditional common and storage commands, etc? bearded-wonder ---------- response 3 02/02/76 14.44 travers mxc All you need is: calcs condvar,atvar‎1231,1010,342, ad nauseaum at atvar How about a conditional statement label? ---------- response 4 02/02/76 14.49 mike b cornell If embedded commands don't offend your sensibilities you can always try the following: writec var,¬0at,101¬1,¬0at,121,144¬1,, Ad nauseum. M¬ ike≤≤≤≤ˇB ---------- response 5 02/02/76 14.57 judy pso In order to avoid endless proliferation of commands, if a job can be done reasonably well and reasonably clearly with two current commands, then a new command (or command form) is not considered desirable.... As was pointed out, calcs n1,n2‎101,201,301 at n2 is easy to read and works right now. ---------- repsonse 6 02/02/76 15.38 mike cornell Anyhow, if you want a conditional -at-, why not make it of the same form as a conditional -do-? Who needs it anyhow? ---------- response 7 02/02/76 16.19 marty smith orl why just conditional stuff like that? why not conditional -inhibit- commands: * * inhibit mood;name='berger',course=any;all;none $$ - = kill * 0 = bad * + = good i mean, jeesh! (just kidding. mike) ---------- response 8 02/02/76 17.07 jones mcl I have a personal preference for the embeded conditional looping unit command, which I think would greatly enhance the.....↓↓? Well folks, this is the third request for conditional at that I have seen in that many years on PLATO, though there may have been some I have missed. TUTOR could use a universal conditional construct (like the PASCAL case, but less verbose?) but unless that is provided, let it be. ---------- response 9 02/03/76 08.32 broadus css The solution that travers and judy like, did not conform to te request! He wants to have a command that will use both coarse and fine grid! Are there any other 'brilliant' solutions to this one? Bob Broadus ---------- response 10 02/03/76 09.33 friedman csa People, if General Notes and Help Notes are going to be combined this way, then some questions like this are preordained! So lay off this guy; how's he to know unless he asks? Out of this whole string of replies, there are only a few which are even civil. The sarcasm is appropriate if this person is an experienced author; if not, it's highly out of place! ---------- response 11 02/03/76 09.55 marty smith mtc Attention bob braudus: the solution mike besosa gave was the only solution. that is, a -writec- with embedded -at-s. after that, everything else was playful kidding. i really would like the -atc- command myself, but don't expect it. ---------- note 53 ecs inc? 02/02/76 14.45 berger mfl Our BASE ECS alottment is changing. Is there any reason to expect this? Incidentally, it is increasing, and I don't mind that at all. ---------- response 1 02/02/76 14.50 golden s That is an artifact of new ecs allocation code. It is not in fact increasing. ---------- response 2 02/02/76 14.54 al9000 german Will lesson 'site' be revised to better indicate what is going on with the base allotment? ---------- note 63 aidserror 02/02/76 15.05 bowery comm From the DATA page goto the writeup on -storen- and press NEXT. ---------- response 1 02/02/76 15.33 jmk pso Fixed. Thanks. You can report errors like this in "psonotes". A general announcement to this effect will be coming soon. ---------- note 65 no charset 02/02/76 15.16 j thiher isumedia Perhaps this would be seldom used but I think the condensor should flag it as an error when you attempt to load a charset from a file that doesn't exist.(changed the name today) I know it does if you try to load common from a file that doesn't exist. At least it would have saved me some time looking for the reason he chars weren't loading properly. jim ---------- response 1 02/02/76 19.06 frankel p The -charset- command isn't really closely examined until execution time. You can check the return the -charset- command gives (see aids for details) and determine the problem that way, at execution time. ---------- note 66 no line 31 02/02/76 15.30 jim pso can system messages be sent on line 32 instead of line 31 as the most recent ones were sent? line 31 is used for consulting and talking. thanks ---------- note 72 key ignore 02/02/76 15.56 bowery comm I'm not sure how many people have experienced this BUT: Recently (last month or so) I have noticed that if I do not press a key for a little while (couple minutes[?]) and then start typing again, the first key is lost. This happens quite often at both -pause-s and -arrow-s. Response is fine after keys start comming through. ---------- response 1 02/02/76 17.42 judy cornell Yeah! judy jeska ---------- response 2 02/02/76 19.08 frankel p I have certainly noticed this at more remote sites, although I don't think I have ever experienced it at my local, direct-wired terminal. We will keep this in mind, and perhaps begin prcbing around. ---------- response 3 02/04/76 12.55 truss cdcc Yep, I've noticed too on the dial-ups. ---------- response 4 02/04/76 13.29 bowery comm Could this have something to do with high site numbers? (another varible in common with many remote sites[?]) ---------- note 73 ecs parity 02/02/76 16.05 artman uimatha Are we supposed to report those beautiful System failure ecs parity messages? (Just got one a couple of minutes ago--and I say beautiful because it bombed just me and not all of you!) eric artman ---------- response 1 02/03/76 14.08 judy pso The system knows internally that you have gotten tossed off. Since it does not specifically request that you report it, I wouldn't bother. ---------- note 85 L limit 02/02/76 19.34 dave infe Can you EXACTLY simulate the effect of running student records under heavy-user-load primetime conditions by running during off-hours with CP time limit (L) set low? If so, what value of L represents the worst-usual- case of heavy user load? Thanks. Dave ---------- response 1 02/03/76 08.51 judy pso The goal is that the system should support 2 TIPS at full load. Working it out, 1000 ms/sec divided by 500 users, gives 2 TIPS. Some use fewer, some use more. I would think that a cpulim of 1.5 would give you a feeling for the "worst-possible" conditions. Trying to get the "feel" of a lesson as a student is very hard. You have to really TRY to go at the pace a student would go. You also have to work for a long time, in order that your TIPS will average in the same way a student's will average. ---------- note 86 SS bug 02/02/76 21.00 anderer ee In service, when it comes to a line of code it can't display, it says 'Next ot skip. Back to quit'. Well, the next works fine, but Back throws you out of service and back to the editor. It hasn't done this before. dg ---------- response 1 02/02/76 21.34 halvorsen pcp Thats because this isn't the old service. I'd say that the word "quit" would sorta imply that. Tim ---------- response 2 02/03/76 08.53 judy pso Halvorsen.... your response confuses me. The BACK to "quit and switch into ID mode" is a very very useful feature. Do you mean that it doesn't happen to work _today_ in SS, or that you don't intend that it should ever work? I would object strongly its loss! ---------- response 3 02/03/76 09.19 michael english It's particularly useful when SDing a long display whose length you don't know, but which has a non-display command at the end. Thus you can stop when you want, without skipping the unrecognized command and showing the next display command, which may be part of a different display. ---------- response 4 02/03/76 09.53 whansen csa "BACK" does seem logical to return to editor. "NEXT1" or some other might be used for the "ID now" meaning. ---------- response 5 02/03/76 13.17 koning csstaff How about being consistent and using the same convention as in the new DS: NEXT to stop displaying, NEXT1 to continue? ---------- response 6 02/03/76 13.46 midden p Hummm... this all seems foreign to me. ---------- note 89 sick print 02/02/76 23.59 george tech Is there any hope for someone fixing the system print routine? The unit tables in my prints have parts of command tags listed as unit names. It is very messy. I have also noticed difficulty in the symbol table, but the abomination in that, was beyond recognition. Thanks in advance. george traynor ---------- note 94 named? 02/03/76 08.42 broadus css what is the name of this notesfile? ---------- response 1 02/03/76 10.33 mike b cornell It says "public Notes" at the top of _my_ screen. M¬ ike≤≤≤≤ˇB ---------- response 2 02/03/76 10.41 michael english It would be nice to know the file name so it could be accessed via such a lesson as "cycle". The same goes for the system features section. ---------- note 97 micro bug 02/03/76 08.54 rick mfl In the past few days, I have been using a micro which I call "edit" to save me some typing when writing certain parts of a lesson. Twice, my micro in ECS has been replaced by a different micro also called edit. (Yes, I'm sure of this--"pack length,answer," is my MICRO-p, "pause" is the other micro's MICRO-p.) Is this a known bug, and if so, what is the cause? ---------- note 101 russian 02/03/76 09.26 a appel uni Does anyone know of a microtable that converts the Russian charsets on plato to the standard russian keyboard layout? I would much appreciate knowing. Andrew Appel ---------- response 1 02/03/76 09.37 b sherwood s In lesson "cccp" (no inspect code) are a micro table and charset for simulating the keyboard layout used in the Soviet Union. Examples of lessons using this are rfly, rshow, and rgeom. This arrangement was created by Peter Maggs (maggs/law) for a Moscow demo a couple years ago. I assume you are already familiar with the charset "russian" in "charsets", which associates letter phonetically for ease of studying Russian. This set is due to Connie Curtin and associates, and its layout is much easier to use if one already touch types English. ---------- response 2 02/03/76 09.52 maggs law Send me a pnote or term-talk if you want any more information on the use of these. Also try pressing term and typing t on the title page of lesson rgeom which gives you direct access to a Russian keyboard display. (Do it fast or you will miss your chance. Alternatively, take lesson rpop and answer no to the first question. I hope you are thinking of making up a learn to type Russian lesson. If you want to see how I do arrows for Russian answers you might want to look at the part of cccp that deals with slides. Be sure to display the lines with H so you can see the hidden font keys. ---------- response 3 02/03/76 10.00 jim pso character sets which can be used by anyone on the system are in lesson "charsetx" (not "charsets"). see AIDS for details. ---------- response 4 02/03/76 10.00 a appel uni This is mainly for my own use-- I own a cyrillic typewriter, and I can touch type reasonably fast on it. I would like the micro to eliminate switching between keyboards (i have a vocabulary drill in my lesson that I use). However, if there is enough interest, I could write a learn-to-type lesson similar to btype (which I wrote). ---------- note 107 SS??? 02/03/76 09.56 snellen medneta What _is_ SS?? AIDS leads to a lesson -service- which, when entered, instructs the would-be user to "press DATA to try this lesson". Pressing DATA, however, merely bounces one back to the Author Mode page, Unfortunately, there seems to be no other way to obtain an explanation of what the lesson is or how it is to be used. --JES ---------- response 1 02/03/76 10.35 corey m Try "ss" in the editor. Jeff Corey ---------- response 2 02/03/76 10.35 celia pso Pressing SS[optional ¬$ of lines] while editing a block takes you to the display editing lesson "service". I couldn't find references to SS in the HELP from the block display, and I admit that the aids description is not as informative as it might be. I'll see what I can do to remedy the situation. ---------- response 3 02/03/76 13.48 midden p This option is disappearing, that is why there is little mention of it around. Notes to pso people about where this is mentioned would be appreciated. Marshall Midden ---------- note 110 errmess 02/03/76 10.22 kehoe shs ... On the last down, 1015 AM, we failed to receive a coherent message outside of 'Sorry...' After that, it was mostly flashing the error light and writing random chars on the screen. ... Perhaps the situation could be corrected. Dave Kehoe Springfield High School ---------- response 1 02/03/76 14.49 rader s We will try. For the record, what site/station was this? ---------- response 2 02/03/76 15.37 meredith medsius At various times we have also suffered the same type of problem. It is not unusual for us, during down-times, to not get any sort of intelligible message. As we are also in Springfield (site 1, sta 10), it may be an anti-educational plot manifesting against us by Ma Bell! Yours in limbo, Mark B. Meredith/medsius ---------- response 3 02/03/76 16.02 meredith medsius ah-HA!! It just happened again...the visuals are nice, but I'd rather know what to expect next and/or when! ?any help? mark meredith ---------- response 4 02/03/76 17.22 snellen medneta As a matter of fact, right here in good ol' Urbana (site 22, stations 5-3,4,5) we didn't get any message at all, intel- ligible or otherwise. The terminals just stopped working; the usual "plato off" sign never appeared. --JES ---------- response 5 02/04/76 11.54 kehoe shs ... Springfield High is at 1-3. As far as the messages...we occasionally get a plato off message immediately after the crash, but that is all the intelligible remarks that we get. Thanks Dave Kehoe ---------- note 111 errnotes 02/03/76 10.24 benvo me Execution Error in Notes Unit Storload State Regular Last Command Stoload Join Sequence finish 2 statlv 1 Occurred When attempting to enter notes after recent crash at about 10:15 am today. (Prime Time) ---------- response 1 02/03/76 10.31 woolley p Thanks... ---------- note 117 rotate 02/03/76 11.20 kane phys Why is it that at some rotation angles the line of writing is not parallel to the rotated lines as in the following: unit test size 1 rotate -37 rorigin 1010 rlocate 0,0 write test test test test nrdraw 0,0;152,0;skip;152,16;0,16 ---------- response 1 02/03/76 12.52 white p That is an old error in the sized writing routines, that I first learned about last week. I expect it is due to the roundoff error, after each character, in wherex,wherey. I hope to be able to fix it, but I don't know how yet. One warning though: after it is fixed, the location of sized writing in many, many lessons will shift up to maybe 8¬0! ---------- note 118 keyboard 02/03/76 11.35 re hilton mtc Does anyone have a drawing of the PLATO keyboard I can use in a lesson? answer by p-note to re hilton/mtc. thanx. ---------- response 1 02/03/76 12.54 warner iu I've got a charset drawing that will do the job in lesson help. See unit keyboard in help, also charset helpchars. ---------- note 119 wasted $$ 02/03/76 11.41 oi cerl if there is not enought memory allocation available why make us waste time and money by answering the phone and getting on the system!!!!!! it would be better not to let us key on if there is not enough space!!!!!!!!! ---------- response 1 02/03/76 12.12 fay o Perhaps the auto-answer ports could be interfaced with the software to give a coded series of tones indicating the computer could not guarantee 1500 wds of ECS. This would determine whether or not you could sign on, I believe. The cheapest alternative would be to disable the auto-ans altogether when 1500 wds could not be promised, but then the user would not know if the auto-ans was broken or if memory was short. Perhaps they could be set to be "off the hook" when memory was short; the busy signal would be your signal a) ECS is tight (ie, the system is "busy"); or b) all the dial up ports are in use. ---------- response 2 02/03/76 12.12 perry uw The check to see if there is enough memory is not made until your password has been entered. It would not make much sense to check memory available at the moment of dial up, since that would certainly change by the time you had finished signing on. Also there are only 8(?) dial up ports, as compared to 950(?) or so terminals that are more permenantly wired into the computer. ---------- response 3 02/03/76 12.18 fay o Ooops. Forgot about that little point. Maybe an exception could be made in the case of the dial-ups? I can see where phone charges could be thus incurred unnecessarily. ---------- response 4 02/03/76 17.36 carter comm The phone company will usually cancel short calls if you pay for the long ones. ---------- response 5 02/03/76 18.25 frankel p (I don't know about that previous response! I've never heard that before!) In many, many cases there are things that can be accomplished when memory is limited -- this often includes editing or talking. And the two or three minute call it takes to find out how much memory is available won't cost you more than $2., I believe. Considering how much you must pay when you do stay connected, I can't believe that that is really significant. ---------- note 121 plato iii 02/03/76 12.11 hecht ed I am in the process of converting an old PLATO III lesson onto PLATO IV and have encountered two obsolete commands that no one here seems to recall--"must" and "diddl". Does anyone know exactly what these commands did? They were used in this set of coding: arrow 1010 answer segments AB and CD must segments,segment must AB,BA must CD,DC diddl are,congruent I am assuming that the "must" command was used to specify alternative answers and the "diddl" command to specify extra words which were acceptable. If this is not correct, please send a pnote describing the proper function of those commands. Thanks. Jim ---------- response 1 02/03/76 13.52 jim pso there are a lot of PLATO III manuals floating around offices at CERL. you can read about it here. briefly: diddl --- ignore these words in the student response for the judging commands for this arrow. put these words in < > for PLATO IV TUTOR must --- replace this command with an -answer- command in PLATO IV TUTOR. There were some (confusing) differences between must ¬+ answer on PLATO III TUTOR call a consultant if more info is needed. jim ghesquiere ---------- response 2 02/03/76 14.18 b sherwood s -must- let you list keywords which should be present. It seems to me that the closest analog at present would be: specs okextra answer horse,pony ---------- response 3 02/03/76 14.34 hecht ed Thanks for the information. Jim ---------- response 4 02/04/76 08.54 berger mfl If you look in lesson projects, I believe that you will find mentioned the fact that these commands may be implemented on PLATO IV. ---------- note 130 mathfreeks 02/03/76 13.46 lombardo ed I would appreciate a little assistance with a problem I received in the mail the other day. It is one of those insane verbal math problems designed for those people who have both math comprehension and ability. I have neither, if you do then have at it and respond to me via p/notes with the answer and method of solution. It is in lesson AVIATION option "H.! Thank you, dal ---------- response 1 02/03/76 14.23 dr stutz mtc !rope is an long as the age of the monkey"....... What are the units of length for the rope (i.e. inches or feet)? ---------- response 2 02/04/76 10.13 lombardo ed That information is not given, all that is available is included. Kinda strange, huh? ---------- note 137 ext. jack 02/03/76 14.37 stark ames Does anyone know the format for the ext. jack. This is for the new type terminal..... thanks ---------- response 1 02/03/76 15.03 ld francis mtc see lesson "external" student mode ---------- note 142 helpcalcc 02/03/76 15.58 putch arizona Can anyone tell me why the calcc statement in line of the first response will not cirflag‎1 when cdex=2? * unit pacdraw do circle1,cdex‎0,2 size 2 at 1814 write P at 2414 write A at 3014 write C do circle1,cdex‎0,2 at 1846 write P at 2446 write A at 3046 write C size 0 * (cont. in response) ---------- response 1 02/03/76 15.41 putch arizona unit coord1 calcc cdex-1.cirflag‎0,0,1 * unit coord2 calcc cdex-1,cirflag‎1,1,0 * unit circle1 at 104+256(cirflag),48+96(cdex) ncircle 48 do cirflag,x,coord1,coord2 * (All variable have been -define-d) ---------- response 2 02/03/76 16.08 dave fuller uimc Because you should have calc_s_, not calcc... The calcc indexes into individual assigned calcs, and the calcs assigns a value to a variable issued in the first tag, ie: calcc n1,i1‎1,i2‎2,i2‎nˆ2,bleep‎funct(gorki) calcs n1,i1‎1,1,2,sqrt(infin),zˆ2 And, since numeric values are recognized as dummy arguments for the calcc, no condense error was issued. ---------- note 145 template 02/03/76 16.04 b sherwood s People have noted that sometimes curriculum options fail to get templated in course records. Apparently this is an intermittent failure, since tests which failed yesterday don't fail today, Please tell us of any failures, with enough details that we can try it, too. Thanks! ---------- note 147 circle at 02/03/76 16.10 b sherwood s There is still some discussion (controversy↑) over what should happen after drawing a circle or partial circle. It has been proposed that the screen position should be restored to the center. However, after a partial circle (or rcircle or gcircle) it seems natural to many people to keep drawing from the last point on the arc. For consistency, it might also be advisable to not return to the center of a full circle. Two questions therefore: 1) Should there be an "at center" following a partial circle? 2) Should there be an "at center" following a full circle? ---------- response 1 02/03/76 16.23 dave fuller uimc Since the center of a circle can always be saved somehow, it would seem to be more logical to assume that the -wherex,y- locations should remain at the end of the arc, or circle. Computing the endpoint of an arc for continuation of a graphic can be messy, whereas the storage of the screen location is not. In any case, if I needed to return to the center of the circle, I would be doing it myself, anyhow, just to preserve program clarity. Using any assumption, such as the default -where- after a circle draw, is likely to cause problems sometime in the future, no matter how sincere the reasoning for making it that way originally. ---------- response 2 02/03/76 18.26 j wilson unidel I would also favor retention of current functions. In addition to Dave's reasons (and because I know of several programs that would require modification), it seems to me that the relation, -draw-:-rdraw- :: -circle-:-rcircle-, should be retained as much as possible. Consequently, all four commands should leave *wherex,y* at last screen position, or -draw- and -circle- should and -rdraw- and -rcircle- should leave ir unchanged. ---------- response 3 02/03/76 18.31 blomme s One warning: anything tied to "last thing drawn" may be a problem, especially considering that the method used in drawing circles may be different on the newer terminals. Only when writing or drawing has an obvious end point should the where be set there. ---------- response 4 02/03/76 18.40 j wilson unidel Maybe then we should stop saying that circles and arcs can be drawn clockwise or counter-clockwise too. ---------- response 5 02/04/76 08.21 judy pso The clockwise/counterclockwise possibility is an important feature to people who care about direction of "flow", such as elementary reading people who want to form letters in the same way they are printed. I doubt very much if it will go away. ---------- response 6 02/04/76 09.38 gilpin peer It's use is important in pseudo-animation, too. I _hope_ it doesn't go away. ---------- response 7 02/04/76 10.03 jmk pso An arc has obvious endpoints whereas a circle does not. How about: rcircle radius $$ returns wherex,wherey to center...(where else?) rcircle radius,angle1,angle2 $$ leave x,y at last pt drawn Wouldn't this permit relocatable drawings consisting of intermixed straight and curved sections? ---------- response 8 02/04/76 11.17 white p Consideration is currently being given to what will happen when the terminal computes and draws the lines for a circle, rather then sending the endpoints of the lines over the phone line. It is expected that circles would then be drawn so fast that the "flow" might not be important. Also, whatever is decided now should be implementable on the "intelligent" terminals. ---------- response 9 02/04/76 17.18 j wilson unidel I'd also settle for dividing the resetting of *wherex,y* between one- and three-tag -circle-s. Larry, do you propose that we should start doing our jumping frogs and flowing symbols manually? Or will there be a -scircle- (for slow circle)? and an -srcirclb-????? It boggles the mind. ---------- response 10 02/04/76 18.37 blomme s While arcs do indeed have a direction defined, they may not be drawn from one end to the other by the actual algorithm involved. In this case, since a direction is fairly clearly defined, it is only a problem of being sure that the system algorithms do a final "clean-up" to reset the final screen position. However, consider that there are still some nasty problems here when one has arcs that go off-screen or are terminated by running out of "windows"--then the question seems to be whether the final where position should be that specified by the command or the position of the last dot (in the direction specified) that is actually plotted. ---------- response 11 02/04/76 18.43 david cerlcc Last dot plotted. ---------- response 12 02/08/76 11.30 shirer vu Last dot plotted for both arc and circle: I am recommending this for consistency...because it is possible to specify an off-screen circle center with on-screen plotting. Leaving wherex,y set to an off- screen location doesn't feel right to me. If this is ok. then would be most convenient to 1) leave at last dot plotted for arcs 2) leave at center for complete circles ---------- note 149 bad erase 02/03/76 16.40 clark lawyer Try this! At an arrow, press SHIFT-SPACE, and then press a couple of letters. Now, press ERASE once and then press EDIT. Surprize! The char that remains on the screen really isn't there at all! ---------- response 1 02/03/76 17.22 hagerman ames So what else is new? ---------- note 158 ecs counts 02/03/76 18.36 golden s New ECS accounting machinery has been installed. It should keep correct records of your site ECS use. Let us know if you detect an error. The new algorithm will not delete authors if there is a student using the same lesson at the site. It will delete authors by pressing -STOP1-, so finish units should be done. At present, author deletion is automatic; however, we plan to make it an option controlled by the site director. ---------- response 1 02/03/76 18.51 lewis unidel Will the new algorithm permit you to run a lesson which is in ESC but under a different site$/≤-≤¬/≤? ---------- response 2 02/03/76 23.40 daleske ames THAT is excellent!! Thankyou!! The finish unit problem has been a hassle, though, in most cases can be properly handled by having a unit on entry take care of the problem. But, allowing the author to remain if a student is in the same site, that is pure genius! Thanks very very much! John ---------- response 3 02/04/76 08.22 judy pso answering lewis: No, you do not get credit for a student in a lesson at a different site. ---------- response 4 02/04/76 08.57 berger mfl That is a tremendous improvement! Thanks! One question - will an author be deleted from a lesson that a student is in if that lesson uses storage? ---------- response 5 02/04/76 13.36 bowery comm Also, will the system reserved var -backout- be set correctly? ---------- note 161 kill stack 02/03/76 19.10 fritz ames If you excecute this code: unit test arrow 1010 join unit write after the join * unit unit specs branch 0 ok no $$ to keep it from falling thru to the 0 0 $$ (originally 0 was a doto endlabel) The write in unit test will not be excecuted. With a little sleuthing, you will discover that branching out from underneath a -specs- DESTROYS your join stack! Have fun, somebody... ---------- note 168 lost notes 02/03/76 23.11 woolley p Apparently an error in the notes program has wiped out several of the early notes in this file, with no chance of recovery. I have not been able to find or even duplicate the error, so please -- if you notice anything strange at all, please contact me or another available system person immediately. ---------- response 1 02/03/76 23.50 koning csstaff I don't know if this is related, but several people have seen notes with responses listed as having no responses. For an example, see note 136 of 1/20. ---------- note 171 code gen. 02/03/76 23.53 stewart arizona In lesson 'printer' is a notes file set up to discuss the various types of code generators on the system, and the possibility of some sort of direct access to the source code in a lesson by these editors. Please express your feelings on these matters in lesson 'priniter'. Bruce Stewart ---------- note 172 storew? 02/04/76 01.01 mike cornell I know that I might get jumped on for making a frivolous request, so let's hope this one's not so.... Wouldn't it be nice to have a judging command that was the universal equivalent of -storen-? In some lessons it could be very useful to drag out each word in the judging copy, one at a time.... it could be of the form: storew n1,20 $$ 20 being the number of chars to save It should take out numbers too, but stored as characters. If people wanted the numbers separately, they could use -storen- beforehand. Naturally, this is not too high a priority, but it does sorta seem logical? Mike Huybensz ---------- response 1 02/04/76 07.48 frye mfl Here here! This would be a perfect addition to the answer judging capabilities -- currently, any kind of word-by-word parsing done has to be done taking into account all the different delimiters TUTOR has...something on the system level that would take care of that would be won- derful! An addition would be to allow any -specs- in affect (i.e. alphxnum,allwords) to alter the word structure for purposes of -storew-. Dave ---------- response 2 02/04/76 10.30 warner iu Yes. This is definitely a needed function when constructing code generators, flexible response analyzers, even dialogues. I would like to see this given a high priority (especially as, as I understand it from judging, the word length and position are stored in the judging "hash tables" and can be used to find the word in the buffer). Meanwhile, some purposes can be found by using the numerical vocabs options to decide what words are being used in an answer without a horrid mass of related but different concepts. ---------- response 3 02/04/76 10.39 fritz ames Oh lord... I wish I wish...! Take a look in lesson "myra" if you'd like to see the first thing I'd zap a storew into and get rid of a ridiculously huge parsing calc! ---------- response 4 02/04/76 11.00 mike b cornell This would also be of use to me. I'd like to see it given a high priority. M¬ ike≤≤≤≤ˇB ---------- response 5 02/04/76 12.25 michael english Same here. The vocabs/concept type of judging can be augmented enormously by a bit of simple parsing on the part of the user, but it invariably requires that words be sucked out of the judge buffer and put into an alphameric array. This simple process is very wasteful when done on the user level since info the system judger already has must be recomputed in order to determine word boundaries. ---------- response 6 02/04/76 13.51 bowery comm Actually I was going to suggest something similar to the -open- and -close- commands that work on words (as defined by the system) rather than single chars. I have been putting it off to try and figure out exactly what it would do. After considering the possibilities, a very logical way for it to work would be: openw vector(0),20 $$ 20 vars available to open into Student enters "some somethings are things" and: vector(0)= some¬0¬0¬0¬0¬0¬0 vector(1)= somethings vector(2)= ¬0¬0¬0¬0¬0¬0¬0¬0¬0¬0 vector(3)= are¬0¬0¬0¬0¬0¬0¬0 vector(4)= things¬0¬0¬0¬0 If the vars needed are not available then a "no" judgement ensues. ---------- response 7 02/04/76 22.33 mike cornell How come no system responses? If i understand bowery's scheme to mean that words too long to fit in one word slosh into the next, I don't like it, because it requires additional programming to tell if a word is a continuation of the previous one. Mike Huybensz ---------- response 8 02/05/76 11.32 tenczar s suggestions read Blomme and I have already thought of implementing something like this, with additional features (like giving the screen x,y coordinates of each word in the answer but...it will have to wait until more important things are done. Thanks for verifying our opinion for the need for such a thing. ---------- response 9 02/05/76 15.19 bowery comm re: mike If you will notice, the format of char codes in the vector is the same one the editor uses. Believe me, (the editor doesn't do anything it doesn't have to do. (especially in the list loop) Beyond this, the command might store the locs of the beginning char code of each word into a buffer like the -findall-, eventual -matchal-, etc.. ---------- note 173 HELP! 02/04/76 01.16 parker arizona To any and all interested PLATO users: I have set up a notesfile in lesson 'proj211' to gather opinions for a class project. The subject that I am dealing with is computer games. Specifically, (1) What educational value do games have? (2) How difficult is it to produce a GOOD game? (3) What is the future of computer games? (4) And any other comments relating to this area. This notes file will be temporary, lasting for about 2 to 3 weeks. I will continue to accept opinions until then. Also I will be periodically clearing out the common and printing it, just in case the file becomes full before that time. I appreciate your opinions. Mike Parker ---------- response 1 02/04/76 01.20 parker arizona PS : In case you hadn't noticed, this note is a duplicate of my original that was stored 2/1. This note got erased somehow. Therefore for the benefit of those who haven't seen it, I have re-entered my note. Mike Parker ---------- note 182 genetics 02/04/76 09.33 lewis unidel Anybody have any lessons on genetics and/or animal husbandry? We are building a mini-library and want to avoid duplication. (lessons indexed by bioindex are known to us) ---------- response 1 02/04/76 10.40 fritz ames You might contact brant/ames, he has several lessons in this field. ---------- response 2 02/05/76 08.59 bonetti ced Look in the medical network catalog (mclcat) for some genetics lessons. ---------- note 186 indenting 02/04/76 10.24 ghesquiere cerl many of the responses to the previous note on indenting (2/2/76, public note ¬$40) discussed implementation details. ---------- i would like some comments concerning the concept of indenting "answer-contingent regular commands". this could be used when TUTOR is taught to new users in an effort to clairify the difficult concept of which regular commands are executed after a student response. of course, the use of indenting would be _optional_. if you choose to use it, fine...if you wanted to continue with the current format, you could do that also. P.S. the current method of indenting (using "¬D") can be fully operable with just a little more work. while any other method would take a week of Wednesdays. ---------- response 1 02/04/76 10.32 ghesquiere cerl a sample unit would be: unit test $$ if implemented, indenting would write some question $$ be O¬ P¬ T¬ I¬ O¬ N¬ A¬ L¬ arrow 1010 specs bumpshift answer correct response write correct response comment wrong incorrect response write incorrect response comment no writec ntries¬,¬,¬,series of hints endarrow between the -arrow- and -endarrow-, all judging command could start at the left margin. regular commands are indented ---- hopefully this would make it clearer that these regular commands are executed only if the previous judging command was matched. comments? jim ghesquiere ---------- response 2 02/04/76 10.43 fritz ames I think it's an excellent idea. One thing I don't understand, though... You keep saying IF indenting is implemented. When I saw your note ¬$40 (with the comments saying "TRY IT", so I did), I went and tried the indenting, and it condensed and worked fine. What additional implementation has to be done? ---------- response 3 02/04/76 11.00 maggs law I think optional indenting would be a big step forward for new authors particularly. When I first started writing in Tutor I found the judging business by far the most confusing part, and the new programmers I have initiated have had the same problem. Would it be possible to indent any command one desired, and to indent optionally either 0, 1, or 2 tabs for further readability? ---------- response 4 02/04/76 11.21 celia pso The only reservation I have is that the new author might be led to believe that indenting or not indenting causes different behavior. For example: arrow 1010 answer some answer calc right‎right+1 write That's right. answer some other answer write Yes! endarrow When "right! is only incremented in the one case, and not the other, the new author will say confusedly, "But it should always be incremented, because it is not indented!" I'm afraid that in trying to clarify, we may just muddle. ---------- response 5 02/04/76 11.24 alan reading Indenting answer-contingent regular commands makes it much easier to see what is happening at an arrow. This practice should be encouraged in new authors not only to make it easier understand arrow processing, but as good coding practice (like defining variables). You can indent any command you want as often as you want: write hello! ---------- response 6 02/04/76 12.50 friedman csa Speaking as one who has taught several programming languages. I would suggest separating the indenting from the explanation of how judging works, for the reason Celia mentions. The student should be taught that he SHOULD indent answer-contingent commands (do we still use that term?), but also carefully taught that the functioning of the commands has NO bearing on how (or whether) they are indented.↓ By the way, he should also be taught to indent the commands within a -doto- loop, or any other kind of loop, and those within the branches which make up an if-then-else type of construct---simply for the increase in readability this will get him. ---------- response 7 02/04/76 13.33 sacco conn As a new author, thank you very much for this conversation. $ j≤ j≤ j≤ joe ---------- response 8 02/04/76 16.53 nate iumusic That last note brings up another point. With the new notes-scheme, will animation now be less frowned-upon than in the last version? Hello??? Nate ---------- response 9 02/04/76 19.18 k mast p No animation please. We will be able to print these files, and no matter how cute it is on a plasma panel, animation on paper is yucky. ---------- note 190 in use 02/04/76 10.34 lombardo ed It occurs to me that while I am sitting here trying to make a change in my code but cannot get access because my lesson is in use that it would be nice if I could at least tell where in my lesson the user happens to be. I don't want to condense him out of the lesson, but on the other hand I would like to know how long I am going to have to wait before I can condense. I suppose I could attempt to get him on TALK option, but I don't really want to disturb him, just find out how long I may have to wait. Isn't there some reasonable way to discover where he is in the lesson? Ah, my wording above is sketchy I notice, I know I can change my code, and have done so, but I cannot sample its effectiveness until I can condense...better? ---------- response 1 02/04/76 11.06 mike b cornell What I usually do in this situation is put a notice on the title page of the lesson stating that development is underway and that deletion is possible. Another author I know has a flag he can set in the lesson's current version which will display a "caution, author at work" message at the bottom of the screen when he begins work. At least this warns the users that they should expect to get deleted, which is somewhat more considerate. M¬ ike≤≤≤≤ˇB ---------- response 2 02/04/76 11.19 daleske ames An even better way is to have a little flag set in a common variable and test it in the -imain- unit/s. If set, any place that the person goes to a new main unit, the note will be seen! John ---------- response 3 02/04/76 11.58 lombardo ed There are several ways to accomplish warning a student that he may get bumped, but I don't want to bump anyone, just want to know where in the lesson the student is. ---------- response 4 02/04/76 12.48 halvorsen pcp Can't the display simply give what unit the student is in? Many times I have wanted something like this so that I could see how long I had to wait before recondensing. Most times, you might want to make a change to a finished lesson that is not currently being tested. In this case, the above "solutions" would not apply. ---------- response 5 02/04/76 18.46 blomme s Those who are running an actual course of real students can look at the display given in the course records editor as an alternative. ---------- response 6 02/04/76 09.36 friedman csa Rick, that's no help. It isn't likely that the author is a course director. It isn't even necessarily true that all the students are in one course, or even that all courses using the lesson are on the same site or campus or part of the country.↓ I think the suggestion would be very useful. I AM a course director, and yet have had this difficulty often. In the case of students using my lessons, they are likely to be spread out over as many as a dozen courses (no exaggeration!). ---------- note 191 voltage 02/04/76 10.34 saltan uicc Where can I get hold of the utility routine DRNET which will draw circuit networks on my command. Please contact me -- if you hava any info. Thanks. ********* The name is Saltan of uicc ********* ---------- response 1 02/06/76 10.02 saltan uicc Please answer 191.... ---------- note 203 keyboard 02/04/76 11.28 maggs law Some time ago there was a list in notes of lessons that helped students to learn to use the keyboard. I am looking for a couple of particularly simple lessons of this type that I think were developed for English as a Second Language. Can anyone help? ---------- response 1 02/04/76 11.29 judy pso try roberta/mfl ---------- response 23 02/04/76 18.28 gilpin peer Marty Siegel has modified the ESL keyboard-teaching lessons for use with speakers of English--he is siegel/peer, but I think he is out of town until next week. Maybe alessi/pcp can help. You would need the OK of roberta/mfl in any case. ---------- note 208 users? 02/04/76 12.36 swenson a There used to be a system defined word 'users' which contained the number of users signed on the system. Has this defined word been done away with? ---------- response 1 02/04/76 12.50 michael english *tactive* ---------- response 2 02/04/76 13.11 warner iu And *usersin*, for the number of users in your lesson. *tactive* is the number of users on the system. ---------- response 3 02/04/76 15.07 swenson a Thanks folks... ---------- note 214 lost note 02/04/76 13.15 lombardo ed I was just reading a personal notes and at the bottom it said Continued Press Next, when I did I received the message saying There are no more notes for you. Is someone playing a cutie on me or is PLATO slipping up? dal ---------- response 1 02/04/76 13.20 harkrader o All the continued message means is that the sender had intended to continue his note, pressed LAB1 to mark the previous note 'continued', and never sent the second part or is still writing it. The continued message is just a flag. Al ---------- note 221 error page 02/04/76 15.14 guerra uimc The new ECS management machinery does not keep your lesson in ECS if you see the condense error page upon condensing a lesson. I have encountered said problem several times today: I would condense a.lesson, see the condense error page, press next IMMEDIATELY (or as soon as I glanced at it to make sure none of the errors were serious) and I would see !now preparing lesson--please wait" at the bottom of the page, indicating the lesson was being recondensed. This isn't intentional, is it? This is going to be fixed, isn't it? ---------- response 1 02/04/76 20.49 chabay chem It isn't intentional, and it isn't the ECS alligator -- it's a bug in the lesson that displays the condense errors. (And I agree -- condensing every lesson twice in the middle of the day is a pain!) ---------- response 2 02/05/76 09.39 friedman csa "alligator"....I KNEW that **** thing was a vicious animal! ---------- note 222 cockroach 02/04/76 15.39 bowery comm I usually don't like to write notes like this, but I am overwhelmed: I have noticed that a frequent argument against implementing suggestion "x" given in NOTES is that it would make some existing lessons obsolete even though the general consensus is that suggestion "x" would make the system more logical ↑≡↑≠ general for the multiudes to come. It seems to me that it would be better to make "some" lessons obsolete than to make the PLATO system obsolete because of calcification of evolutionary processes if ya get my meanin if ya catch my drift. Comments? Amendments? Weightings? Considerations? Your opinion is humbly valued, at least by me. ---------- response 1 02/04/76 15.43 berger mfl I agree. Nothing bothers me more than having to work around an inconsistency in the system, and being told that the inconsistency will remain, because it would stop "a couple" of lessons from working. I am aware of the responsibility of the system to keep lessons running, but there are many cases in which a change wouldn't result in too much work to make a lesson work correctly. What's more, the system currently is making changes that make our lessons stop working... so I can't see any real difference from the status quo! ---------- response 2 02/04/76 16.22 fay o There seems to be an inconsistency operating at more than one level here. The system personnel have, in the past, made conversions with full knowledge that the conversion would affect a subgroup of users in general. The usual procedure is/has been to give as much advance warning as possible. The warnings have typically been given a lot of publicity with the assumption that the affected groups are responsible for receiving and acting on the warnings. ---------- response 3 02/04/76 16.58 bowery comm I am NOT indicting any person or group of people (not even sysprogs). What I AM indicting is an attitude which: 1) we are all subject to from time to time 2) is overly conservative as is and 3) seems to be on the increase. I say this because in the past people who adhered to a particular belief for very good reasons, were unable to communicate those reasons because my criticizm was taken as a personal attack. The fact that we all are learning is something we must not lose site of. ---------- response 4 02/04/76 18.48 frye mfl Mike's and Tim Fay's arguments BOTH tend to support what Jim is suggesting -- that while certain suggestions for improvement in the system are rejected on the grounds that the change "would affect or make obsolete existing lessons", others which ARE implemented DO affect those same existing lessons... See any foreign language programmer for num- erous examples. This inconsistency in justification, in my eyes, needs to be examined more closely -- perhaps the next time something is needed but alters some current convention, we can remember these facts.↓ I'm afraid that I personally cannot provide relevant ex- amples either for or against this point, since I've been out of communication for so long. PLATO could well be a main educational mode in the near future -- lets not make it a confusion of inconsistency at the same time! G. David Frye ---------- response 5 02/05/76 09.44 friedman csa Well, there are literally thousands of hours of effort which have gone into the creation of the many lessons on this system. For a change to come along which simply makes some of them stop functioning is simply out of the question, in my opinion. After all, an awful lot of those authors are no longer on the system at all.↓ If someone says to me "find all your lessons which use command -x- and change them, because -x- is going to work differently"...well, I don't even know where to start looking in our 175 or so lessons! And even if I do find -x- in one of these lessons, it may take me hours of work to understand how the lesson works and rewrite it! ---------- response 6 02/05/76 09.49 friedman csa So, when changes are proposed, I don't object to extending the language or even to changing it, IF the changes can be made in such a way that properly written, functioning lessons continue to function properly (after an automatic conversion, if necessary).↓ But I DO object to having to spend dozens of hours going back over functioning lessons (whose original authors are no longer around) to find and fix something which has been working up to now.↓ After all, this "experimental" system is ALSO very much a production system. Or do you want all serious use to cease until the language is "finalized"? NO language is EVER finalized, in this imperfect world of ours. But programs (lessons) DO have to continue to function properly without being constantly rewritten.↓↓ I guess you really rattled my cage on this subject! ---------- response 7 02/05/76 11.55 gilpin peer Altho there will always be disagreements about particular cases, it was my impression that system policy in these matters was balanced just about right. For example, the changes by Larry White to the -circle-/rdraw-/etc. commands originally were thought to entail fixing up lots of lessons manually. They were to be implemented anyway because of the tremendous advantages they provide. Notice was given in a very responsible fashion, and, as far as I know, authors generally approved. As it turned out, a means was found to avoid obligatory manual fixing, which demonstrates again that thought _is_ given to author problems occasioned by change-overs. Admittedly, like the rest of us humans, sysprogs vary in the amount of care and consider- ation they show...and certainly some users have more clout than others when they demand/resist a change...but overall, the balance seems about right to me. ---------- response 8 02/05/76 11.57 fay o George: You are very right and the sysprogs must (and, from what I've seen, do) make every effort to "automatical- ly" convert existing lessons to the new format of the given feature/command. The recent conversion of the datain/dataout commands is one such example. In cases where an automatic conversion is not possible (re: going from "router" to "mrouter") systems support has been most generous. The issue I think Bowery is raising, as pointed out in the response immediately preceding yours, is what he is interpreting as the "wrong" official reasons which seem to be given for NOT implementing various changes. ---------- response 9 02/05/76 12.13 friedman csa Well, I don't mean to be attacking Bowery or the idea that changes should and must come. But the discussion up to the point of my previous reply seemed to me to be saying something like "to heck with compatibility, make changes anyway, comaptibility is not a valid reason for not making a change." If I misinterpreted, ok, I back off.↓ What I said earlier was not intended to be directed at any person, but was meant to state a principle that I think ought always to be kept in mind. So far, it is only sometimes (but most of the time) kept in mind. In other words, I'd not complaining about present practices; I, too, think they're just about balanced, and I wanted to make sure they do not become unbalanced. ---------- response 10 02/05/76 12.15 friedman csa By the way, the conversion from "router" to "mrouter" did not, as far as I know, change any working lessons (did it?), and so is not the kind of change I was addressing. ---------- response 11 02/05/76 12.58 fay o (Well, it did sort of..."router" was made obsolete and generally speaking, there was no easy way to convert (including modules, etc) to the new version. I guess its more of a general feature than a specific command as far as examples go. It was, perhaps, an irrelevant example.) ---------- response 12 02/05/76 14.02 carter comm I must strongly agree with Professor Friedman on this point. It will be increasingly difficult to get people to produce quality software if it is certain to be inoperable in six to twelve months. ---------- response 13 02/05/76 15.10 bowery comm I would like to reiterate a condition stated in my original note. The reference to "general consensus" was there for a very good reason. No one need remind me that this is not a democracy. The definition of what is "general, is pretty much a function of who the people with the right signons are credulous to. This brings up the very important point of how decisions are reached and what variables are weighted to what extent. I would like to see information on this process made more public. This would NOT make decision making a "democratic" process. It WOULD give the "general" user population more of an idea of what the dynamics are that mold the properties of the computer space they operate in. ---------- note 226 ecs off 02/04/76 16.26 clark lawyer I don't know if this matters, but at our site (law) today, the individual ECS usages did not add up to the total ECS use shown. It was off by 250. ---------- response 1 02/04/76 16.35 fay o When you add up the totals presumably you aren't double counting any lessons. Are you adding about 400 for each user at the site? This is roughly the number of extra words of ecs needed per user for various pointers in the execution of a lesson. ---------- response 2 02/04/76 16.40 clark lawyer I did not add any lesson twice. The total ECS shown was actually less than the sum of the individual values. I did not add any extra (i.e., 400/user). ---------- response 3 02/04/76 20.38 andersen s ECS use shown will be less than the sum of the individual values if two or more individuals are using the same lesson or common - the lesson or common appears twice but is only charged for once. ---------- response 4 02/04/76 21.52 berger mfl Making it even harder, storage is charged for each user, but "site" doesn't break down the charges. Things seemed to work nicely today! ---------- note 230 condense 02/04/76 16.47 kane phys I think it would be useful to be able to request that a lesson be condensed (and the condense error page displayed without necessarily requesting that space be found for it in ECS or that running students be deleted. This would allow authors to check on condense errors when ECS is short at their site or when they have made changes in a lesson in which students are running. Since it would not be in ECS, no one could run in such a version of a lesson. ---------- response 1 02/05/76 09.16 golden s Sorry...ecs must be obtained into which the condense occurs. So, that couldn't be done without charge. ---------- note 236 pnote size 02/04/76 18.41 michael english It would be nice if the PNOTE buffer size were as large as the GNOTE buffer. It's impossible to save a section of a large GNOTE and insert it in a PNOTE. ---------- response 1 02/04/76 19.21 k mast p It would be very nice indeed. Would somebody please look into doing it? ---------- response 2 02/04/76 19.33 dave fuller uimc Things are sounding grim, indeed... ---------- note 237 wrap-rap 02/04/76 18.42 warrens uw I haven't read notes for a little while so this might already be known, but here it is anyway. I read in responses by storea, limited by jcount. When the line is exceded and wraps around, the characters are on the same line a they started on untill the position of the arrow. At this position, the character is plopped one line down. ...11↑ˆ↑ˆ↑‎↑ˇ↑ˇ≤¬62222222222222...... 11111111... So it looks like this where the ‎ is the point where the wrapped line drops.It is repeatable. Also repeatable is the RAPID erasure of the second line that is displayed when asking for term-consult and no consultant is on system. Thanks! ---------- response 1 02/04/76 19.17 midden p problem a) have you been able to reproduce this on other terminals? If so, it would be nice to have a detailed example of the problem (like for example: this arrow in this place, and this and that, and then...). ---------- response 2 02/04/76 19.24 fritz ames It happens to my on all 4 of our terminals here. The instance that comes immediately to mind is where I have an arrow at 201, and the first char to wrap around stays on the wrong line. ---------- response 3 02/04/76 19.34 midden p Strange, very strange. Works just fine for me. Anyone have any ideas whats happening here? M.Midden ---------- response 4 02/04/76 20.11 frankel p Looks like it only occurs in monitor mode -- anybody with an example that works in normal mode, please let us know! ---------- note 244 bug killed 02/04/76 19.24 woolley p I believe I have fixed the bug that overwrote some Public Notes last night. There's no way to retrieve the lost notes, but at least it shouldn't happen again. ---------- note 250 varian 02/04/76 20.46 alan nuc It seems to me that for the past month or so, the varian printer 'would be ready in about a week'. Before that, it was just plain down. Can we please have an accurate estimate as to when it will be up and ready to be used. Alan M. Malakoff ---------- response 1 02/04/76 21.35 kawell o Well, I am working on it. But it has LOTS of problems. Maybe, the latest fix might work tonite. But then again. it might not. We will be sure and announce it when it is ready. Hang in there. Len ---------- response 2 02/04/76 21.55 alan m nuc Thank you for the information. I don't mind waiting as long as I know someone is looking into the situation. AMM ---------- response 3 02/05/76 11.38 rader s (It didn't work...) ---------- note 252 aargh! 02/04/76 21.24 schramm cs317 I'm trying to get the system to accept an -answer- command in 2 different character sets, e.g. answer ≥kl7h-≥ Whenever I try to condense this, I get an error message. Since I'm writing a bilingual lesson, I need to circumvent this problem. Any suggestions? ---------- response 1 02/04/76 21.31 fritz ames The problem is in the minus sign. You have answer kl7h-, and -'s are illegal in answer commands. (Along with all other arithmetic operators and most [all?] punctuation). Replace the - font char with something else and you'll be ok. ---------- response 2 02/05/76 09.23 judy pso Paul Tenczar has been working very hard the last weeks (months?) to fix this and similar problems with the punctuation and arithmetic marks. We're waiting anxiously for his final version!! ---------- response 3 02/05/76 11.21 tenczar s yes, soon you will be able to completely define the "value" of every visible character...so that in the alternate font you can "define" what each character is: letter, number, punctuation...etc...by the way, there are currently over 26 differently defined entities that are treated differently during answer judging. ---------- response 4 02/05/76 16.05 dave fuller uimc Incredible! Waiting with batted breath... ---------- note 259 q's ¬+ ch 02/04/76 22.51 kent unl Running late at night as I usually do during non- prime time hours when PLATO is frequently in the "yo-yo" mode, I get plenty of opportunity to watch the condense queue message in action--which leaves me with the (rela- tively unimportant) question: What determines the Priority in which lessons condense? There's obviously some priority scheme in operation. Also, is there any way to effectively re-execute the last -charset- command encountered--regardless of the circumstances? It can be very upsetting to be in the middle of a lesson and suddenly find the r/w memory of the terminal screwed up for heaven-only-knows-what reason! ---------- response 1 02/04/76 23.28 rick mfl To effectively execute the last -charset- command in your lesson, include a unit like the one below for each -charset- that might need reloading... unit chars term italics at 1525 write Reloading the italics... charset $$ clear the charset pointer charseet charsetx,italics $$ reload italics at 1725 write DONE! Press NEXT end $$ finish up the help sequence * (TERM starts a help sequence) Then, if you're using italics for the alternate character set, and they suddenly get zapped, just type TERM italics! ---------- response 2 02/05/76 00.55 blomme s The condense queue is explained in aids--try "queue" for example. ---------- response 3 02/05/76 07.55 clark lawyer Strange things have been happening to me in the queue recently: I will get the message "loading condensor" on the screen and then the terminal seems to freeze. After about 25 secs or so, I press SHIFT-STOP and reenter the queue only to find that actually I was 3 of 5. I get passed up while loading the queue! ---------- response 4 02/05/76 08.27 berger mfl If you press SHIFT-STOP to reenter the queue, it places you back in the proper position. The "loading condensEr" message means that the condensEr is being rolled back in. Wait for it - you'll be next to condense - automatically - after the condensEr is back. ---------- response 5 02/05/76 08.47 clark lawyer ok...I thought it meant that my lesson was loading the condenser ( sorry about the misspell!) ---------- response 6 05/05/76 12.25 snellen medneta The misspell isn't entirely your fault, Clark; the message you were referring to really is spelled . . .loading condensOr. --JES ---------- response 7 02/05/76 21.13 kent unl My question was not about how to reload the last character set in lessons that I'VE authored. but in other author's lessons. In my own lessons it's very easy indeed; but what about when I'm running someone else's lesson in which no such provision was made. If it isn't there already, there should be some scheme for anyone using PLATO to do this--students as well as authors. Perhaps another -term-, ...hmmmmm?.... ---------- response 8 02/05/76 23.02 daleske ames A good standard to set up is not another system's term, but a system-wide standard -term- "term" (or "terms") that will display a list of the terms available to the user. Under that list, then would be -term- "chars" (or whatever) and an explanation that it will allow the user to reload the the charset. This feature is VERY useful in loading a charset for a person in monitor mode without having to clear the charset flag, re-enter the lesson and then reload the charset. By the way, in order to save some code, it is nice to have unit char term char charset do charset press back This way unit "charset" may be used in other places. John ---------- note 261 aargh2! 02/04/76 23.08 schramm cs317 The problem that cropped up as described in note 252 wasn't solved by the response. I even tried inserting a specs okspell, but it didn't work. Any further suggestions? Also, why can't I get this thing to enlarge cyrillic letters? ---------- response 1 02/04/76 23.38 rick mfl I tried the following and it worked ok for me: unit test arrow 1010 answer ¬W≤¬X≤¬Akl7h $$ ¬W≤¬X≤¬A is font here, for readability If you keep the "-" in answer command, you'll keep getting that condense error since TUTOR doesn't like authors to use punctuation like that in answers. As for enlarging Cyrillic characters, you should investigate the "lineset" feature. In a nutshell, PLATO doesn't expand them automatically because sized writing is actually line drawings. If you and this character. in altfont, how should PLATO interpret it for a large size? A box? A tiny x? Two horizontal bars? Two vertical bars? The regular characters can be sized because the system programmers made up a set of co-ordinates for PLATO to use. ---------- response 2 02/05/76 01.33 warner iu The command answer ¬W≤¬X≤¬Dkl7h- will never work as long as the - sign is in the tag. (Once again, ¬W≤¬X≤¬D stands for the font code.) You can get around this in two ways, either exact ¬W≤¬X≤¬Dkl7h- (exact will accept - signs, but will not mark up spelling or capitalization errors), or put -=ab answer ¬W≤¬X≤¬Dkl7hab (changing the - sign to a legal group of characters before hitting the answer match.) If this is a new foreign language charset. I would suggest _most_ strongly≤≤≤≤≤__≤________ that you not put letter characters on punctuation signs! ---------- response 3 02/05/76 09.30 judy pso Silas-- your response (avoid punctuatioin like the plague) is correct TODAY, but Paul is getting very close to allowing the author to specify that "minus should be treated as a letter, and not as an arithmetic operator". Won't that be lovely?? For a -lineset- of cyrillic characters, which can then be used for -size-d writting, contact: curtin/mfl or nolen/mfl or cooper/mfl ---------- response 4 02/05/76 10.28 bailey nuc Incidently, the Cyrillic charset used my the Russian lessons at Urbana was designed to use the "-" for the so-called 'soft sign', because there are 33+ letters in the alphabet,and because of the frequency of occurance of the 'soft sign' which suggested a convenient key location. Thus far, the Russian programmers have satisfactorily worked around the problem (but not without some anguish). ---------- response 5 02/05/76 11.36 tenczar s see my response 3 to note 252 for a little bit of info as to what is going to happen soon ---------- note 267 ¬, 02/05/76 07.56 a appel uni According to aids. ¬, can be used only with writec. What do I use for this: (in place of ¬,) keytype n3¬,,¬,,¬,;¬,:¬,a¬,b¬, (This gives me a condense error) Andrew Appel ---------- response 1 02/05/76 08.21 clark lawyer keytype n1,,,.,;,:,a,b PLATO realizes that the second comma is actually the key and not the delimiter ---------- response 2 02/05/76 10.41 whansen csa Consistency may be a hobgoblin, but it is a comfortable hobgoblin. A single notation for conditional commands (including mark in command field, delimiters, and null action indicator) would be a great boon. ---------- note 275 batching 02/05/76 09.23 jmckown physio How about this: a provision for automatically running computations during non-prime time. A user would call lesson 'batch',say, and specify the locations of his code, data, etc. Then late at night the job would be automatic- ally executed on a time sharing basis with the flesh-and- -blood users that happen to be on. Something like this would be very useful for those of us involved in laboratory applications of PLATO or, of course, anyone who needs a fair amount of computation but can afford to wait if he doesn't actually have to sit at the terminal. It is also a way to use back- ground time (that now mostly serves the midnight gamers) in a way that any taxpayer would approve. Are there any serious problems with this sort of thing? ---------- response 1 02/05/76 09.32 judy pso I would think that the very biggest problem would be the undependability of the system. You'd REALLY have to write your program carefully to have it recover correctly from crashes. ---------- response 2 02/05/76 10.31 fay o Hmmmm....other than that little problem, Judy, I know of several current applications of PLATO which would be vastly improved if such a system were operational! It sounds like an excellent idea to me. ---------- response 3 02/05/76 11.11 guerra uimc I agree that this would be very useful. As for Judy's objection: the people who would use such a feature would just have to be aware that there are certain risks. If you have a "batch" program (i.e. a program that does nothing besides calculations and data manipulations) that runs for more than a few minutes, the problem of system crashes is something you have to worry about whether you're sitting in front of a terminal watching it run or whether the program is activated automatically. (continued in next response) ---------- response 4 02/05/76 11.31 guerra uimc For example, I have a batch program that takes from 5 to i5 minutes to run. The way I activate it is to sign all 10 of our terminals into the lesson and let them run. The output from the program has to be completely accurate, i.e. no data can be skipped and no data can be processed twice. It has cost me much sweat to make sure the program can always pick up where it left off, if it is interrupted either by a backout or a plain old hard crash. Thus, from my standpoint, there is nothing to lose and something to gain. I'm going to run my programs anyway. Shall I arrange to have it done in way that is convenient both for me and for the system? Or shall I continue doing it in my present way, which uses as much of the system's resources as 10 terminals can grab, as opposed to the amount the system can actually afford and feels like allocating to the task? ---------- response 5 02/05/76 12.33 frye mfl Just a question of clarification: Is "jmkeown"s suggestion one of allowing TUTOR-style code to be executed batch-style, or rather to make better use of the FORTRAN, etc. capabilities of this machine? David Frye ---------- response 6 02/05/76 13.34 jmckown physio I'm not familiar with whatever FORTRAN capabilities this system has. Where do I look? Any way batching could be accomplished would be fine, but what I had in mind was auto-execution of TUTOR. ---------- response 7 02/05/76 13.56 judy pso Just for information's sake.... jmckown/physio (physiology) is a different person from mckeown/com (accounting) It's confusing to have two such unusual and similar names both working from UI! Almost as confusing as Jim Parry (parry/arizona) and Jim Perry (perry/uw) ---------- response 8 02/05/76 14.10 michael cornell Or especially: mike/cornell (Mike Huybensz) mike b/cornell (Michael Besosa) michael/conrnell (Mike Oltz) michael/english (Gary Michael) ---------- response 9 02/05/76 14.43 jmckown physio My name is John McKown. Actually I was hoping for a dialog on the subject of batching, but it sort of went bananas there. System responses would be especially welcome (on the subject of batching, hopefully). ---------- response 10 02/05/76 15.21 guerra uimc If anyone else writes a response to this note about names that look alike, I'm going to send them a letter-bomb! (through p-notes, of course) ---------- response 11 02/05/76 16.01 dave fuller uimc It would seem that these people wanting batch privileges would be at the mercy of the internal structure of the monitor, the operating system (the present OS is marked for extinction in the not-so-distant future, as mentioned in "projects".) and file structures, which the designers of PLATO would prefer invisible. Perhaps a controller - program could be devised, that signs on terminals automatically, and re-signs them in upon system-bringup. I understand a program similar to this exists, but works mainly with non-extant (pseudo) terminals. ---------- response 12 02/05/76 16.43 rader s We do various batch jobs in-house (TUTOR-batch), but have not gotten around to defining what a user -batch would do--what impacts, what capabilities. There is of course the further possibility of general batch, but that is even farther away (aside from in-house batching which we have always done). ---------- response 13 02/05/76 21.11 koning csstaff Some other systems have what they call 'pseudo-terminals' which are artifacts of the operating system that look like terminals to the jobs running on them. Input is done by having a different job press keys on the 'terminal' and output is done by the driving job reading the output buffer. On a system I used this allowed one to do simple batch processing without much of a change to the OS. With Plato's graphics abilities, the saving of the output would probably be difficult/impossible, but the input should be feasible (since -press- already exists) and a running program can always avoid doing output by storing all results in a dataset/common. ---------- response 14 02/05/76 21.33 kent unl As a would-be plato number-cruncher, much of this discussion interests me but it also leaves me puzzled. Would't a batch job be run as a separate job from plato (under NOS) and thus be unaffected by plato crashes? Or isn't this a multiprogramming environment like, say, IBM's OS/MVT or /MFT? Anyway, my preference would be APL to FORTRAN-- much more compact and POWERFUL! (except for complex numbers) Pardon me if my ignorance of CDC hardware and program products is showing, but the only exposure I've really had with them has been through plato (which provides a ____≤≤≤≤very distorted view). In fact, I'd like to see a whole series of lessons made available describing CDC products: everything from system architecture through assemblers to how plato "works" at the systems level. ---------- response 15 02/06/76 02.58 dave fuller uimc Koning: Such facilities as the pseudo-terminals initiatable on the DEC-10 and simliar systems exist on this system. They provide the "tutor batch" service that Bob Rader mentioned. I stated I would be in favor of the type of service already performed (such as the "autosignon" of terminals without human intervention once the system comes up after reload/crash) for "real" terminals and not the "fake" ones. Kent : Perhaps, one of these days when NOS is finally implemented on the planned dual-6500 system, we'll have processing time to burn. I would personally hate to be working systems staff, and have to contend with lord knows how many batch jobs that were commited to be executed during the night. It would further limit the systems staff's "realm", to have to guarantee further OS uptime to service all the users. ---------- response 16 02/06/76 03.03 dave fuller uimc To compound the misery, the programs that would be run in batch mode would have to be developed and com- piled sometime. I do not know how many people there are like me that enjoy staying up to all hours programming, but suffice it to say that would have to be the case, since there would definitely NOT be enough time to allow compiles/test runs in the day. These are only my opinions, and do not reflect the wills of the systems staff, but I would think they are more or less universally valid... ---------- response 17 02/06/76 08.33 fay o a) The systems staff would not lose much sleep over tutor- style batching (once the software was modified to handle it). The burden would be upon the individual requestor of tutor-batch jobs to program his job to run w/o human inter- vention and w/o mods to the existing tutor structure. b) These types of jobs would probably NOT be guaranteed any different priority than ordinary, human-type users during non-prime time, They would probably be guaranteed no more cpu than the avg user and I'm sure the systems staff would not want to implement the feature if it meant not interrupting tutor-batch jobs currently running. As on other systems, the bulk of the responsibility for the job running properly, given the constraints of the system, is upon the user him/herself. ---------- response 18 02/06/76 08.57 golden s We do have the capability of running repetitive batch jobs written in TUTOR. The capability is quite limited, but I can provide it to users. Several groups already use it for daily data analyses. The system uses it for all sorts of jobs such as destroying obsolete binaries, sorting lesson useage data into the accounts, etc. ---------- response 19 02/06/76 10.23 jmckown physio From the discussion so far I infer there may be quite a few users who need batching. I'd like to send up the following trial balloon,mainly as something to be re- placed by an improvement. A TUTOR program with a name like batcher accepts jobs during the day by storen-ing and storea-ing data and code locations. A batch big board,so to speak, would pro- vide feedback on the statuses of previously submitted jobs. At night a phantom terminal signs on and executes the pro- cessing part of batcher. The 'use' command, generalized to accept a variable string as tag, lets the job appear to PLATO as part of lesson batcher. All results are stored in datasets. Protections against interruptions might be limited to detection of the interruption and starting over on the interrupted batch. I think even I could program that much protection into batcher-- the user could then forget about ---------- response 20 02/06/76 10.30 jmckown physio crashes altogether. A clever programmer could probably provide more efficient protection by requiring some simple flags be included in each batch's code. This outline probably has lots of weak points due to my basic ignorance of programming and this system. I hope some of you more qualified respondents comment. ---------- response 21 02/06/76 14.57 dave fuller uimc PLATO has the facility to initiate pseudo - terminal runs. I was referring, in the second part of my note, to REAL batch under the machine's OS, as suggested by Kent and others. ---------- response 22 02/06/76 15.35 jmckown physio Of course. I get the impression that real batching has something less than sizzling priority, and I would't presume to argue with that. If, however, a stop-gap measure could be hammered out,then some of us would sure use it. The time sharing arrangement I outlined was offer- ed in that spirit--not as an alternative to something more powerful that might come later ---------- response 23 02/07/76 17.07 barr econp I think the idea by jmckown is both practical and desirable. An auto-signon terminal(possible puesdo terman- al) would sign a student into a router. From the router the student would jumpout to lessons execute the code, return, set a flag and jumpout to the next lesson etc.. If the system crashed while executing the lesson the "student" would jumpout(restart) in the same lesson. would be up to each user how his lesson would deal with the crash. The router and the bigboard lessons should be easy to write give existing features of anything. This system would not require any managemen by operators. The only effect this system would have on current users is to transfer some primetime demand to nonprimetime. mike barr ---------- note 282 erroreport 02/05/76 10.01 grandey park At present it is very difficult to report errors in lessons to the authors because there is no way to determine the unit name. Many times a student will say, this lesson will not accept this answer but that is what the help says I should type. Sometimes there are problems with erase and inhibit erase. It would be very very nice to be able to term - unit and get the authors signon, lesson name and unit name. Or maybe better yet, when an error occurs, we could term -error and the author would automatically get a pnote containing the lesson name and unit and the name of the person to contact. How about it systems people? ---------- response 1 02/05/76 10.38 fay o (Hope you don't mind a non-system response.) The existing machinery can meet such a need. A special function key can be made active in all units (one of the -op type keys, for instance) which would do nothing to the current display but would execute an -output- into the datafile with the "baseu" and Mainu" information, along with some student info (perhaps an -outputl- showing relevant values in the bank variables). Along the same lines, the institution of the much talked about -termop- would do the same thing without tying up one of your function keys and would thus minimize the amount of existing code which would have to be re-written. Finally, the institution of an on-screen student notes type system (if it is being considered and if it ever comes about) would do nicely. ---------- response 2 02/05/76 10.42 silver ve There's no way you will get the file name of a lesson into the kind of -term- feature you describe, But it shouldn't be necessary. I think a term -zproblem- or some such to display the mainu, baseu, and perhaps the mode would be valuable for reporting errors, and give away nothing that isn't available by other means. ---------- response 3 02/05/76 10.53 ahasic med Until the systems people get moving on this, you can help the student in reporting problems by putting the following code in your "imain" unit (or do it on an "op" key if you are worried about appearance): at 101 $$ or other appropriate position write "lesson identifier"/ showa mainu This will give the student a way to identify the lesson and the main unit. It is included in all the SBMS lessons. ---------- response 4 02/05/76 11.22 fay o Silver: if you are referring to the suggestion I make in my first response when you say "there is no way to get the lesson file name with such a term feature" you are only half right. The -output- itself does not specifically record the current lesson in the datafile, but the other student datafile info will reflect the current lesson. Alternatively, the file name could be the first 10 chars of the tag to the -output-. It's really up to the way the author wants to code it. ---------- response 5 02/05/76 11.28 ld francis mtc problems in your own lessons have many, comparatively easy solutions. the larger difficulties come when you are using someone else's lessons. You'd like to have the errors fixed so your own students would cease to have trouble, but the lessons are generally not familiar to you, errors may not be reported until sometime after everyone's memory has clouded, etc. I think the original note is directed more toward the case when you cannot make coding changes or additions in order to provide the needed info. ---------- response 6 02/05/76 18.23 mcneil a uicc If you are running the course the student is in you can goto the running students display(3) to get the lesson and unit info. ---------- note 288 errtype=2 02/05/76 10.48 friedman csa Can anyone suggest why my students frequently get sent to my router with errtype=2 (which Aids says means either fatal condense error or attempt to jumpout to a second router lesson), when the lesson in question is NOT a router lesson, and no fatal condense error shows up when I try it the next day? No author is working on the lesson, either.↓ HELP!!!??? ---------- response 1 02/05/76 11.06 warner iu One suggestion: did you misspell the name of the lesson in your router? A jumpout to a lesson that does not exist results in a return with errtype=2. ---------- response 2 02/05/76 11.19 rick mfl Could we PLEASE have some variable to give better disctinctions between error types? This was men- tioned before in Note ¬$391 of 1/27/76. Could we at least have a systems response? ---------- response 3 02/05/76 11.49 white p "Not enough ECS" is a fatal condense error. ---------- response 4 02/05/76 12.31 friedman csa Well. It's probably the "not enough ecs", I guess. Thanks to those who made suggestions.↓ The request for more information than "errtype" gives is a very good one.↓ Aids says that errtype=3 means not enough memory. How does this differ from "not enough ecs"? Is Aids wrong?↓ If true, I wish Aids listed "not enough ecs" as a cause of errtype=2. It's very misleading to consider it just another condense error, since a fatal condense error usually implies something wrong with the lesson. ---------- response 5 02/05/76 12.45 rick mfl I made a test router and sent a student through. and when there wasn't enough memory for a -jumpout-, errtype was indeed 3, both with and without an -inhibit jumpchk- command. ---------- response 6 02/05/76 14.48 white p I stand corrected... Not enough ECS is a fatal condense error. but is reported to the router with a different errtype. Another possibility is "Condensor unavailable". ---------- response 7 02/05/76 16.57 friedman csa I'm disappointed...I thought that ECS was it.↓ The frequency with which these turn up makes "condenser unavailable" somewhat unlikely, but possible.↓ ---------- response 8 02/06/76 12.18 white p errtype: 0: pre-execution for jumpout command start of condense 1: execution error output buffer overflow/illegal action request 2: fatal condense errors of type other than 3 missing unit on two argument jumpout! 3: condense errors of certain types 4: exec error in finish unit 5: stop1 out of condense queue after 30 seconds. ---------- response 9 02/06/76 16.44 white p The following "condense" errors have returned errtype=3: No ECS, but still within site allocation Lesson Deleted No room in ECS for common Lesson tables full (unlikely) Site memory allocation exceeded "lesson deleted" is applicable as a "memory" error only for authors, so on the next version of Plato, this will return errtype=2, not 3. This should remove one occurance of incorrect errtype=3 errors. ---------- note 291 backups 02/05/76 11.07 berger mfl I think lesson "backups" would be much more useful if one didn't get deleted for lack of memory. ---------- response 1 02/05/76 11.25 fay o Eh-heh. Yes, it is a problem, and we have considered making the lesson of the type which registers only 1500 wds of ECS (much like Aids or the editor). We can still do this, but we are really wary of aggravating the ECS shortages by adding still another couple of thousand words to the already large ECS overhead. ---------- response 2 02/05/76 11.38 berger mfl I can understand the problem, and agree in spirit that it would indeed aggravate the ECS problem. But the lesson HAS been advertised as a "system" feature in notes, and one would therefore expect it to be always accessible. ---------- response 3 02/05/76 11.46 rader s In fact the lesson is protected on the next version of PLATO. ---------- note 299 jan flake 02/05/76 11.32 hecht ed Question--is Jan Flake still on the system? One of our elem. educ. classes wishes to use the lesson on Polya's style of questioning and needs some information about it (and an ok to use it). ---------- response 1 02/05/76 11.50 michael english She is on the Florida system. Dr. Travers or Dr. Hicks in the Education Building where you are at might be able to give you some information. ---------- note 309 health 02/05/76 12.01 hecht ed Another question--one instructor is interested in using some lessons on health occupations for her class. Who can I contact for an ok to use lessons in this area? Please send a pnote to hecht of ed. Thanks. Jim ---------- response 1 02/05/76 12.50 celia pso tebby of pso can answer questions about subject matter and who to contact about existing lessons. ---------- note 314 nitpick 02/05/76 12.32 snellen medneta I'd like to submit the following for the "nitpick of the month" award. Often, when condensing a lesson, one is treated to the message ". . .loading condensor" when, in fact, it really _should_ say, ". . .loading condens_e_r". No big thing, but we wouldn't want to spread the impression that us peeple her in plaitoeland dont no how too spel gud. --jES ---------- response 1 02/05/76 12.54 jim g reading This was the subject of a debate some time ago ( a couple of summers ago, I think ). The outcome of that debate was that since the condensor did not really "condense" the spelling was up to those who implemented this piece of computer (or is that computor? ) software. Therefore you are not assured of the nit pick award. Sorry, it was a nice try! ---------- response 2 02/05/76 16.47 rader s Yes, it was felt that we could invent a word if we wanted too. On the other hand, after much research I have concluded that one should not invent this particular word, and the normal senses of "condense" are not so different as to matter. I am now favoring the spelling "condenser" in my own work, and hope to see it propagated. ---------- response 3 02/05/76 21.43 kent unl ...I think a propagator is a large reptile which is in very little danger of becoming extinct.... ---------- response 4 02/06/76 17.54 rice nuc No. no...you're thinking of an _allegory_. ---------- response 5 02/06/76 23.52 kent unl ...No, I'm afraid you're quite mistaken. An allegory is a closely related creature most noted for its habit of wantonly ripping its prey to shreds before devouring it. ---------- response 6 02/08/76 14.49 lombardo ed NO NO, that's a professory, coming close to extinctsion in the jungles of reality! ---------- response 7 02/08/76 16.59 kent unl Everybody's confused but me! A professory is an entirely different creature--but it's true about its danger of extinction in the jungles of _reality_, since the published evidence of its existance is almost invariably _abstracted_! ---------- note 320 inspect? 02/05/76 13.26 s zweig iumed Would it be too hard to place in the editor on the top line, (where block name and space are located) the word/s INSPECT/INSPECT ONLY if in fact the author _is_ editing the lesson in Inspect mode?? T'would surely be a minor thing, but a boon to forgetful folk. Are you there,_dpf_/ Ʃteve ---------- response 1 02/06/76 09.43 p cohen med Just to provide the usual system answer, a lot of things might be nice to have in the editor. However, each addition will further slow an already sluggish editor, so a compromise must be made. I agree that an "INSPECT ONLY" header might be nice to have, but I trust that you can work around this nuisance. Just be thankful for what we have. --paul ---------- resposne 2 02/06/76 10.02 guerra uimc I really don't think it's worth it. I think that would make the top of the line edit page irritatingly crowded. And furthermore, I can think of only a few cases where you would get to the line edit page without knowing you're in INSPECT ONLY mode, so it would mean we would have to put up with an irritatingly redundant INSPECT ONLY message on the line edit page in the 98: case, as against the 2: of the time when it is useful. After all, if you're in INSPECT ONLY mode and you don't know it, you'll find out after a few keypresses anyway. ---------- response 3 02/06/76 10.38 don emerick mtc not to mention the burned out dots at the top of the panel problem. ---------- note 324 when eos? 02/05/76 14.03 hagerman ames Does anyone have an estimate of when the core will be (mostly) repaired? It would be nice to run on full ECS again...... Sauron ---------- response 1 02/05/76 14.56 fay 0 HUH? Unless PLATO has taken to playing tricks on me, like flipping various toggle switches all by its lonesome, we are currently on full ECS. What you probably are ex- periencing at the moment (and undoubdtedly will continue to experience during the day time) is the result of the heavy demand being made on our ECS by all our different users. ---------- response 2 02/06/76 08.51 avner s If you are referring to the high frequency of ECS errors resulting from the stresses of the shut-down over the Xmas-New Year holidays, past experience suggests that the worst is well over but that we will still be seeing some effects for the next 30-50 days. Incidently, over 70: of system unreliability during January resulted from these ECS problems. Al Avner ---------- note 325 specs tag 02/05/76 14.07 p cohen med As suggested by Paul Tenczar, I am writing this note to learn if there is any desire for the following proposed TUTOR feature. Presently, there are -specs- tags "okextra" for allowing ALL extra words in an -answer- command, and "okxvocab" for extra words in a -concept- command which are NOT included in the preceding -vocab/s-. Is there any interest in a new tag for allowing extra words in a -concept- command which ARE included in the preceding -vocab/s-? Discussion? --paul ---------- response 1 02/05/76 14.25 michael english Sweet Mother of Mercy! I have prayed for this feature for months. Last time I talked to Paul about this he didn't know if it were possible, and it would be a year before it could be done even if it were. At least one other systems person said the entire theory of the -vocabs/concept- structure made it impossible because the hashing procedure employed at condense time threw away information needed to decide what words were in the -concept- and what words, therefore, were not, and could be used to ignore extra intra-vocabs words not relevant to the particular -concept- being processed at any one moment. With a -specs okextra- that worked on -concept- you could do some real natural language processing! ---------- response 2 02/05/76 14.33 stan smith chem The present system is very difficult to use because of unexpected student use of words which are in the vocab list. To get around this, in some lessons, I have had to use a different vocab list for every question. It would be very helpful if we could in addition specify words which can not be in the response. ---------- response 3 02/05/76 15.48 silver ve What Stan and Michael says, goes for us, too. Otherwise, we may end up using two vocabs at the SAME arrow in some instances! Is there a real possibility of doing this? ---------- response 4 02/05/76 16.57 fritz ames It would help immensely in the CS project I'm doing for next quarter! (As if it could be implemented by then...) Please, please? ---------- response 5 02/06/76 07.54 mont csa another vote for! ---------- response 6 02/06/76 09.24 danielson csa I agree with all the previous responses. In my last use of concept and vocabs, I discovered that 37: of the misunderstood student inputs were considered wrong simply because they contained a _single_ extra vocab keyword. A feature such as suggested would improve things very much. Ron ---------- response 7 02/06/76 09.38 bonetti ced This new specs would be greatly improve some of the lessons I am currently working on. One more vote for. ---------- response 8 02/06/76 12.42 tenczar s thanks for all the input...while it is currently impossible to do because of the way the concept command is currently implemented, I will try to re-implement the concept command to allow this much needed feature in it. Since it might take awhile, please find your own solutions in the meanwhile. It is on my list of things to do. ---------- note 326 site 02/05/76 14.09 mcneil a uicc Computers are ideal for doing repetitive tasks. One task that has become repetitive and time consuming recently is monitoring the site ECS/users and adjusting what is going on at the site. The present ECS manager does not serve our needs here at uicc. Only a cont. monitoring of site enables us to keep our sanity. I'd like to propose a site-manager -- a lesson that would be in the control of the site director(s), written by the site director(s), and controling the site. -cont- ---------- response 1 02/05/76 14.09 mcneil a uicc This lesson would need the following features: 1) ability to monitor what lesson/course/ecs/user type is at each terminal at a site. 2) send messages to any terminal in the site. 3) press STOP1 at any terminal in the site. Other handy features would be: 4) ability to approve/reject attempts to condense/jumpout to lessons. 5) catch users starting to sign on. -cont- ---------- response 2 02/05/76 14.10 mcneil a uicc The basic operation of this lesson follows: 1) At class-changing time datain next period's scheduled courses/user priorities. 2) On user signon, check if user is in scheduled course if so let in, else follow site specific procedure. An example could be: if non-scheduled then check ecs for room, check if any open terminals left (we get non-sched users hogging terminals often). 3) When user wants to jumpout, check if scheduled, check if ecs available (if not delete non scheduled users). 4) when ecs left goes negative delete some non scheduled users. -cont- ---------- response 3 02/05/76 14.11 mcneil a uicc When ever a deletion is necessary a message would be sent to the terminal warning of impending deletion. I also believe that these lessons should be only quasi- system lessons--they should be under the control of the site directors. AL9000 ps I believe I could write a manager in 1500 words. ---------- response 4 02/05/76 14.34 deiss phar I pushed for a similar idea about 18 months ago. I called it a site router. It has not been well received by anyone ....... systems, pso, users, etc. Unless you have better luck than I, you'll have to write a router that routes for _all_ courses that use your site. Good luck anyway. ---------- response 5 02/05/76 14.54 mcneil a uicc 18 months ago we didn't have as many students ¬+ courses to cope with. The situation has reached the agravating level here at uicc (and from what I hear also at 'mfl'). ---------- response 6 02/05/76 15.39 weible german It certainly _is_ annoying here; especially when one unsched- uled physics student can tie up 40: of our base allotment by going into a certain mammoth physics lesson (which shall remain nameless). Before he is discovered and booted out, any number of legitimately scheduled students may have become so frustrated at their inability to get into a lesson that they have already gone for the day. A site director lesson would be a real help I think for any medium to small sized site. ---------- response 7 02/05/76 15.52 silver ve Yup. This has gotta be a pretty general problem. I think that cs guy came over here after he left mfl... ---------- response 8 02/05/76 16.44 simons cs196 i third(or fourth ?) the motion... a automated system manager would beat the heck out of the operator always locking out all the other nonscheduled students. this way, we would only be kicked off WHEN necessary, and not until. many a time i have wanted to edit a lesson, but couldn't since 165 was restricted. ---------- response 9 02/05/76 16.56 simons cs196 also... it would be nice for the system manager to keep one terminal open for a possible scheduled student. when someone tries to signin on that one, he is allowed on only if he is on the schedule. if he is, then some non-scheduled user is informed that he will have to leave to make room for the class. this way no one is going to the operator complaining that there isn't a terminal. you can determine the person to kick off, by ecs usuage.lesson in,etc. i have also seen cases of where scheduled students cannot get into their lessons because of the ecs. many do not know(or don't care) that they could complain to the operator about it. with the manager,it wouldn't happen as it would knock off nonschedulers until there was ecs. ---------- response 10 02/05/76 20.16 stan smith chem It is clear that managing the ECS at a site takes a lot of time that could be better spent doing other things. The basic scheme which I use is simple enough that it might be possible to have it done automatically. Basically if there is plenty of ECS, let all students work. When the difference between current ECS and used gets to less than n, a parameter set by site director, the temporary site restrictions are turned on. When this ¬Δ gets to be less than nˇ1 (nˇ1o200). "square", whose name is (gasp!) micro1, is a function key just like COPY. Square1, CR, sup, sub, etc. do not have names, because they are not function keys, but rather keys which insert an octal code into the internal string. square1 ONLY inserts an o76 into the internal character string, whereas MICRO causes a table look-up which _may_ or may not cause an o76 (usually + other characters) to enter the internal character string. It is useful to understand these distinctions, especially since CR ¬A¬D TAB and MICRO ¬A¬D FONT are typed using the same keys, but are different _types_ of keys. If square1, sub, CR, etc. were given names, it would become doubly important to understand this distinction. ---------- response 5 02/06/76 10.40 judy pso in passing.... The name "micro1" for the square key was removed from AIDS when the function keys were shifted around. (Do you remember trying to get used to the new location of ERASE? I think that long ago, "square" WAS micro1 and "square1" was font1...?) Clearly, "micro1" is not an appropriate name for square and Blomme said he planned to remove it. But one can't just _remove_ features from TUTOR, so we merely took it off the list. Probably square should now be named square. I think you should regard "micro1" as being an unstable name.... ---------- response 6 02/06/76 11.12 michael english I hope micro1 isn't removed without first creating a substi- tute. I don't know how to make it a -jkey- otherwise. ---------- response 7 02/06/76 18.00 blomme s Summary: some of the names are a mess; you know it and we know it, but there are several snarls involved that make the amount of work necessary far greater than that needed simply to add a new key name. It will take (1) a determined person (2) several days of re-doing the whole business for changes to get done...in all likelihood this may not happen for quite a while, but the naming problems are known and noted. ---------- note 367 records 02/05/76 21.59 galcher uimatha It might be nice in the next version of the course records editor if the following were given with the information available in the "last edited by" and the course information page: last edited by: editor course: mycourse date: 02/05/76 time: 19.52.44 last action: add student And of course there could be some fairly informative messages for that last section, such as: add/delete student change name deleted notes changed course info, etc... How about it folks? ---------- response 1 02/06/76 08.20 b sherwood phys Will keep in mind. ---------- response 2 02/06/76 08.21 marty smith mtc i'd rather have my own course editor. this one is not what i need to run students. i mean, different sites use plato in a different way; so why should we be stuck with what seems best for somebody else? ---------- response 3 02/06/76 08.50 michael english What fate the -writer- command, which would allow such? ---------- response 4 02/06/76 11.09 d sherwood espo Also, it would be nice if one could see at least part of the data page of lessons, so one would know where to send one's pnotes of praise/suggestions. ---------- response 5 02/06/76 13.54 nate iumusic While we are/were on the -writer- command: Would it be possible to find out approx, how long we have to wait before we can at least write into student and router variables? It gets a little difficult trying to explain to a proff that this router variable, and not this one, and you must edit it as an integer, does this powerful little thing to this router lesson. Would be nice if I could inclued it with my lesson-editor and not have to bother explaining how to use that clumsy varible editor in course-records. Nate ---------- response 6 02/06/76 14.18 marty smith mtc i would settle for that for the time being. ---------- response 7 02/06/76 20.08 b sherwood s It is intended to implement some form of writing into student records. No time scale -- sorry. By the way, I would point out that setting bits in rvars even via some curriculum editor program isn't going to work too well if someone at another location would like to use your curriculum..... ---------- response 8 02/07/76 01.26 nate iumusic HUH? All I'm using the router vars for is for a router-pointer to determine what lessons that particular student is to get and what he is not. It just affects the router and nothing else. I don't understand that last response at all, Bill. Maybe a pnote explaining? Nate ps. granted, it is trivial, but I am curious folks. ---------- note 368 template 02/05/76 22.01 andersen s Template curriculum option in records should now be working correctly. ---------- note 380 lost/found 02/06/76 08.34 greve o We have a lost calculator in the operator's office (room166 CERL). It is a Texas Instruments model SR10. It has social security number scatched on it's side. To claim it have an ID with you. Also there is a check here that was lost this morning. It is made out to Cal Finkel. Thank you ---------- response 1 02/06/76 09.34 carter comm Whoever the owner is, if you would like to sell it, send a pnote to trollip of arlc. He is starting a calculator museum and would like one of these antique models. ---------- note 385 extnd H 02/06/76 08.51 buz phar Would it be feasible to extend the "H" option in the editor to show "n" lines...i.e. Hs would show 5 lines...or at least to the end of a continued command, like a calc, or write. OR possibly even the option (like the "L" option) to dis- play the entire contents of the block in the "H" format? Low on priority list...but a somewhat useful feature. *buz* ---------- response 1 02/06/76 09.08 mike cornell I hate to be rude, but I cannot see any real need for this sort of thing. In any event, it would probably require lots more programming, and look sloppier. What might be a good idea would be NEXT1 to see the next line, and BACK1 to see the previous line, but they are low priority ideas anyhow. Mike Huybensz ---------- response 2 02/06/76 11.03 judy pso One line of "H" information is all I can absorb at once, but I sure would appreciate being able to go forwards (I don't even care about backwards) without having to re-request the H option. ---------- response 3 02/06/76 11.11 mike b cornell I agree with Mike and Judy. This should be fairly simple to implement. M¬ ike≤≤≤≤ˇB ---------- response 4 02/06/76 12.49 buz phar That would be good...I hadn't thought of that. It's just a bother to find a hidden character in a calc statement when said calc is 12 or 13 lines long, by doing H1...H1...H1............ad nauseum [sp.] *buz* ---------- response 5 02/06/76 13.49 tenczar s the "going forward one line in the H-option" has already been put on the work list of -p- programming. thanks for your concern. ---------- response 6 02/08/76 19.44 marquart ee How about an indicator that points to the end of the command field (just a simple ¬X above it). This would make it simpler to check out certain errors like accidentally starting a tag in the command field. Also, if the number of characters in the line could be displayed, it would be helpful. This information must exist already, and only needs to be displayed. ---------- note 387 diskaccess 02/06/76 09.04 maggs law For lesson design purposes, I would like to know what the systems programmers consider to be the upper limit on the reasonable number of disk accesses per student terminal hour that a lesson should be allowed to have. The limits on ecs and TIPS are pretty well defined, but I don't recall information on disk access limits. Since in some circumstances there is a clear tradeoff between ecs use and disk access, I would like to know which way to go in plannning future lessons. ---------- response 1 02/06/76 10.07 b sherwood s Take a look at the system limits notes of 9/24/74. Go to system feature notes and ask for that date. ---------- note 395 no.191 02/06/76 10.03 saltan uicc Someone please answer no.191...thanks ---------- response 1 02/06/76 11.13 judy pso Contrary to frequent appearances, the audience of NOTES does not have universal knowledge. Perhaps nobody _has_ an answer to your question! Maybe you didn't ask the right question. I never heard of DRNET, but I wondered if you were asking about routines which print through an external hookup to the terminal? Or were you asking about a network drawing program developed by a graduate student in EE? (Ask tebby/pso about that one.) ---------- response 2 02/06/76 12.30 harkrader o I've never heard of DRNET, either. But perhaps Doug Brown can shed some light on your woes, if he hasn't been in touch with you already, He has been working on a very complex circuit analysis program, which I believe is in an operational state. His signon is 'doug brown' of 'p'. Al ---------- note 404 limits ii 02/06/76 10.51 fay o The preceding discussion (public note 387) and the "limits" discussion (sys notes, 9/24/74) have brought to mind what is currently a fairly acute problem. (I recommend users in general review the "limits" discussion of 9/24/74...it is illuminating.) This note assumes the reader has read the above references. Lesson authors and the instructors responsible for directing student use of PLATO curricular materials seem to have lost sight of the fact that efficient use of the PLATO system cannot be made when a number of students are assigned a number of different lessons (intentionally or by default because of self-pacing). This is one reason it _seems_ many times that we must be running on half-ECS. To para-phrase the "limits" discussion referenced above "completely in- dividualized instruction cannot be promised/projected at this time". (-LAB-) ---------- response 1 02/06/76 10.59 fay o As many site directors are currently painfully aware, students who must complete a large number of PLATO assign- ments on their own time (ie, w/o reserved terminals) fre- quently run into the problem of lack of ECS. The reason, of course, is the rate at which ECS is used up when a large number of different lessons are in use. This is very frustrating to the student and no less frustrating to the site director who tries to accommodate them. My recommendations to all instructors are: A) As far as reasonably possible, attempt to arrange the use of the same lesson by several students at the same time. It is quite easy with the system router to make up a new index weekly with only one or two lessons available to the stu- dents. B) As far as reasonably possible, reserve terminals for use by your students and exhort them to make good use of this reserved time. (-LAB-) ---------- response 2 02/06/76 11.04 fay o There is no reason that course directors should be ignorant of the limits on system resources (these are known and are available on-line). In many cases, course directors can (and should) make a greater effort to use the available re- sources in a more efficient manner, Your students will be less frustrated and will get more out of their PLATO curriculum. Tim Fay ---------- note 406 Yawnnnnnn 02/06/76 10.59 chen mfl Why would anyone want to sign in and program at the between 2400 hours and 500 hours????? These people who can stay awake are crazy......please respond (especially the people who usually does this....) Yawnnnnnnnnnnnnnnnnnnnnnnnnnnnn ---------- response 1 02/06/76 11.13 d sherwood espo Why do you ask? If people want to be night owls, that's their business. ---------- response 2 02/06/76 11.17 fritz ames As one of those "who usually does this": We here at Ames have four (4 countem 4) terminals on which to work. When you have approximately 8 different classes running students on these four terminals, it is often quite difficult for an author to get much time to work during the day. Consequently, you work when you can -- late at night/ early in the morning. Besides, it's the only time we can have extended work sessions (upwards of 18 hrs long for some of us hard-cores!). Is this ok with you, sir? ---------- response 3 02/06/76 12.17 michael matha My list: No interruptions (common, and not unwelcome during the day) Quiet Abundant ECS Abundant CPU Shorter condense queues And because of the above, increased efficiency and greater concentration. Besides, some of us are insomniacs. Oops--extra added occasional benefit-- All those features marked: *NOT YET AVAILABLE DURING PRIME TIME* ---------- response 4 02/06/76 13.11 parrello mfl Perhaps because foreign language instructors require their programmers to perform superhuman tasks. the mfl tag team "The difficult is due today, the impossible was assigned yesterday." ---------- response 5 02/06/76 14.45 dave fuller uimc (Do I hear a touch of sad violin music in the background of that last response?) ---------- response 6 02/06/76 16.13 bonnie matha ...and the bars are closed. ...and the sun isn't out. Why would anyone want to work during the DAY? YUK! ---------- response 7 02/06/76 16.33 alan reading Bonnie's got it. ---------- response 8 02/06/76 16.50 berger mfl I hope it's not contagious.... ---------- response 9 02/07/76 00.02 kent unl I hope I can talk Bonnie into sharing some of it with me....... ---------- response 10 02/07/76 01.12 tricorn hebrew What I don't understand is those people who are getting up when I am going to bed...or the people who schedule 9 o'clock classes...or banks and stores that close an hour after I get up...I am also doing the early birds on this system a favor by not using some of their resources, and am rewarded for it by them not using up mine...besides, it's morning somewhere Δ ---------- response 11 02/07/76 09.12 daleske ames When, during the day an author cannot condense a lesson without being dumped and each of the preceeding reasons apply, the night is much more than a piece of sky that twinkles with that clear, forceful star "Orion". To walk abroad in the early morn, when the system's just gone down, I see the sunrise and hear the birds. What more could one ask for? ---------- response 12 02/07/76 14.24 parrello uimatha ... Some way to shut them **** birds up when I'm trying to get some sleep. ---------- response 13 02/07/76 22.39 kent unl ....And maybe a whole new PLATO system to double existing capacity, if that's what it takes.... ---------- response 14 02/08/76 14.58 lombardo ed I just had breakfast and I'm gonna go back to be until a reasonable hour like 2200 or so. You people are crazy for being up at such an unreasonable hour, if God had wanted man to be up during the sunhours he would have built sun- glasses into our eyes. Yea night people, not to be confused with ladies of the night! ---------- response 15 02/08/76 16.18 ciciora cerl Crion is a constellation, not a star. ---------- response 16 02/09/76 07.56 daleske ames I meant to say: forceful star of "Orion" (Betelguese) I know that Orion is a constellation cause that is my very favorite of constellations in the sky! John ---------- note 415 address? 02/06/76 11.25 hagerman ames Does anyone know the mailing address for Larry Niven, the SF author? ---------- response 1 02/06/76 12.20 roper siu You can probably contact him via Ballanitne Books. Quetzal ---------- note 423 dot size? 02/06/76 12.53 buchin iu just out of curiosity........ does anyone out there know the dimensions of a single dot? and, while were at it, what is the space between them? red bush ---------- response 1 02/06/76 13.23 shafer m The dimension of a single cell is 10-12 mils. Cell spacing is .0167 inch, center to center. This is on a standard 512-60 unit. Hope this helps. GLS ---------- response 2 02/06/76 13.39 warner iu Easier to remember that the distance between dot centers is 1/64 inch. ---------- response 3 02/06/76 14.00 golden s No, that's 1/60 inch. ---------- response 4 02/06/76 14.05 nagel vu About .42333 mm for you metric fans...... ---------- response 5 02/06/76 15.10 wood medsiuc Obvious pun category: All this over a silly half a millimeter? ---------- response 6 02/07/76 00.05 kent unl Of course! The other half regenerated into a very serious millimeter. ---------- note 434 notes ? 02/06/76 14.30 nudelman css I seem to remember hearing about this problem before, but anyway... I just signed on and had a NOTES flag on the author mode page. Pressing 'N', 'd' resulted in the "You do not have any notes right now" page. Is this a problem with the new groupnotes? * Mark Nudelman * ---------- response 1 02/06/76 14.42 dave fuller uimc The last time you were on the system, probably you received, and deleted some notes, leaving you with none. The system then crashed, and your records were not updated to reflect the "off" notes flag. ---------- note 444 charseterr 02/06/76 17.05 bereiter matha ERROR in CHARSET! lesson=marilyn.block=pizza error word=000006400032000150000 ---------- response 1 02/06/76 19.56 b sherwood s Answered by pnote. ---------- note 445 over-write 02/06/76 17.09 koval uicc Did any of you fabulous people out there in platoland ever try to rewrite a calcc that spanned several lines??? If, per chance, you have, you may have noticed that when the arrow shows up for the corrected line-o-type, it is placed on the _very_ next line after the calcc command. This would be all fine and dandy except for the fact that whatever you type at the arrow will be obscured by the second line of the calcc. I dont mean to be picky or anything, its just that it gets awful hard to edit a line if you cant even read it. I would be eternally grateful if this horrendous problem were to be fixed. truely aggrevated Jed Tull ---------- response 1 02/06/76 19.16 frye mfl A cheering note -- it's been griped, complained, muttered, and cursed over ever since I can remember -- and that goes back some 2.7 years.... Dave ---------- response 2 02/06/76 20.05 b sherwood s Huh? When CREATING your long -calcc-, why not press NEXT every 50 characters or so, putting additional arguments on later lines? ---------- response 3 02/06/76 21.57 warner iu Meanwhile, press the white switch on your terminal before you type in your new -calcc-. This works fine but takes a lot of memory space -- yours! ---------- response 4 02/07/76 09.17 daleske ames This problem is even much worse for those of us that tend to use the ever-much-faster C¬≤vR to reach the return point set by the arrow. To have a list of index-letters imple- mented with the carriage returns and then try to edit it and following lines is very difficult. Silas's method should be written up with the rest of his super ideas and kept in aids. They tend to help greatly. John ---------- response 5 02/07/76 10.43 b sherwood s How is the CR faster? ---------- response 6 02/07/76 16.05 warner iu It is not faster, but it is _much_ smaller in source. The statement ↓a↓b↓c↓d↓e↓f takes up only 2 words, while a b c d e f takes up six words! ---------- note 448 partial 02/06/76 19.03 d zweig iu Can we possible speed up the process of partialling out blocks for condensing? I would suggest making all the info now given -helpop-able. Also, activate stop1 directly after changeing the "markers" so reploting wouldn't be necessary. ---------- response 1 02/06/76 23.36 kemp uicc I don't think the auto stop1 idea is so good. What if you have a 2-part lesson, and all blocks are condensable, and you just want the 1-a and 2-g blocks condensed? There is no way to control what blocks in a different part of a lesson get condense flags, just the all-on, all-off isn't going to work here. ---------- response 2 02/06/76 23.55 d zweig iu I didn't mean forcing a stop1, but rather make it so that one could press stop1 immediately after deciding what one wanted to be condensed. ---------- response 3 02/07/76 12.50 perry uw Perhaps pressing "P" should take you to a page and arrow structure like the page in terminet where you select which blocks to print. This would take longer to plot and to select, but you could see all blocks at once. Alternatively, it could made available on some other key, leaving the present structure pretty much as is. ---------- response 4 02/07/76 15.17 golden s I agree, a way to work on all parts at once and elimination of the help would be better. ---------- response 5 02/07/76 16.17 k gorey research I don't think that working on all parts at once is necessary or desirable, unless as an option. Many times I want to partial out all blocks except one or two, and I do not want to have to wait forever to see my whole lesson printed on the screen as in multiple block copy. I do favor the elimination of the help information except as an option, and oppose the full screen erase done at the end of marking the blocks. Kevin ---------- note 456 sick print 02/06/76 23.49 george tech There is at least one obvious error in the PLATO IV print routine. Is there any chance that someone may direct their attention to this problem, or shall i resign myself to looking for units that don't exist and symbols that look like they were constructed from a hundred randu commands? (Belive me, they aren't MY units or MY symbols. Anybody lose any?) george f. traynor ---------- response 1 02/07/76 01.38 midden p Try not to use implied multiplication. If you feel that you need someone to respond to you, try "doug brown" of "p". ---------- response 2 02/07/76 20.19 george tech The print problem, which plagues me, is not the result of implied multiplication. Thanks for the reference, Marshall. george ---------- note 461 psychiatry 02/07/76 00.15 kent unl I'm going to be giving a demonstration of PLATO to the director, a psychiatrist, of our regional mental health center and he's the sort of person who's never heard of using computers for anything but "standard d-p jobs": pay- roll, inventory, etc. So I was wondering if there are any good lessons that someone might recommend (besides the ones available through the regular sample lesson routers like "sample") which would demonstrate the value of PLATO in medicine and/or psychiatry--perhaps even for psychotherepeutic purposes! ---------- response 1 02/07/76 07.00 mike cornell Talk about psychotherapy, 'conquest' could drive anybody same..... ---------- response 2 02/07/76 17.22 hody med try the health science network catalog lesson: "mclcat" and be sure to test out yourself each option you want to demo ahead of time or contact a member of the health sciences author group for specific recommendations... ---------- response 3 02/07/76 17.34 bt williams rhrc Response via pnote ---------- note 469 GNOTE 02/07/76 11.33 woolley p Group notes files may now be created through accounts. ---------- response 1 02/07/76 13.32 michael english Thank you, Dave. ---------- response 2 02/07/76 13.36 golden pso Announcing psonotes! psonotes is a groupnotes file read regularly by all pso staff members. Please send any questions, comments, or suggestions concerning pso services to this notes file. If your contribution is a question or request for some help, we will get back to you via pnote or somehow. ---------- note 475 microbug 02/07/76 14.36 a appel uni Problem with system???????: In my lesson, I used microtable cccp,rmicro I changed this, and am now using rtype,rtmicro However, after condensing many times, I still get cccp,rmicro. I do not think the problem is with the tutor code, etc. What gives? ---------- response 1 02/07/76 16.34 white p You had -force font,micro-, and the fonted part of both micro tables the same, so you couldn't tell the difference between them. ---------- response 2 02/08/76 19.19 maggs law To amplify, the original mico table, which I wrote, was to be used largely with "force font." In such cases if, for instance the student types an "a", the computer will look in the fonted part of the microtable opposite "font-a". ---------- note 478 notesnotes 02/07/76 16.13 woolley p Just a reminder about group notes file 'notesnotes'. Currently there is a discussion about notes-editing privileges for multiple signons. ---------- note 481 wherzit 02/07/76 17.02 krol mxc anyone know where the game of sim is hiding? i played it a year ago and forgot where its at its an angular type game you play against plato ---------- note 485 storen 02/07/76 20.07 thomasson music The *storen* command doesn't work for negative numbers.↓ For example, if you type -1 you get -111. If you type -4 you get -114, etc.↓ I hope it is just this version of Plato.....↓ Jim Thomasson ---------- response 1 02/08/76 11.45 alan reading Is the tag of your storen an unsigned segment? ---------- response 2 02/08/76 12.08 parrello uimatha Not likely. It would be very hard to display an unsigned segment and get -111. ---------- response 3 02/08/76 12.11 alan reading oops ---------- response 4 02/08/76 17.40 tenczar s thanks for pointing out this error...fix should be on next deadstart tape ---------- note 488 notesblitz 02/07/76 23.07 hinton ssu Has anyone else got the feeling that the new NOTES system results in a blizzard of notes that cannot be kept up with, or is it just that I haven't been able to sign on as much as usual the past 2 weeks ? Have we developed communication to the point where understanding is difficult ??? ndh ---------- response 1 02/07/76 23.24 artman uimatha It does seem to me that the number of notes has increased greatly, but it may be that I didn't get into help notes as much as I used to think I did. Is anyone ambitious enough to go back and count for some averages? too lazy to do it myself, eric artman ---------- response 2 02/08/76 02.04 anderer ee While I too am too lazy to do an accurate count, I have th feeling the number of notes has increased significantly in the last week. Why, I know not. (Of course, the reason may be notes like these, followed by replies like these.......) dg ---------- response 3 02/08/76 03.16 kent unl The "notes blizzard" does exist! I guess people will just have to be selective about what they choose to read..... ---------- response 4 02/08/76 06.50 fay rtv362 I suggest that we make available an extra 10 or 20 char- acters for giving titles to these notes. This would allow much, much more descriptive titles and allow the notes read- er to be much more selective in which notes to peruse and which to ignore altogether. How 'bout it, systems? ---------- response 5 02/08/76 15.02 lombardo ed I second that idea... ---------- response 6 02/08/76 19.21 maggs law A third vote--newspaper headlines have served this purpose very well for years--they suggest a scheme we cannot easily improve upon. ---------- response 7 02/08/76 20.30 mcneil a uicc Thats a great idea, I feel buried by just a day's backlog of notes. ---------- response 8 02/09/76 10.34 railing iu It is not so much extra notes, as it is the combination of notefiles. As one who used to read all of both note- files, There seems to be less notes to read than there were before. So the blizzard is really more of a dry spell than a snow storm. Malkin ---------- note 490 microgripe 02/08/76 00.31 fritz ames Is there any reason why you can't have both characters and function keys in a micro assigned to the same key? What i'm trying to do is initiate judging if a person hits ', set a flag that he has pressed it, and -judge exit- so he can finish typing. The problem is, if I map the ' into some jkey'ed function key, the ' never appears in the judging copy, and i can't detect where the person typed '. If i could assign "'", funct. key to the ', it would work fine. Is there any solution other than storea word,jcount move "'",9,word,jcount+1,2 judge continue loada word,jcount+2 ok judge exit (urk -- not even considering getting the correct *where*!) PS: I could use -search- once the word is completed, but I thought I'd try to find a faster way. ---------- response 1 02/08/76 06.50 koning csstaff I guess the main problem is that the current format of a micro table does not readily allow for that. If the micro contains normal characters, they are left justified in the micro table. If it contains a function key, it is right justified in the slot (try the LAB option in the micro editor). If the slot is empty, it contains 000. How would you combine the two things and still have an easy to decode format? A possibility would be: 0010200000000000202 for 'ab' NEXT. This would be more of a pain to decode so you don't really want to do that for each keypress unless there is a real need. ---------- response 2 02/08/76 19.59 marquart ee I have had need of the "regular-key then function-key" type micro also. I think it would be a simple way to do many things that I curently do with an incredible pile of -calcs- and loops. --------- note 493 AIDS 02/08/76 02.40 ian conn which pso do we send complaints about AIDS/ should say somewhere, like on title page... anyway...about write up on -search- no mention is made of the character limit in no search [for you interested viewers, it is 15000-10(c-1), where c is starting search char] life is rough, ian ps. speaking of which, how come S-buffer doesn't work in these notes? ---------- response 1 02/08/76 02.45 fritz ames 1) Comments about AIDS should be putin "psonotes". Hit "c" from the main notes page, then hit HELP. 2) There is no inherent limit in search. However. there is a limit of 1500 words of common that can be in central memory at once. Since 1500 words contains 15000 chars, trying to search for more than 15000 chars (starting from nc1, char 1) goes outside the range of your common and gives you a bounds error. 3) The Save buffer DOES work in here. common myra,stats,322 storage 322 comload nc323,1,322 stoload nc1,1,322 These lines were just Saved from my lesson and IS'd in here ---------- response 2 02/08/76 15.46 ian conn right...thanx for the info on AIDS reports... the problem with the save buffer seems to have fixed itself...i kept saving my lines, then returning and when i hit 'IS', nothing happened other than a screen replot...now its working. ---------- note 498 zap??? 02/08/76 07.37 koning csstaff I was just reading NOTES a few minutes ago when I suddenly found myself facing the 'Press NEXT to begin' page. There was no transition through the author mode page that is so typical of a STOP1 pressing type backout. No message on the bottom either. I signed back on and tried to write a note. After 5 seconds: zap! And once again before I even got a chance to enter notes. What is happening to me? Is it (1) a system bug, (2) a hardware error in the terminal, (3) someone playing around with privileged system features? I am getting worried a bit....... ¬↑ˇPaul Koning ---------- note 499 editouch? 02/08/76 09.44 fay rtv362 I just noticed this morning (as did other users in the room with me) that the touch panel is activated every time you enter a block to edit it. Thus: PLACE: TOUCH PANEL ON: block directory no block (show mode) yes block (replace) no This is (at present) repeatable at the other touch panel- equipped terminals in the room. What gives? It is a minor problem (annoying when one wants to point to lines in the editor). (Don't know what other circumstances will activate panel... those are just the ones I noticed.) ---------- response 1 02/08/76 12.46 t little research The reason for the new times in which the touch panel is turned on is that there has been a change to the pause command. It now looks and see if you are allowing the touch panel to be touched(i.e. pause keys=all) and if so then sends an enable even if there was't a previous enable in the unit. All the places you mentioned I believe use a pause structure of some sorts. Todd ---------- response 2 02/08/76 15.45 jones mcl pause keys=all? I thought I always had to explicitly enable to turn on the touch panel, if keys=all turns it on, then we should have: pause keys=allonthekeyboard $$the old keys = all that won't enable pause keys=allbutext $$so touch is active but not ext. this should enable? pause keys=allbuttouch $$keyset and ext, no touch. this should also enable. ---------- response 3 02/08/76 16.31 andersen s Should be fixed now. ---------- response 4 02/08/76 20.33 mcneil a uicc I wish notes like this would go away when the error is fixed. ---------- note 500 record err 02/08/76 11.19 shirer vu Execution error (index err) in course records while attempting to template all students in course into a new module. error in lesson records/twrite regular calc (18th command) ---------- response 1 02/08/76 20.54 b sherwood s Thanks. I believe this has now been fixed. ---------- note 506 octal err 02/08/76 13.26 clark lawyer The code below does not produce a condense error even though the octal number has 21 digits(?).. Is this the way it is supposed to work?? I almost freaked out trying to figure out why the code failed occasionally. calc n1‎(n1$mask$c007777777777777777777)$union$(055$cls$54) Repeatable in other lessons J. Clark Kelso ---------- response 1 02/08/76 13.53 woolley p You can add as many 0's to the front of the number as you want. It doesn't affect the value of the number and is not considered an error. ---------- response 2 02/08/76 14.43 clark lawyer hmmmmm...Why?? It seems to me it should be flagged as a condense error... ---------- response 3 02/08/76 16.24 k mast p Leading 0's are ignored. This is quite typical not only for computers, but for people as well. ---------- response 4 02/08/76 18.19 al mflu Use "rmask(54)" ---------- response 5 02/08/76 20.37 clark lawyer aha! a new command I did't know about...thanks greatly! ---------- note 508 ntries ‎ 0 02/08/76 14.01 nudelman css If a student goes to a help sequence on an arrow, when he returns to the arrow, the system variable ntries will be reinitialized to 0. Is there any way to prevent this short of storing ntries in a student variable and adding it to the new value of ntries ? * Mark Nudelman * ---------- note 509 publicatio 02/08/76 14.03 r swallow hum Could someone tell me why I havent recieved two PLATO publications I ordered in December,1975? I really need them. If you need the names(titles)let me know. The order was prepaid. Doug ---------- response 1 02/08/76 14.26 golden s You had better send a letter to Dave Self. Tell what you ordered, when, and the method of payment. ---------- response 2 02/08/76 14.45 r swallow hum ok thanx. ---------- note 514 recordsbug 02/08/76 14.57 baker me Excution error in lesson records. Unit twrite Current state regular Last command calc 18ˆtˆh command index error Fred Baker ---------- response 1 02/08/76 16.31 andersen s Should be fixed now. ---------- note 523 packc ordr 02/08/76 16.55 tom s mfl The new -packc- command is real nice, but wouldn't it be more consistant to have the first variable be the character count and the second variable be the packed string? It will be confusing to remember which command (-pack- or -packc-) has what order. ---------- response 1 02/08/76 20.57 b sherwood s Maybe old 3-argument pack commands should be converted? Now that there are so many string-oriented commands with arguments of the form "address,count", it might make sense to change the old -pack- format. What do people think? ---------- response 2 02/08/76 21.02 warner iu I agree with this. It is difficult to remember which end is which when there is only one exception. ---------- response 3 02/08/76 23.36 blomme s Yes, let's convert the old -pack-s. ---------- response 4 02/09/76 10.29 guerra uimc Hurray! ---------- note 526 suggestion 02/08/76 17.46 cole langlab Why don't you have an option so that when you are in a charset. common, or vocab block you can press -NEXT1- and -BACK1- to get to other blocks just like in regular blocks? ---------- response 1 02/09/76 09.14 mike b cornell I have a feeling that its because the charset and common editors are not in the editor lesson. M¬ ike≤≤≤≤ˇB ---------- note 527 lost ring 02/08/76 17.59 dimitrief english LOST: An Urbana High School 1977 Class Ring with an Orange Stone, and the Initials "AD" on the inside of the band. It was lost in the vicinity of the second floor around room 203. If anyone should find this ring, please contact Alex Dimitrief at 328-1645, because it was a gift and means an awful lot to me. There will be a reward to the person who finds it. Thank you. ---------- note 531 putv 02/08/76 19.31 dave fuller uimc The following line of code, which appears not to have been written since 1/25/76, does not condense as of yesterday: putv 076710 00000 00000 00000,2,0,1 ---------- response 1 02/08/76 20.20 white p That's right. The executor knew only how to handle variables for the first and third tag, but the condensor allowed any expression. So that you would know it wouldn't be done correctly, the condensor was changed to give you a condense error on such things. (In your example, it would be very unlikely for it to execute correctly!) The first and third tags must be variables on putv. ---------- response 2 02/08/76 21.44 dave fuller uimc Thanks, Larry... I was wondering if it really worked the way it was intended to, and since I did not write the coding to begin with, guess I'll have to go and find out who did. Thanks again... ---------- note 543 -restore- 02/08/76 21.47 warner iu Suggestion for programmer of -setperm- and -randp-, why not have a -restore- command that would replace a number in the same manner that -remove- removes a number, in both the setperm and modperm lists? I have just written code to do just that in the 2-argument setperm in lesson iu10, and know that it is not that difficult. The general user would find it helpful in constructing a permutation quiz. since it would allow an option to skip over an item entirely and return to it later in the run. ---------- response 1 02/08/76 22.48 blomme s That is a possibility; I would note however that remove does not operate on both lists but only the "modperm" list and not the list currently in use. Perhaps what might be useful would be a simpler way to initialize a list with entries "missing" than is currently possible (presumably one does a setperm, removes the desired items, then does a modperm if using the system kept lists). I'd be interested in other examples and other problems in this area. ---------- response 2 02/09/76 10.55 judy pso I think one would want the -restore- to balance with the -remove-, i.e to modify only the -modperm- list. *** end of notes ***