+++ plato iv group notes +++ general interest notes 1975 notes beginning oct 17, 1975 file onotes38 printed at 4:26 am on june 12, 1976 ---------- note 0 prison job 10/17 13.55 siegel ppp The PLATO Prison Project is looking for an experienced TUTOR programmer to do sophisticated data collection and routing-type programming as well as some instructional programming. The position is half-time. Please contact Marty Siegel, project director, via personal notes (siegel/ppp) or call 333-3247. If you know of someone who may be interested in the job but who may not see this note, please inform her/him. Thank you. ---------- response 1 10/18 12.29 wood siu area code? ---------- response 2 10/18 12.46 keith s matha ---------- note 10 term calc 10/17 19.42 fisher conn I found a very troublesome, for me, inconsistancy in the system calculator. It is when calculating with octal numbers. The actual line that I used is: (0777≠2ˆ9)$diff$0777777777 I was trying to get : 000000000000777000777 Plato gave me : 017417770000777777777 (To get the correct answer, type '0777000$diff$0777777777') I don't know if there is any way to fix this, and still be able to handle LARGE numbers, but people who use term calc for things like this should at least be warned of what it does. Any comments/critisms? fishswine ---------- response 1 10/17 19.54 fisher conn To be more specific on what i am saying in the note, (0777≠2ˆ9) = 0777000 0777000$diff$0777777777 = 0777000777 (0777≠2ˆ9)$diff$0777777777 ¬= 0777000777 (0777≠2ˆ9)$diff$0777777777 = 017417770000777777777 HOWEVER. int(0777≠2ˆ9)$diff$0777777777 _=_ 0777000777 !!! hopefully yours fishswine ---------- response 2 10/17 19.56 dave fuller uimc Your exponentiation is causing the trouble. If you write this expression as: (0777$cls$9)$diff$0777777777 You will get the expected number. The calc compiler automatically normalizes cases involving exponents, so you'll have to rely on the shifting operations to accomplish bit manipulations. Such is life on a floating point machine. (Note that 2↑ˆ9↑ˇ alone is an integer. Operating on this quantity forces the normalize.) ---------- response 3 10/17 20.04 ian cs425 The point was the the floating point normalization was _not_ done on (0777#2ˆ9) as a stand-alone expression, but _was_ done when it was used as a component of a more complicated expression. One of the minimal constraints on any computer should be that equal quantities can be substituted without any discernable effect. In this case, there is apparently an invisible temporary variable being used, with an implicit conversion taking place.It is incredibly difficult to debug anything when a component of an expression changes form in a manner like that. If (0777#2ˆ9) was displayed as a floating point variable, the error would have been nowhere near as mysterious as it actually turned out to be. ---------- response 4 10/18 15.45 b sherwood s If you do bit manipulations, use $cls$ and $ars$ to shift things, not multiplication or division by 2ˆN. Basically, TUTOR handles exponentiation thru floating-point arithmetic. For part of the reason why this is, consider 2ˆnˆ1, where at execution time n1 might have the value -3. TUTOR will correctly compute 2ˆ-ˆ3, or 0.125. Note that many computer languages would give 0, or some such inappropriate result. ---------- note 19 newans 10/18 17.58 koning csstaff The new ansfer features look nice, and so does -miscon-. But where is the analogous -cant- that was promised a while ago? Having that would complete the set...¬↑ˇ≤≤≤≤≤P¬≤↑ˇK ---------- response 1 10/18 22.18 hinton ssu -miscon- and -answerc- !! HOORAY !!! When will they come up on Prime Time ? Thanks. ndh ---------- response 2 10/19 13.41 fritz ames FanTAStic! ---------- response 3 10/20 02.29 t little batch is that anything like transfr? Todd ---------- note 21 cheap mail 10/18 22.19 zweig iu In the spirit of saving money, could users at remote sites be alloved to specify if we will accept less than 1st class mail of printouts? Sometimes, the next day (almost)delivery isn't necessary for me, at least. I would guess this would be a "*list" type command. Leaving no comment in the code would assume that 1st class is required... --dan ---------- response 1 10/18 23.32 al mflu *shove in an envelope and pay enough for xx class mail will do ---------- response 2 10/19 03.32 bonnie matha I think the post office requires that computer printouts go 1st class. (That's what I was told by a man with an eagle on his pocket.) Anyone know for sure? ---------- response 3 10/19 13.39 fritz ames Well, last week I received a print of a listing which was accidentally sent 3rd class... (And, I might add, which took 3 _WEEKS_ to get here!!! You could practically WALK from Urbana to Ames in that amount of time!) Moral: If you want your listings before next leapyear, have them sent first class. ---------- response 4 10/19 14.45 warner iu Hear, hear! I know of a cause where it took 6 weeks for a 3rd-class printout to travel 220 miles from Champ to Bloomington. I also know of a case where a 1st-class printout was printed at 3:15 p.m. on Thursday and arrived at 8:30 a.m. on Friday! ---------- response 5 10/20 08.51 olson ced The postoffice doesn't have any particular rules for printouts. (How would they know its a printout?) CERL could send them anyway they want to......I prefer 1st class myself..... DˇO ---------- response 6 10/21 13.50 bl abbott mtc what about those other remote sites like this one where we wait TWO WEEKS sometimes for prints...and you have been sending them first class...first class what?...mule? ---------- note 27 print who 10/19 14.49 warner iu In line with zweig's suggestion, I would like to make another. How about saving the name/course of a person requesting a print, and having that information printed out somewhere on the print? If it already is, could you please tell me where? Someone at Indiana University is requesting prints of datafile "goldat". Four of them have arrived at my office, since they contain mailing instructions to mail the prints to me. I would like to send them on to whoever requested them, but I have no idea who! ---------- response 1 10/20 08.54 olson ced I think one way to get around that would be for the person who requested a printout to send a pnote to the systems personnel who are in charge of printouts. The systmes personnel always send the printout to the name and address contained in the first block of the lesson so this suggestion should take care of that. However, I don't know who's in charge of printouts.....Anybody know? DˇO ---------- response 2 10/20 12.04 michael english That isn't going to help Warner. It is lack of forethought on the part of the original requestee that is the problem, but exhorting authors to think a little more is not likely to improve the situation. Silas' original suggestion is simple and reasonable, and would go a great way toward minimizing the problem. ---------- response 3 10/20 13.31 olson ced I concede....... DˇO ---------- note 28 miracledit 10/19 16.27 nathan syfrig iumusic A couple questions about the new editor features: a) Since I would think that most of us agree that the new options are worth saving, could you give us any idea of what you would consider an overload of disk accesses? How muh cautin and restrain should we use on those features? b) Any idea of how long the usage will be tested before a final decision is made about the premament implementation of these features (I would assume the part-indicator will be permament..I'm talking about the new search features (known as the Y-search))? In closing. THANKS FOR THE NEW FEATURES!!!!!!!!!!!!! Nate ---------- response 1 10/19 17.05 clark lawyer One question. How does one stop a search once it has started (or is this possible). I have found that stop1 will work sometimes but that is dangerous due to condensing problems. Other than that, I think it is a much better system. THANKS FOR ALL THE NEW FEATURES!!!!! J. Clark Kelso ---------- response 2 10/19 17.05 keith s matha The STOP key will stop the search. ---------- response 3 10/19 17.30 silver ve I noticed that the M option now involves a complete lesson search, initiated by pressing NEXT. I would just as soon have a block search on NEXT, and a complete search on next1. ---------- response 4 10/19 18.17 koning csstaff I second that. ¬↑ˇP¬≤↑ˇK ---------- response 5 10/19 18.41 char mathw i noticed this only happened after all occurrences in the PRESENT block were exhausted ... the search then moved on to the FIRST occurence in the NEXT block, which one used to initiate himself utilizing NEXT1 ... -search- just does it for you ... you can always change the "M" option or kill it, after noticing that the -search- has jumped you a block ... i vote to leave it as is.. ---------- response 6 10/19 19.22 daleske ames However, to cause an extra disk access to save you time and sometimes find that you dont want to go to the next block is somewhat wasteful. I vote with silver. (or perhaps does the system load more than one block at any one time? Such as the next1 and back1 blocks?) ---------- response 7 10/19 20.43 frankel p I vote with how do we know, when you hit NEXT1, whether you want to go to the next block (as is currently the case) or want to start a whole-lesson search? Or do you want NEXT1 to activate the whole-lesson search someplace other than the normal block editing display? ---------- response 8 10/19 20.47 berger mfl Please! Leave it as it is. I'm happy. ---------- response 9 10/20 00.20 hanoka ed Why wont it find a unit at the first line of a block? mike ---------- response 10 10/20 00.54 nate iumusic Never has a search of any type worked that way, and frankly, I'm glad for that feature...although I myself have tried to do that same braino, when it says "not found" and I KNOW it exists, I can conclude from all this data that I've gathered, that it is currently at the top line, which is indeed the case. Then I get to go, "You dumb clown!! Blind while at the terminal, eh?". Anything else? Frankly, I'm not too worried about the refinements right now...I'm just thankful we have the features, and want to make sure we CONTINUE to have them! ---------- response 11 10/20 08.19 mike b cornell I don't think we ever got a response to the original question. Namely: What guidelines, if any, should we try to adhere to in order the insure that these features will be permanent. Systems? M¬ ike≤≤≤≤ˇB ---------- response 12 10/20 10.05 golden s The important thing, I think, is that you not program in a way that depends on these features every time you edit. i.e. don't start scattering units in any old place and then have to use a lesson-wide search to find them. The lesson structure and the block names should still be your main aid in finding sections of code. ---------- response 13 10/22 08.16 mike b cornell Good! If that's all there is to it, there won't be any problem. You'd have to have the patience of Job to use the new search every time you needed to find a unit, since it is inevitably rather slow (compared to the flying fingers of an editor who knows where his units are). The feature is great for the exceptional case (or editing late at night when your mind is zonked anyway), but I don't think there will be any utilization problems. M¬ ike≤≤≤≤ˇB ---------- note 38 characters 10/19 19.36 daleske ames It has been brought to my attention that some of the char- acters developed in the three charsets for the Creature Character Design Contest would possibly be wanted by those other than the creatOrs, most specifically, the read group. There is a giant ape of excellent proportion and most nasty attitude, as well as scorpions, giant rat, and many critters of mythical nature (the best of which is the Griffin). The name of the lesson with the charsets in it may be obtained by sending a pnote to daleske/ames. I will be deleting the characters this Friday evening, so do hurry. John ---------- note 39 signon err 10/19 20.32 mark d csa A terminal at site 9 was stuck on the 'type the name of your course' page. Arrow was active, but there was no way to back off to sign-on page. Since a student had typed in a garbage name, there was no way to continue the sign-on. mark d ---------- note 49 pagemessag 10/20 10.08 mike b cornell I think it's time to revive the debate over changes to -term- talk. There has been plenty of talk in the past about how frivolous talks were annoying while people are working, but how nobody wants to miss an important talk by turning off his/her talk option, or -term- rejecting. I suggest that a simple solution would be the following: Add to the -talk- the option to send a brief message along with the page. So that if I had to talk to somebody about an exec error in their lesson, he might see: "talk: mike b/ cornell: exec error in you lesson." He would, of course, probably want to answer such a page. On the other hand, something like "talk: mike b/cornell: how's you mother?" might well be ignored by a busy author. Even better than ignoring it, however, would be -term- busy to replace the jolting -term- reject. Any comments? M¬ ike≤≤≤≤ˇB ---------- response 1 10/20 10.28 john r reading Those who annoy you with term talk could bug the heck out of you with phony 'important'messages or such things as 'talk: melvin snerd/hoho/you're a *':$[].' ---------- response 2 10/20 10.32 berger mfl And there's no problem with TERM reject other than the fact that too many people take it personally, when it only indicates that the person is busy. If people stopped taking "rejected" as something per- sonally, there would be no need to change the (accurate) message. ---------- response 3 10/20 10.44 mike b cornell But that still doesn't solve the problem of missing talks you might want to get, even though busy. As for the sort of messages mentioned in the first response, I would point out that it is easy enough to ig- nore a page like that (it has no effect on the paged per- son's processing), while it is somewhat harder to abruptly end a talk--especially if the pager is suffering only from mis-placed friendliness, rather than malice. ---------- response 4 10/20 13.34 olson ced Possible another suggestion would be to have term reject send a message along with it such as..... term...reject/ send pnote DˇO ---------- response 5 10/20 13.37 judy pso Clever people could use the proposed message option to write, not just on the bottom line, but all over your page. What's wrong with p-notes? If it is important, it probably bears writing down. ---------- response 6 10/20 17.06 copeland pfw Not all of us have your IRON constitution, Mike. I for one vote that the 'rejected' message be changed to something less drastic! Or, as an alternative, maybe the pagee could be allowed to type in a 20 char (arbitrary) message to the effect of 'busy now send pnote', or any message that the pagee would deem appropriate for the pager to see. your friendly neighborhood Wizard ---------- response 7 10/20 17.19 wells ee well, if you like a message a little less drastic try going into term calc. It will send a message 'wells is busy'... ---------- response 8 10/20 17.20 clark lawyer As to Judy's response...Isn't it possible to remove the offending keys?? (super,sub,back-space,etc.) Certainly this could be done with a simple put command. Right? JCK ---------- response 9 10/20 20.29 lombardo ed I think it would be a great idea to have some sort of short message ability when trying to contact someone via term talk. The point is well taken that frequently an author is too busy to sit and write to friends, not because he doesn't want to talk, but because he has to get some work done. The very problem he is working on however could be the subject of the talk request. A short optional note would be most helpful to all concerned, as well as a short note of explanation why the contact cannot be established. ---------- note 52 seminar2 10/20 10.34 john r reading Reminder:↓ WHAT: Seminar WHEN: 1:30 p.m., Tuesday, October 21st WHERE: CSL auditorium main floor-SE corner Goodwin¬+Sprngfld WHO: Sharon Hunnicutt↓ Sharon Hunnicutt is working with John Allen at MIT in the development of a synthetic speech system. The group has developed a text-to-speech system which, for a piece of unrestricted text, can decompose words and consult an 11000 entry morph lexicon or else employ a grapheme-to-phoneme and stress production algorithm, and generate synthetic speech. Ms. Hunnicutt will discuss the nature of the group's work and play a demonstration tape. One of the goals of the group is make the sytem compatible with PLATO so that 'speech' commands can be sent digitally to an audio synthesizer at the terminal. ---------- note 53 ¬owritec¬1 10/20 10.44 fritz ames I know how embedded stuff is generally frowned upon, but... I would like to propose an embedded -writec-. This would come in very handy in cases like:↓ write No, a * space ¬W writec critter,cat,dog,cow write goes * space ¬W writec critter,meow.,bowwow.,mooo. ↓which could be replaced with↓ write No, a ¬oc,critter,cat,dog,cow¬1 goes ¬0c,critter,meow, bowwow,mooo¬1. This would be ESPECIALLY handy in cases where the write is a response to an incorrect answer, and isn't erased correctly if separate write's and writec's are used. Is something like this possible? ---------- response 1 10/20 12.28 loitz acoustic Try using the two argument form of the embeded showa. Example: * at 402 write Plato is currently running in ¬0a,'ptime',(-ptime≠5)¬1¬0a,'non-ptime',((ptime+1)≠9¬1 * (could also use≤≤≤¬↑ˆ¬W¬W¬W¬↑ˇ -(ptime=-1) and -(ptime=0) * * I admit it is akward but it works! ---------- response 2 10/20 12.58 fritz ismus Ouch... ---------- response 3 10/20 12.58 iezek ames I think you missed the entire point, loitz. Fritz was not looking for an akward solution to an akward situation. multiplex ---------- response 4 10/20 13.54 judy pso awkward In the example given in the note, one would reasonably expect kitten,kitty,tomcat,puss,pussy to be synonyms for "cat". In that case you would not want to write "cat says meow", but rather "kitty says meow". Slightly less convenient than the proposed ¬owritec¬1 but much more flexible: (see next note ¬D ¬D ) ---------- response 5 10/20 13.55 loitz acoustic Right....just thought I would mention there is a possible way to get an embeded writec not necessarilly a good way but it does work. ---------- response 6 10/20 14.02 judy pso (This is the 2nd part of my earlier note.) at 1010 write What says bow-wow? arrow 1210 storea n1 specs calcs anscnt-2,n2‎'bowwow','meow','whinny' write The ¬0a,n1¬1 says ¬0a,n2¬1. answer (dog,hound,mutt,puppy) wrong (cat,pussy,kitten) wrong (horse,pony) This format allows you to use the student's word or phrase in you feedback, rather than the fixed phrases of a -writec-. ---------- response 7 10/20 16.38 fritz ames But you gotta admit, an embedded writec would come in handy. ---------- response 8 10/21 10.06 little t research sure and so would embeded graphing commands and embeded draws, and embeded rdraws, and embeded circles, and embeded calcs, and embeded do's, and embeded arrows, and embeded plots, and emdeded lessons, end embeded units, and I'd like to see this note embeded in ......... Todd ---------- response 9 10/21 11.20 schreiner csa Any of you guys ever heard about block-structure? Like ALGOL68 (...) where ( ) always means a begin/end block and thus the distiction between statements and expressions starts to fade... As we are patiently waiting for TUTOR XX. ---------- response 10 10/21 12.57 whansen csa See lesson -kaids-. A version of the KAIL language described there has been implemented and I will be glad to talk to anyone wishing to use it. (Unfortunately, we have to write the compiler in Tutor and compile into Tutor, so it is very slow.) ---------- note 59 i/o error 10/20 13.09 hart mfl received following error message: DISK I/O ERROR; unrecog (message truncated at this point) while returning to block directory of gerling2 in sys editor at 1:12pm. ---------- response 1 10/21 00.16 blomme s Thank you (most of the actual disk failures return only that obscure "unrecog" message--we just have never sat down and gone over the more than a dozen different cases that can be detected and sorted them all out. If we ever DO get a permanent failure in some area of a disk. we get the complete information regarding the problem another way). ---------- note 60 i/0 error+ 10/20 13.16 hart mfl repeatable disk i/o error: get sys editor, go to gerling2 enter some block, say B, return to block directory, attempt to access block C, error message appears. ---------- response 1 10/20 14.13 frankel p Strange. Fixed now, although some information in block C may have been lost. Sorry. ---------- note 68 microfiche 10/20 15.50 george cerl Lesson "slide" is available to all users who would like to examine a PLATO microfiche. Lesson "slide" was formerly called "susan". "susan" no longer exists. george ---------- response 1 10/21 08.19 avner s How are we going to break this to Susan? ---------- response 2 10/21 23.54 warrens uw I don't think we need to break her anymore, sounds like she's already been done in! Randy ---------- note 73 LOOK! 10/20 17.47 jeremy mflu Hi I've been playing around with some graphics and if any one would like to take a look a "jeremy" I wouldn't really mind and you might find it very intresting jeremy/mflu ---------- response 1 10/21 09.59 little t research gads ---------- response 2 10/21 21.07 char mathw ditto ---------- note 74 eraseless 10/20 18.30 parrello uimatha Continuing the note now in old general:↓ PLATO is missing full screen erases. This has happened to me at CERL, FLB and the med site (the only places I work), and most often in the editor (where I spend most of my time on PLATO). This missed full screen erase is _NOT_ the hardware problem whereby a terminal freezes and the clear key must be pressed. A page is simply overwritten by a new one in places where such things should not happen. I am not alone in experiencing this (I am told it happens in "notes" a lot), and it happens too frequently and without warning too pin down any pattern. However, the most recent occurrence for me was when I hit "+" on the main editor page and got part 3 written on top of part 2. Once again: this is NOT specific to any terminal or site; the terminal does NOT freeze until the clear key is hit. ---------- response 1 10/20 18.37 brand matha Does it deep on overwriting until the clear key is hit? Or does it just overwrite once? ---------- response 2 10/20 18.46 fisher conn When it happens, it just happens one time, i have never seen three pages written over each other, but at least five times a day, I find I am looking at two pages at once. fishswine ---------- response 3 10/20 20.49 parrello uimatha It is exactly like a single full screen erase packet got lost somewhere along the line. Very bizarre, and at times extremely frustrating, since replots may be difficult. ---------- response 4 10/20 20.52 kravitz cerl Believe it or not, this same problem just happenned to me between two replies of this note. Pressing next got me to the next page, with a full screen erase... no clear was required. This is not the first time that it has happenned to me, though it is the first time in about a month. ---------- response 5 10/20 21.09 clark lawyer This has happened to me also, although not to the extent that some people have experienced it. The screen just gets written over as if plato skipped over the erase command. Most frustrating at times. ---------- response 6 10/21 11.04 mcneil a uicc This happens to our syudents also...on the average about 4 failures a day. Always just 1 erase is missed. AL9000 ---------- response 7 10/21 11.11 hinton ssu It happens just about everywhere, too: it happens to me in the editor about five times a week: it happens while reading NOTES, about once a day, it happens in lessons.... I wrote a note about this back in July (I think), and there was no explanation for it then...as there does not seem to be now. ---------- response 8 10/21 11.13 berger mfl And when it happens to me, the terminals are NOT overheated. ---------- response 9 10/21 13.19 hinton ssu Right--we have had some problems in maintaining good operating temperatures, but the missed erase occurrences show no correlation at all with room temp. ndh ---------- response 10 10/22 17.59 judy cornell And it just happened to me, here! judy ---------- note 78 micro-edit 10/20 20.31 koning csstaff It would be nice to have a command in the editor that allows you to load a micro table (and perhaps a charset). This is useful if you are entering font characters for which there is a microtable, like some of the CS charsets. Perhaps a T command? It could work like the editor in NEWS. ¬↑ˇP¬≤↑ˇK ---------- response 1 10/20 21.07 judy pso I hope you are aware that you can hit "C" and "M" on the author mode page to load -charset- and -micro-s. I would like to be able to specify in my lesson directory (or some such place)-- "when I am editting this lesson, I want microtable 'suchandsuch' to be active". ---------- response 2 10/21 03.46 koning csstaff Yes I know about the C and M features. They are nice but while editing something more flexible is often needed. I like the idea of an automatic microtable specified in the lesson data (with an automatic charset as a matching feature) but there are cases when one wants to switch between 2 microtables while editing a lesson. Example: while editing a CS lesson I need microtable 'key' for the keytop characters and table 'big' for the block capital letters. Having to switch between them by backing out to the author mode page and using the M feature is kind of painful. In short, I would like to see this implemented. It should be pretty simple, one would guess. What is the word from systems? ¬↑ˇ↑¬≤↑ˇK ---------- response 3 10/21 04.18 midden p Most things are fairly simple to implement. Considerable thought is usually given before something is done, though. That explains the saying, "that'll take at least six months before it's even considered" from some users. Automatic loading of microtables upon editing a lesson has a somewhat lower priority on lists of things to be done someday. There are many other things that will be done first. Character sets? Hummmm... Marshall Midden ---------- response 4 10/21 08.39 judy pso No, NO! I would much rather have automatic loading of microtables, than of charsets. After all, once the charset is in your terminal, it stays there until you load something else on top of it. But the microtable specification gets lost each time you change from author mode to student mode--which can be a dozen times in an hour. I think automatic loading of charsets should be viewed with great caution. Consider... you have loaded a charset for editting, then you look in aids which uses maybe two characters for illustrating some point, then you return to edit, and you get the whole set reloaded. Of course, you can press STOP and kill the loading, but you would have to kill the loading each time you changed states from then on. It would be a big nuisance. ---------- response 5 10/21 15.20 blomme s AUTOMATIC loading of a charset seems a disaster to me also--imagine waiting for a charset to load in order to be able to get in and make a change to one (non-write type) statement! ---------- response 6 10/21 19.15 koning csstaff I guess there is a good reason against automatic charset loading, but I would still like the automatic micro load... P¬≤↑ˇK ---------- response 7 10/21 19.29 daleske ames Perhaps the answer would be a mixture of the two options that would work somewhat similar to any other temporary systems records. During the time a person is on the system, one may set a charset or micro-table to be loaded for the editing process. Then, when sign-off has occured, or when the person indicates no further need for the feature, it will be cleared. This would decrease the unnecessary loading of both the micro and the charset but would allow use of the feature. John ---------- note 88 edit bug? 10/21 09.09 moll rhrc While in replace mode in the editor. I discovered that next-ing to the last line in the block was not showing the last line of code or filling the copy buffer. The problem was repeatable by any test case. I can't seem to find any mention of this in notes and it does appear to be a bug. ---------- response 1 10/21 10.21 friedman csa OH! YES! I was seeing this yesterday also, but didn't realize what was happening, He's right. New bug.... ---------- response 2 10/21 10.41 frankel p Right...it'll be fixed as soon as PLATO comes up again (I hope). ---------- note 97 M option 10/21 11.49 iezek ames The new M option (where after all occurances in the block have been exhausted the press of next does not get the familiar "not found" but instead initiates the full lesson search, perhaps jumping one into some other block) is unwieldy. I use to count on the not found to tell me when all occurances in the block were found. (Makes sense, doesn't it?) Now these unnecessary searches. It appears strange that so much worry was expressed that the full lesson search should not be used unnecessarily and yet full lesson searches that are completely undesired are foisted upon us by this M option. I also know that this has been discussed before, with little accomplished. Perhaps some change could be made in the function keys pressed to use this M option. M. Iezek ---------- response 1 10/21 11.59 frankel p Some people like it the way it currently is; others prefer the old way. I don't care myself. So who is going to decide? ---------- response 2 10/21 12.35 warner iu Well, if it is done the usual way, there appear about 26 responses to this note, saying "One vote for", and "I vote against", and if the "for"s outweigh the "against"s, then the thing goes to the systems staff. Let me start the thing off by voting for a return to the same block. It is definitely surprising to press NEXT and find yourself 3 blocks away! ---------- response 3 10/21 12.38 nelson ced I second warner's vote. Not only is it surprising, but sometimes it's downright disgusting to find yourself far away from where you thought you were. ---------- response 4 10/21 13.12 whansen csa This question is also discussed in note 29. My vote is that something a little more drastic than NEXT should be required to initiate possibly many disk accesses. Similarly, NEXT1 and BACK1 are too little to switch between blocks. Suggestion: Next block: CR, NEXT Previous block: shift-‎, NEXT (The keys at the left are already used to move around in text. CR and shift-‎ could echo as "next blk" and "prev blk", respectively.) Continue M: M NEXT1 (Does not reset search string) ---------- response 5 10/21 13.31 perry uw X-searches from the M-option should be confined to one block when pressing NEXT. ---------- responses 6 10/21 13.41 fritz ames Perhaps you could specify when you enter your search string whether you want it to search only one block or not? ---------- response 7 10/21 15.06 darlene pso I agree with the last note. Upon finishing the modify directives have NEXT do the M only in the one block, and a SHIFT NEXT set it to go thru entire lesson. It would then be more like the x (next) search of a block and the X (shift next) for search of the lesson. ---------- response 8 10/21 15.25 silver ve Aha! I like that, Darlene... ---------- response 9 10/21 16.10 john r reading I second Darlene's notion. ---------- response 10 10/21 16.13 rr padilla mtc Three on Darlene. ---------- response 11 10/21 16.34 kimble ustaf An addition to darlene's scheme of things... While pressing NEXT or NEXT1 would initiate a single block or whole lesson search, it would be handy to: 1. Press NEXT for single block search/modify. 2. Cycle through all occurrences in that block, eventually getting the "not found" message. 3. Have the option THEN of pressing NEXT1 to expand search to the entire lesson, if desired. ---------- response 12 10/21 20.06 walter physio One vote for Darlene. ---------- response 13 10/21 21.11 char mathw after reconsidering .... make that one more for darlene ... ---------- response 14 10/21 21.14 iezek ames One vote for Darlene.. this is exactly the kind of idea I was looking for when I wrote the note originally. Systems.. its up to you now.... Mark Iezek ---------- response 15 10/22 08.54 fay o I like the idea of first getting the "not found" message, with the option of hitting NEXT _again_ to initiate a full lesson search. ---------- response 16 10/22 09.54 p cohen med Actually, I am in the habit of hitting NEXT several times when the "not found" message comes up, especially when I am frustrated by an exceptionally slow system. So I wouldn't be in favor of fay's sug- gestion. But darlene's idea shines! This would also prevent me from pressing BACK instead of NEXT at the second arrow of the "m" option, assume that BACK was a -jkey- and that the string to modify _to_ was set, and wind up with garbage. First satisfying both ar- rows with NEXT presses and then being given a choice of NEXT or SHIFT-NEXT would solve these mishaps! --paul ---------- response 17 10/22 22.33 frye mfl If there's a reason to vote, then I'd be in favor of a) leaving it the way it was (preferably), or b) implementing Darlene's suggestion. I find it just a leetle hard to believe that all these people have developed non-correctable habits already! Those options have only been around a short time, folks.... Dave ---------- response 18 10/23 22.15 fisher conn Definitely a vote for Darlene! ---------- response 19 10/24 13.56 bl abbott mtc one more for darlene ---------- note 105 site 10/21 13.32 rick mflu Execution error in lesson "site" unit "sdisp1" Current state - regular (pre-arrow) Last command - sbread 17th command Error type - (blank) I was on the last page of the site display, and had just sent a message to someone on that page. When I pressed NEXT to go to the next page, I got the above error. ---------- response 1 10/21 14.47 b sherwood s Thanks. Will look for. ---------- note 108 4 pages? 10/21 14.06 czwornog mflu This is a first for the screen erase problem: 4 pages. After typing in my name, I pressed next and no erase. Since this wasn't too unusual, I entered my course and pressed stop1 and no erase. This left me with three pages overwritten on the screen.Since I knew how much the foreign language people are interested in this type of thing, I went to get someone to see this.Then I pressed next and a VERY unusual thing happened. Four pages on the screen at one time. But the fourth page was the welcome to Plato page. Since I entered my password, I was expecting to complete the signon procedure. What happened? ---------- response 1 10/21 14.35 meers wright There is a time limit on the signin pages.... When you went to find that person you went over the time limit, and were bounced back to Press Next to Begin. ---------- response 2 10/21 14.48 warner iu Making 5 pages in all. Not too unusual lately...recently had a page of AIDS and 5 other things written over without an erase! ---------- response 3 10/21 15.05 k mast p I don't understand what the difference between this and the already known hardware problem in the terminal is. In other words, did you have to press the CLEAR button in order to cause a full screen erase? or did the terminal eventually recieve a full screen erase and act correctly? ---------- response 4 10/21 15.19 czwornog mflu The terminal recieved a full screen erase when I pressed next after the fourth(fifth?) page was written. Incidentally,I didn't know that there was a time limit on screen display while you are at the password page. ---------- response 5 10/21 16.11 dave fuller uimc I have not had too much trouble with this problem lately, and the times I have it has been consistent within a couple of terminals. I'd complain to the hardware people about this. I also noticed that the complaints are coming lately from MFL... D. Fuller ---------- response 6 10/21 16.35 artman uimatha Lost another one between pages 5 and 6 of user list. Yes, this is at FLB, but it did not take an manual clear, just caught the next full screen erase and is now behaving normally. (I hope.) hopefully helpfully, eric artman ---------- response 7 10/22 00.11 warrens uw If this helps any, I got it way out here. Yesterday the 20th, I had a 3 pager at least. Thought it was just something I did. I'm not sure but I think that I STOP1ed to start over at -press next to begin-, but don't quote me. Randy P.S. I guess that was two days ago according to the above date. ---------- response 8 10/22 08.06 parrello uimatha When I worked at CERL all the time, it happened there too. Also, during the afternoons, it happens in room 280 of the med building, where I work afternoons. ---------- response 9 10/22 17.41 weeks uimc Had 4 pages skipped a few days ago here on a cold night. Required manual clearing to get erase. ---------- response 10 10/22 22.29 frye mfl I would suggest making a list of ALL terminals which have, at one time or another, lost a full-screen erase. I'd be willing to bet that The Red Sweater has worked at every one of them. Dave ---------- response 11 10/22 23.04 john r reading A better suggestion. Most of us automatically hit the clear button at the first occurrence of a skipped erase. If everybody, systems people especially, conditions them- selves to letting the garbage pile up for a page or two, it should become clear whether the old terminal failure is the cause of the problem. ---------- note 120 art lesns. 10/21 16.14 wickham unidel We are interested to see who else is developing art material on PLATO. We are just starting out and would like to know who's doing what in the way of art lessons? leave a pnote to wickham of unidel or nichols of unidel thanx Chas. Wickham ---------- response 1 10/21 19.19 ostuni iu Perhaps you should specify more, Do you mean art history. Or do you mean lessons as a art medium? (Like graphics.) ---------- response 2 10/22 11.13 celia pso Look in the book "PLATO Curricular Materials" by Elisabeth Lyman, and then if you need more information, leave her a pnote, She is tebby of pso. ---------- note 125 erasedrop 10/21 19.25 koning csstaff This early morning I saw a new version of the dropped erase problem that may help to clarify matters. It seems to be that at least some people think that it is a software problem. But I noticed a dropped erase during the preventive maintenance period (in the display "plato is down for preventive maintenance... This is a recording..."). That just can't be software... So it's either in the terminal, the transmission lines or in the central hardware. P¬≤↑ˇK ---------- response 1 10/21 19.32 frankel p Or it is in both the software and the tape recorder system. The tape recorder system has been dropping frames here and there ever since we started, I believe. (Although I still think it is mostly a terminal problem.) ---------- response 2 10/21 21.23 john r reading Wellll, we've had a lot more teachers this year than last year complaining about garbaged screens. If it is a terminal problem (and I had assumed until the rash of notes that it was), it's far more pervasive this year than last. ---------- response 3 10/22 11.29 michael cornell Of course, because the average age of a terminal is increasing every day. ---------- response 4 10/22 14.28 berger mfl I would think it would be fairly easy to tell if it is a hardware or software problem. Why not just send (with a hardware device) "erase" command words to a terminal, and see how many times it fails to respond. Since it happens to me at terminals that seem to erase pretty reliably, and some of them are fairly new, I can't believe that it's a hardware problem. ---------- response 5 10/22 16.04 michael english I don't know whether it's a hardware problem or not, but I just lost an erase pressing BACK from the block display page. A second BACK took me to the AUTHOR MODE page. erasing the overwritten pages in the meantime. I did not touch the CLEAR button or the STOP key. ---------- response 6 10/22 18.11 judy cornell And it is happening out here too...on both our old and our new terminals.... judy ---------- response 7 10/22 19.24 koning csstaff The particular terminal I was talking about is quite old (serial ¬$11) but I have seen erase drop on lots of other terminals as well, but never more than one in a row, and never seen a case where the terminal did not recover by itself. In other words, I didn't need the clear button.¬↑ˇ≤≤≤≤≤P¬≤↑ˇK ---------- response 8 10/22 21.23 artman uimatha Just got one myself--classic symptoms of anything but the terminal. And, oh yes, this was right at home, in 1658 CERL. hopefully helpfully, eric artman ---------- response 9 10/22 22.26 frye mfl Ah, good old terminal ¬$11. If you get a dropped -erase- with that one, be glad that's all you lost... An experienced PLATOite -- Dave ---------- response 10 10/23 16.49 koning csstaff But it's been working fine for me...........¬↑ˇ≤≤≤≤P¬≤ˇK ---------- note 129 records 10/21 20.22 dave fuller uimc When you are in the course records editor, and are inspecting a student's record, STOP replots the screen. This is annoying, especially when you know what you are going to do, say, change a password or user type. Is there any reason why STOP funtions in this fashion? D. Fuller ---------- note 134 leslists 10/21 22.35 berger mfl Two questions regarding leslists: The "universal" variable jumpout was put on the "obsolete" list because we were told that we would be able to attach as many leslists as we wanted. The "aids" description of leslists now tells us that we may only attach one (although this MAY change someday). In light of this, can we have the variable "jumpout" and variable "from" removed from the "obsolete" list? Secondly, how about an optional third tag for the leslist command to allow a codeword check: this would eliminate the necessity of having the common codewords match in utility type lessons. ---------- response 1 10/21 22.55 carter comm I agree that it is time to get off of the pot on this one. Some decisions need to be made one way or the other. I hope some way is found to accomadate those who note that the leslist feature doesn't provide all the capabilities of the previous methods. If variable jumpouts are definitely on the way out, then set a date when they won't work at all--hopefully not before the end of the term. The present system simply encourages the authoring of more obsolete lessons since people really don't take the obsolete notice seriously any more-- its been there so long. ---------- note 141 site 10/22 09.18 meers wright I would like to suggest an option for the site directors controls that would enable us to change the message seen when a student with a restricted course tries to sign on. Oftentimes it would be advantageous to change the message to reflect something like 'There is a class now.............' . The revamped message and restriction list options are fantastic. ---------- response 1 10/22 11.52 broadus css I wish that someone would put something other than a pause keys=all at the end of the This site has restrictions page. I don't feel like having the guy peeping over my shoulder see all but the first letter of my password. Bob Broadus ---------- note 143 block-name 10/22 11.11 nudelman css A minor problem: I was editting my lesson≤ (name on request) around 11 am today and saw a block name on the block directory which I didn't recognize: "kar:rla". I entered the block and the name was given as "kardla", which IS the correct name. Somehow the directory replaced the 'd' with a ':'. The name was correct when I left and the error was, of course, nonrepeatable. * Mark Nudelman * ---------- response 1 10/22 11.15 halvorsen forestry Sigh. Another terminal error problem... ---------- response 2 10/22 11.53 broadus css more like line errors I would say.. ---------- note 149 erase talk 10/22 13.25 ka lutz usc several days ago we were photographing one of our lessons and i received a talk message at the bottom of the screen. i never really noticed before but that line is not erased when the caller gives up. So i tried doing a term talk just to erase left overs (talk:name end of talk) but that erased the bottom 3 (or was it 4) lines of the screen and took some of the display with it. So i had to go through the lesson again to get the cummulative display back where it should be for photographing. To get to the point, can you please erase the talk message from the receivers screen when he/she chooses not to answer? Or was it not erased due to our infamous phone line error? kathy ---------- response 1 10/22 13.57 stone research it is not usually erased. i prefer it that way so i know that someone called while i was away from the terminal. why don't you just shut your talk options off while you are photographing? that way you will never get the talk message at all. maureen ---------- response 2 10/22 15.07 michael english Agreed. The "so-and-so called" nature of the talk option is one of its nicest features. Turn off your own options rather than inconvenience everybody ---------- response 3 10/22 21.42 john r reading exactly. ---------- response 4 10/23 09.14 zweig iu I think a term -calc- would have only erased the bottom two lines, that might have solved your problem. ---------- response 5 10/23 11.34 ka lutz usc not to belabor the point because i am now convinced that i should have just turned off my talk options. BUT I still think it would be nice to be able to erase that message (maybe by hitting an e or term erase) since one may not always remember to turn off options before a demo or may choose not to answer a particular page and not want it to remain on the screen. I certainly wouldnt want it to inconveniece others....... actually, it seems that thouse of you who stay signed in at a terminal while not using it are actually the ones doing the inconveniencing of others!!!!! ----------response 6 10/23 14.15 ka lutz usc p.s. the talk option is for paging....p-notes are for indicating that "somebody called" . ---------- note 167 modify M 10/22 23.30 bob y reading as requested several days ago, i really wish that the modify option would NOT continue to search the next blocks when i just press next; most often i am more interested in making sure that i have found the last occurrence of an object within that block, and would prefer a wraparound within that block rather than going to the next block. ---------- response 1 10/22 23.40 blomme s I agree; the full lesson search can always be set as an "override" option--and there are also other ways already suggested to make the full lesson search be set at the time of the modify string entry. ---------- response 2 10/23 00.34 midden p Next time the system comes up full lesson X search will be an option on the modify entering page. Marshall Midden p.s. talk about people getting upset with this! ---------- response 3 10/23 01.36 koning csstaff Beautiful!!! ---------- response 4 10/23 01.52 halvorsen forestry Thanks, Marshall! ---------- response 5 10/23 09.55 p cohen med Thanks, Marshall! (Please don't blame us for getting upset...we are just nervous about breaking our newPLATOy.) --paul ---------- response 6 10/23 15.27 midden p (I was talking about myself getting upset at it!) ---------- note 171 notes bug 10/23 01.33 koning csstaff When trying to reply to Help note ¬$565, I found to my amazement that rather than the usual space of 100 for the note, I only got 37... I could only put in 6 lines. It was not in the text because I cleared the note and saw "SPACE 37". It is not because the notes space is full be- cause I tried to enter another reply and got 100 words. I finally deleted the note completely and re-entered it, at which point I got 100 words....... Wierd, for sure! P¬≤ˇK ---------- response 1 10/23 09.23 judy pso Were you trying to modify an answer which you had entered (even slightly) earlier? If there has been a change in the notes file between your two entries, then you only get as many words as your original response. ---------- response 2 10/23 16.55 koning csstaff No, it was a new reply, and besides what you suggest is not true; I have changed (added to) notes before and never found that it only gives me the amount used rather than 100. For another example of that, look at note ¬$186, which has a line appended to it later (I would think), and the fact that this paragraph was added later also....¬↑ˇ≤≤≤P¬≤↑ˇK ---------- response 3 10/23 19.55 blomme s Judy's remark is exactly correct as it was stated; if additional notes are added between the time you enter the note and the time re-enter and try to change it you will get only the space your original note took and not 100 words. ---------- response 4 10/24 08.27 judy pso (I have no idea why a _new_ reply would give only 37 words-- which I guess was your original puzzler.) ---------- note 175 please res 10/23 09.22 berger mfl Please - could I have a system response to general note 135: I need to know, so I can write my lessons properly. ---------- response 1 10/23 13.50 curly iu Yes, there are many others who would like to know also. A-kept-in-the-dark-curly≤≤≤≤≤¬↑ˇ↑ˆ↑ˆ↑ˆcurly ---------- response 1 10/23 15.25 bowery comm I would like, for the sake of my children's children, to request that a change be made in the function of the function keys on the author mode page so that they better approximate their names. ie: DATA inspect a program SHIFT DATA inspect last block edited LAB try cut lesson Other advantages: The old problem of doing a SHIFT DATA by mistake would disappear. LAB would be consistant with the present SHIFT LAB to try out a lesson already in ECS (if present) from inspecting/editing a lesson. ---------- response 1 10/23 16.02 halvorsen forestry I have to agree that LAB is a better key to use for trying out a lesson and is more consistent. Possibly the options now available from the SHIFT-DATA page will be available from the help pages where they should be in the first place. ---------- response 2 10/23 16.16 bruce iehl mtc There are some people who have not yet adjusted to the previous change which eliminated the student mode page. I suggest there be 2 next keys, one for what you think you want to do, and one for what plato thinks you want to do. >>conserve function keys<< ---------- response 3 10/24 02.09 k gorey research Although it may be difficult for the dyed-in-the-wool platoites to adjust to, I believe that a change in these function keys would help to improve the logical consistency of the system. something we are sorely in need of. ---------- response 4 10/23 07.44 maggs law I must agree. Since I like many people have never adjusted to the most recent change I would have no relearning process to go through. I really think we would be making it easier for a lot of future programmers. ---------- response 5 10/23 08.07 golden s I still favor the NEXT key for viewing the lesson as a student and the EDIT key for editing! ---------- response 6 10/24 05.58 friedman csa Despite the nice mnemonic of Bill's suggestion, let's leave EDIT alone to do its normal job. ---------- note 185 notes err 10/23 16.42 anderer ee Possible point of interest---after the two crashes that just occured, when I signed in again, I had the 'NOTES' message, along with the arrow indicating p-notes on the notes page. But then received the message that I have no notes, which is correct. Problem? (Make that 3 times out of 3) ---------- response 1 10/23 17.36 golden s No, that does not indicate a problem. You have to sign-out to get rid of the note indicator. ---------- response 2 10/23 20.27 anderer ee It has been my experience that once I have read a 'new' p-note (ie, one I haven't seen before), the marker on the author page disappears. But in this case it appeared three times when I had received no notes, and had none in my file. On these three occassions, I had just signed in, and each time after I attempted to read my notes, the marker disappeared. It shouldn't say I have notes when I don't, should it? ---------- response 3 10/23 20.41 fisher conn When the system crashes, your records have not been returned to disc, as is what happens when you sign off via stop1. Therefore, the system still thinks that you haven't seen your notes, which, in your example, you have deleted. A helpful fishswine ---------- response 4 10/23 21.20 anderer ee Thanx for the explanation. dg ---------- note 192 recondense 10/23 20.52 fisher conn I was just on the Condense Error page for a second. pressed next, and got on the 'now preparing lesson' page. Was it an unusual circumstance, or can we except to get this again? just reporting an inconsistancy. A hopefully helpful fishswine ---------- response 1 10/24 08.42 judy pso This _did_ occur quite frequently a while back. The explanation was that, while you are meditating about your condense error page, there is no "student" in your lesson. Hence it is eligible for deletion to make more space in ECS. When ECS was very full, the deletion routine was so busy deleting things that your lesson got deleted before your ever got to it. This was fixed..... your lesson is not deleted while you are on the condense error page. I expect what you saw (in non-prime time, I notice) was just a temporary aberration. ---------- response 2 10/24 09.06 fay o I disagree, Judy, based on my own recent experiences. On several occasions during prime time this week, I have had my lesson deleted from ECS before I could hit NEXT from the condense error page. That is, I had errors which were known to me already, so I spent X time (X¬<1 sec) looking at the error page. Still got deleted several times. ---------- response 3 10/24 10.22 perry uw Same thing happened to me yesterday. There must be a (small) bug somewhere. ---------- response 4 10/24 11.24 campanini mfl Same thing was happening during prime time at FLB the past few days. ---------- response 5 10/24 13.32 judy pso Well... I guess it's not fixed after all! ---------- note 198 Fiat 10/24 08.26 olson ced If this note shouldn't be here, I apologize.... 1969 Fiat....124 Sport Coupe FOR SALE≤≤_≤≤≤_ ____ Engine: 4 cylinder, double overhead cam Transmission: 5 speed I'll sell real cheap...like 350.00 leave pnote to olson/ced or reply here... DˇO ---------- response 1 10/24 08.32 mike b corneel This note shouldn't be here. General Notes are archival--they are stored virtually forever, so that "future generations of PLATO users" will have the benefit of a large accumulation on empirical information about the behavior of the system and features and services not docu- mented elsewhere. Since PLATO authors in 1980 will have no interest in you Fiat for sale, this note--if it belongs in notes at all--belongs in Help Notes, which are recycled. I know the names are misleading--something which should be fixed when (if?)the new notes system comes up--but that's the way it is. M¬ ike≤≤≤≤ˇB ---------- response 2 10/24 09.54 brant ames General notes are NOT virtually forever. A number of my general notes have been deleted (by someone). They all concerned systems errors. George Brant ---------- response 3 10/24 10.07 jmk pso try lesson "wantad" for this sort of thing. ---------- response 4 10/24 12.40 little t research I find it hard to believe that someone is deleting your notes for the specific reason that they contain references to system bugs. If such a thing has occured I'm sure there would be many interested people. Now if there are other reasons as to why the note might have been deleted that's a different story. Todd ---------- response 5 10/24 15.10 fritz ames I can vouch for Dr. Brant. Several times he has run into errors in the system, reported them in notes, and checked for responses later only to find his note had mysteriously disappeared... ---------- response 6 10/24 15.35 sellers arizona How about date and time of such happenings...perhaps a personal log for recording future happenings of such nature? If you do this perhaps it would be a help to the systems people in tracking down the person and/or cause. ---------- response 7 10/24 15.39 brant ames I have given information to a systems person relating to the last such deletion of (one of my) general notes. The situation is being investigated. The note in question actually had three responses and was deleted a few days after I had entered it. George Brant ---------- response 8 10/24 16.50 tenczar s I would sure like to know about deleted notes. The only cases I know of...had the deleted notes _replaced_ by a note stating that this note was deleted for such-and-such reason(such as bad language or some system error that causes a system crash (which, of course, shouldn't be advertized until fixed)) The system staff realizes that the whole openness and usefullness of -notes- would be jeopardized by censorship! ---------- response 9 10/24 20.08 nate iumusic I have seen a couple mysteriously deleted notes... the one that comes to mind at moment was one written as a bad (and harmful joke) written by one of the authors down here before we got rid of him for doing weird things (like editing courses he had no right to be in, etc.) oooohhhh booooyyyy.... Nate ---------- response 10 10/27 08.32 olson ced Good Grief.....What have I started.....!!!!!! DˇO ---------- note 204 proofs 10/24 09.49 gelder cerl Five lessons devoted to proofs of elementary theorems in algebra are available for your inspection and use ---proof1,proof2,..etcetera. Please send your comments and criticisms to either gelder, cerl, or to jmk, pso. Happy proving. HMG. ---------- note 208 leslist 2 10/24 11.00 berger mfl Since I still haven't received a system response, and it is quite important to me, I have re-entered this note: Two questions regarding leslists: The "universal" variable jumpout was put on the "obsolete" list because we were told that we would be able to attach as many leslists as we wanted. The "aids" description of leslists now tells us that we may only attach one (although this MAY change someday). In light of this, can we have the variable "jumpout" and variable "from" removed from the "obsolete" list? Secondly, how about an optional third tag for the leslist command to allow a codeword check: this would eliminate the necessity of having the common codewords match in utility type lessons. ---------- note 210 use 10/24 11.25 maggs law Wouldn't it be nice if the system could use some of those long night time hours to access disks and update a file which would let us know for any block of code, charset, micro or leslist which lessons it was used by. Another nice feature would be a service which would automatically -- overnight -- change all names of references to a lesson when the name of the lesson was changed. ---------- response 1 10/24 12.37 little t research ARGHHHHHHHHHH the people that work at night will love you. Todd ---------- response 2 10/24 16.28 koning csstaff What the hell... Run it under low priority (like cpulim =1 or so...) Shouldn't hurt too much... ---------- response 3 10/25 10.21 dave fuller uimc Just another thing to worry about when testing the system, heh? ---------- response 4 10/27 14.57 tricorn hebrew What if the name is being changed to make it 'unlisted'? All a person would have to do is put a name in a leslist or something and he would keep a string attached to the lesson.... ---------- response 5 10/28 17.44 michael english Such "unlisted" lessons should have internal protection. Secrecy of file names has always seemed idiotic to me. ---------- note 215 plato off 10/24 14.42 golden s The CDC engineers wish to make some circuitry changes in one of our communications devices (DDP). So, PLATO will be off from noon to about 5pm tomorrow, Saturday the 25th. ---------- note 221 exec error 10/24 18.31 oberpriller arizona This is something that would prove very useful: How many times have you been running a lesson that is not yours, and you get an execution error? You have to copy all the info down by hand, then write a p-note to the author or put the info in help-notes. How many times have you just ignored the exec error just because it was a pain to hand copy all that info and then type it in again? How often have you, as an author, gotten a note about an exec error in one of your lessons, and there is not enough info given to track it down? How about an option to do an S like in the editor, on the exec error page. then go to "notes" and do an IS? This feature shouldn't be too difficult to add. dave ---------- reponse 1 10/24 22.05 roper siu I agree. Quetzal ---------- response 2 10/24 22.14 hinton ssu Agree also: too often I find that I STOP1 out, and then remember later that I should have copied down all the data, or I get in touch with the author, and have omitted one of the pieces of information... ---------- response 3 10/25 09.34 avner s Something allowing such notes is indeed being considered. All things come...... ---------- response 4 10/26 18.00 al mflu I don't see what the big deal is. The exec error page never tells the author anything, anyway. Even IBM with their numbers does better! ---------- response 5 10/27 08.12 judy pso It depends, Al. If I have a lesson which I believe is working and then receive a p-note saying your lesson has an "exec error" in unit such-and-such, it tells me a LOT. ---------- note 223 lab-inspec 10/24 20.08 koning csstaff If I use the LAB key to look at a lesson in INSPECT-ONLY mode, then go into the charset editor, then return to the main editor, I am no longer in inspect only mode. It would be nice if the inspect only mode were preserved. It is not that the charset editor doesn't know about this form of inspect-only; it could very well be made to restore the setting when returning to edit..... ¬↑ˇP¬≤ˇK ---------- response 1 10/26 10.57 white p This will be fixed, someday. ---------- note 226 print stat 10/25 01.11 parker ualern The introduction of accounts allows those privledged persons to change the names of lessons with the touch of a key. This feature has potential hazards when requesting status of lesson print-outs. I have discovered that if a lesson doesn't currently exist when a user attempts to check the status, the search for the lesson will not be performed. I suggest that the status search be performed regardless of lesson existance. ---------- response 1 10/25 09.23 clark lawyer I doubt very seriously that account directors are going to start playing games with lesson names, sizes, etc... Account directors should always let interested parties know what they might be doing to a lesson and in most cases I would hope that account directors do nothing unless asked by an author. The whole reason for setting up accounts was to 1) give greater control to authors to set up lessons quickly, etc... 2) to take some of the load off of sheila and other members of PSO 3) to give another classification to lessons. (maybe there are some good system reasons also!) I do not believe that account directors are going to start making arbitrary decisions about other peoples lessons. They should merely accept requests from authors as PSO did. There should be no problems as you described because lesson names, etc., should only be changed at the author's request. J. Clark Kelso ---------- response 2 10/25 18.03 sellers arizona As far as I can tell the reason for lesson dependency is to to insure that the searcher gets feedback about the correctness of the lesson name he entered. Am I correct about this system persons? This being the case and assuming the change is easy to make, the best approach would be to go ahead and make the search but inform the searcher if the lesson does not exist, or inform the searcher that the lesson doesn't exist and give him the option of searching anyway. As for leaving the facility the way it is...I think it should be improved in some way because in repect to the present consideration the configuration is not an example of good programing practice. ---------- response 3 10/27 08.54 friedman csa I don't see what you guys are griping about. When you request a printout, a check IS made to see that the requested lesson exists. It should be a rare event for a lesson name to be changed, or the lesson to be destroyed, between request time and printout time, For that rare event, would you really have them create a lot of cumbersome machinery to search the account log to see why the lesson name has disappeared? I don't see that it would be worth it. Or are you saying you don't trust your account director? If that's true, you are in a very bad way. ---------- response 4 10/27 09.18 fay o A) Account directors and other users of the account lessons are solely responsible for remaining enough in touch with each other's activities to avoid problems like those mentioned. I don't think this is a system problem. B) There have been a few (stress FEW) cases where a print was requested, then the name of the lesson was changed (effectively destroyed as far as PLATO is concerned) be- for the print was done. Result: the print never gets done. IF a user is requesting a print in order to have documenta- tion BEFORE making some devastating changes (in the code OR the name) he/she should ALWAYS have the print firmly in hand BEFORE effecting those changes. This "rule of thumb" has been the norm for as long as I can remember. In this case it would seem to apply especially well to accounted lessons. ---------- response 5 10/27 11.22 rader s As Tim has pointed out, people who are asking for prints when someone is going to change the name of the file are in trouble--and I don't think it is worth putting in machinery to tell you that this situation has arisen. For those who change the name after the print is done, but want to check the status anyway--I am not too sympathetic. It is correct that the check could simply be taken out and let you search for any name, whether the file exists or not--the current arrangement prevents mistyped names from generating unwanted searches through the disk files that keep this information. So which is more prevalent-- mistyped names, or files that have disappeared since the print request? ---------- note 230 -score- 10/25 11.31 b sherwood s Ruth Chabay and I are curious about what use is made of scores generated by the -score- command. As an example, in physics "homework problems" are worked for a score, and the score is part of the student's grade. A physics "gradebook" (see lesson gradebook for a sample) shows the student these grades (as well as others). Instructors using the system router can see scores obtained by students, but at the moment there is no display for the student to see his own scores. Should the system router provide this information to the student? Presumably instrcutors do make some decisions based on scores, so it seems plausible to show them to the students, too. ---------- response 1 10/25 22.02 hinton ssu Though I have not been using -score- to generate them, I think that students ought to be shown their scores when they complete a quiz, jst as they would get back graded quizzes, etc. in a "regular" (sorry) course. What is the value of using -score- for this ? If I am storing a student's score in,say, a variable defined as "grade", what would a reason be for using -score- instead of a -show(a,s,or whatever) grade- ? (or "grade/100" for percents ?) If there is a reason for using -score-, I cer- tainly would like to use it (I have read AIDS on -score-). But I have not yet seen the reason for calc-ing a var into "lscore" rather than just storing the var. ndh ---------- response 2 10/26 00.50 tom s mfl I have found "score" useful for passing data back to a router for decision making. In this case, I wouldn't want a student to see the score because it is meaningless to him. ---------- response 3 10/26 08.43 stan smith chem In my lessons which have scores used for grading, the lesson gives the student his "score" at every place where points are calculated as well as the total score when he leaves the lesson. The system router is used to summarize the data for the instructor. It all works very well. ---------- response 4 10/26 10.50 judy pso In response to hinton, I think the question is Should the system router show the student his score in addition to marking lessons completed with a * ? This could remind the student of his relative strengths and weaknesses. There might be occasions where the student would be allowed to re-do a lesson to improve his "score". ---------- response 5 10/26 14.16 roberta mfl In reference to the question as to whether or not the system router should also display scores, I believe that the instructor setting up the router should be the one to decide whether the scores should be shown. I would hate to have to have the score shown in all cases. ---------- response 6 10/26 19.58 hinton ssu Well, I wasn't thinking that the sysrouter should auto- matically display scores: they might be meaningless in a "raw" state: or *lscore* might be used for some purpose not immeidately evident to the student. What I meant was that when a student takes a "test", "quiz", or other formal examination of some kind, he ought to be told how he did: it just sems like good teaching. ---------- response 7 10/26 21.33 b sherwood s Agreed. We assumed that of course any lesson would show the student how he was doing. We were asking only about displaying scores stored for long times. ---------- response 8 10/27 08.58 friedman csa The cs router also accepts and stores -lscore- values; a utility lesson, csrecords, can display this information to the cs student. It happens that most cs lessons do not (so far) set -lscore-, but whatever, the student can see it anytime he wants. (He can see his own info only, of course.) ---------- response 9 10/27 09.23 fay o I support the idea of making the showing of scores by the system router an optional arrangement. I, too, sometimes use the values of -lscore- for purposes other than really recording a meaningful measure of student performance. In other words, Bruce and Ruth, the instructor should be able to turn the "score showing machinery" at will; I would object, however, if it was one of those infamous "either/or" decisions which would necessitate re-initializing the router in order to change the mode of operation. ---------- response 10 10/27 18.02 michael english To be useful at all, you would have to keep a bit string and let the person who sets up the instructor file (not always the instructor) specify WHICH lessons in the catalog should display scores, since the catalog owner would presumably know which lessons use -score- in that way. What has developed is that people are using *lscore* to pass information between lessons/router since the system has no easy means of providing an argumented jumpout capability. Perhaps work should be done on this so that -score- can be used more consistently for its original purpose. ---------- note 231 TIPS 10/25 17.44 broeker aero After condensing a lesson and then returning to the author page,I noticed that the lesson used 51.6 TIPS. I tried the lesson again and this time it was 128.7 TIPS. What gives? ---------- response 1 10/25 18.50 little t research If there is nothing in the lesson and you go into it you use all the processing for the overhead of the timeslice but it is ended immediately since there is no code to execute, i.e. lots of processing for a very little time. Todd ---------- note 239 ¬0size¬1? 10/26 12.31 weible fl An interesting glitch observed in embedded sizing. The following code: at 1611 write Therefore, although our relative pronoun will still be masculine in gender and singular in number,¬0size,1¬1 it will have to be in the accusative case, because it will take the place of the direct object of the verb in the clause it introduces.¬0size,0¬1 condensed normally through the end of the line containing the first of the embedded size commands. The following lines, however, appeared two lines higher than they should have, overwriting that part of the text. As this occurred on a Sunday, it may simply have been due to a fault in the running version of PLATO. But it seemed to me to be a serious enough problem to warrant inclusion in general notes. A consultant found it just as baffling as I did. ---------- response 1 10/27 08.27 judy pso Monday morning and it is still doing it.... Further details.... The misplaced text always seems to appear on the 2nd line of the -write- statement: at place write something something something ¬0size,1¬1 misplaced !misplaced" always appears at place+100. ---------- response 2 10/29 23.01 white uimatha Apparently the sized writing is using the current *where*, which is only updated at the end of the -write- command. I don't think it will be an easy fix. ---------- note 240 notes 10/26 12.59 lynch unidel I like the HELP and GENERAL notes very much, but are there any good notes lessons that don't primarily concern them- selves with the PLATO system? I've seen pad and aggwk and don't think they are worthy of further mention.¬↑ˇ¬↑ˇBill Lynch. ---------- response 1 10/26 13.03 silver ve Lessons "discuss" and "medtalk" are good examples of such lessons. ---------- response 2 10/26 13.22 mcneil a fl or micronotes or peripheral ---------- response 3 10/26 13.33 berger mfl You don't indicate what it is you're interested in. If you want to set up your own notesfile, then guerra of uimc has developed an excellent system. Contact him for details. If you are simply interested in reading other people's notefiles, ken barr, ken 8 of cccs, has a list of fairly important notefiles. ---------- response 4 10/26 16.36 silver ve Three more that don't deal with Plato: "wantad" and "pad", and "news" which is really an electronic newspaper. ---------- response 5 10/26 20.56 dave fuller uimc Perhaps we need a listing of notesfiles. Any takers? ---------- response 6 10/27 01.11 daleske ames Lesson -gaming- has a section devoted to communications lessons. [section g] I will attempt to contact all of the authors of the previous lessons to have them added to that list. Before entry to this list, authors must give their "OK". If you have any lessons to be added, please let me know. (whether the author or a user)... I would especially appreciate additions to the Educational Games list. These need not be "games". Any lesson that uses more than droll text to get an idea across will be useful. Lessons -west- and -wmg- are good examples of educational games, as well as the math index games. [I guess this has gotten a little off of the original note, but the extra seemed to fit...] John ---------- note 245 mysterydot 10/26 15.26 al mfl Why is -TERM- cursor 1 dot off in fine grid? Placing the cursor all the way to the bottom or all the way to the left and pressing "f" says that I'm at 1, not 0. Pressing ¬X (or ¬A) moves the cursor to 511 but says it is at 0. Pressing ¬X (or ¬A) agan does not move the cursor but finally displays 511. Not only does it do this, but if you move the cursor from dot 2 to dot 3, or back, it moves 2 dots. Any chance of fixing this? Alan ---------- response 1 10/26 15.30 parrello mfl Actually, when moving horizontally, the glitch is between dots 2 and 3, but vertically, it is between 6 and 7. ---------- response 2 10/26 22.45 midden p Fixed upon next reload, either that or it'll be off more often. (-1$mask$0777)¬=511 glitch stricks again! Marshall Midden ---------- response 3 10/27 01.14 daleske ames I hated to have to fix my cursors with a calcs, but it seemed necessary. In this case, my fix was with: calcs x,x‎511,x$mask$o777 How did you do yours Marshall? Is there a better way? John ---------- response 4 10/27 09.05 parrello uimatha Blasted -0 always getting in the blasted way, blast it... ---------- response 5 10/27 14.10 koning csstaff Try checking for o77777777777777777777 or whatever -0 is ---------- response 6 10/27 17.19 midden p Yes, try the following: at locx-offsetx+_512_,locy-offsety+512 * works fine and dandy. (assuming locx and locy are between 0 and 511) The -at- command masks to 9 bits, so the overflow is lost. Marshall Midden ---------- response 7 10/28 13.15 white p Bruce and Paul, it isn't a -0 getting in the way, it is the fact that we are using a ones complement machine, and (-1$mask$o777)=510, not 511 as we happen to want. Yes John, there is a simpler (and faster) way: calc x‎(x+512)$mask$o777 ---------- note 248 mess-garb 10/26 17.01 wood siu Lately, the messages sent out during a long crash are being garbled by the time they get to siu. Does this problem occur at other sites? If so, I would prefer a simpler format to give the pertinent data instead of the more elaborate format. Ray Wood ---------- response 1 10/26 17.57 al mflu It appears that all of Dave Frankel's tape recordings are bad. I don't think anyone has gotten one right yet. The best I saw was: Sorry, Plato is off. We expect Plato vice will be unavailable for at least the next half hour. ---------- response 2 10/26 20.42 nate iumusic they used to be fine...but I gues they wear off too quickly...oh well....(I remember when they plotted all those things correctly....even the frog croaked in style).... ---------- response 3 10/26 20.46 roper siu However, the messages during Saturday's planned crash were a joy to behold!!! Except of course for the fact that they meant PLATO was down... Quetzal ---------- response 4 10/26 21.06 frankel p Yes, folks, our new tape recorder (which happens to be 15 years old) seems to work much better. The "back in half an hour" tapes have been run on the previous tape recorder, which doesn't work nearly as well. However, we will be switching over completely to the new (old) machine within the next few days, provided we figure out how they are interacting with the microwave system...by the way, microwave sites, how did the message sent Saturday afternoon look to you? ---------- response 5 10/27 01.19 daleske ames The down messages came through excellently here at Iowa State. I watched them from 16:30 to 17:22 when the system came back up. Only a few times did the display "act-up"; most usually, the acting-up was the arrival of too-many (!) full screen erases. Dave, did you insert some extras? "Norbert" (as we affectionatly have titled the PLATO eating creature) was quite a delight! Thankyou systems for the little extras. John ---------- response 6 10/27 08.37 berger mfl We couldn't read a single display at FLB... ---------- response 7 10/27 08.51 stan smith chem The tapes simply did not work in chemistry. ---------- response 8 10/27 09.51 lieber m While the error rate was very high at sites in Chicago it was still possible to get the information on how long the system would be down. dick ---------- response 9 10/27 10.04 frankel p Mike Williams told me this morning that the site controllers on the microwave sites need to be re-adjusted. So, after that gets done, I'll take another poll. Thanks... ---------- note 261 + - answer 10/27 04.25 tenczar s I would like to change the interpretation of "+" and "-" during word-type answer judging (via commands like -answer-) Currently, when "+" or "-" occur within a word...they are illegal during condense time for authors and treated as punctuations for students (i.e., "Anglo-American" just become two words). I would like to make "+" and "-" behave as legal "letter- like" characters for both author and student for this case in which they occur _within_ a word (an entity starting with a letter). Thus, hyphens will become legal input for authors and code into a one-word entity. Thus, "Anglo-American" and "Agˆ+" will be legal one word entities. However, this means that authors who "handle" hyphenated cases by just including two words (or by making a phrase out of them)...will have to re-do these commands to put in the hyphen. OK? ---------- response 1 10/27 08.37 gravel chema Personally, I would greatly appreciate seeing this feature adopted. I would make writing the code for accepting chemical symbols as answers much simpler, and would also make it more readable. I would hope this feature would not be too much of a burden to those who would have to change their code. .....Phil ---------- response 2 10/27 08.38 berger mfl Please! Since the answer machinery would handle Ʃ and Δ as expected, I never could understand the inconsistency. ---------- response 3 10/27 08.46 stan smith chem This change would help very much! It would be nice if we could also use the hyphen in cases where the word begins with a number. For example: answer 2-bromobutane or answer 2,3-dibromo-4-chloropent-1-ene At present both the , and - must be -put- to other characters. ---------- response 4 10/27 09.00 hinton ssu Yes indeed ! ---------- response 5 10/27 09.17 darlene pso Yes, please allow the + - in -answer-. The ease of using them in the future would override any work in changing my present lessons, that "had to get around" the + and -. ---------- response 6 10/27 11.29 jmk pso It would make things much simpler for some russian stuff. ---------- response 7 10/27 15.48 dave fuller uimc For sure! This would make things lots easier in judging listing options. ---------- response 8 10/28 11.47 brant ames I think it would be a good additon, even though I'll have to rewrite some of my lessons. George Brant ---------- response 9 10/28 22.43 rich cs196 I agree! Having to put all possible responses in -exact-s is time consuming!! ---------- note 264 llist sigh 10/27 08.29 berger mfl Go ahead! Ignore me! See if I care... Regarding leslists: When leslists were first established, the variable-tag "from" and "jumpout" commands were put on the "obsolete" list. We were told that leslists would do everything that those commands did and more. In light of the fact that leslists are entirely inadequate for those purposes, particularly because we no longer even have the promise of being able to attach more than one, please take the variable-tag "from" and "jumpout" off the obsolete list. Secondly, to make things easier for "universal" utility routines, could an optional third tag codeword be put on the leslist command to do common codeword checks on the system level similar to the way dataset codewords are handled? ---------- response 1 10/27 10.00 fritz ames Aren't you grateful for the IS option in NOTES, Mike? ---------- response 2 10/27 12.44 curly iu Will some system person answer the man ....he's been asking long enough. curly≤≤≤≤≤¬↑ˇ↑ˆ↑ˆ↑ˆcurly ---------- response 3 10/27 13.10 chabay s The question of leslists, variable jumpouts, etc. is currently under discussion among the systems staff. When this finally gets straightened out, we'll let you know. Sorry that it's inconvenient -- but hang on just a bit longer. ---------- response 4 10/28 16.51 broadus css is that a 60ˆtˆh of a word? ---------- note 281 -contest- 10/27 11.13 fay o ANNOUNCING THE FIRST (and probably the last PLATO DOWN TAPES DRAWING CONTEST User response was so underwhelming toward the new drawings in one of the PLATO OFF tapes that we would like some contributions from the users!!! Besides, none of us can draw. SOOoo....see lesson "tapedisp" for details and rules (in the author mode ONLY!!). Please read the rules in the INSPECT MODE to avoid attaching the file. We will attempt to determine the final winners within ten days. The closing date for entries is midnight, Nov. 5 (Wednesday). Winners wll be announced within 48 hours of that time. Happy drawing! ---------- response 1 10/27 11.50 frankel p Bribes are payable, as usual, to myself. ---------- response 2 10/27 15.45 dave fuller uimc Arsenic carab cookies? ---------- response 3 10/29 14.35 zielinski csa What is the Prize? All contests must have a prize, said Pooh. ---------- note 285 bulletn bd 10/27 11.59 friedman csa In teaching the Computer Science courses on Plato, we have felt the need for an instructor-to-student communication system similar in concept to a Bulletin Board: a means of leaving messages for a more-or-less long period (more than one signon), addressed to the entire class. For example, these would include notices of upcoming exams, announcement of unusual Plato down times, etc.↓ An attempt to fill this need is now available in lessons csmsg1 and csmsg2. Lesson csmsg1 will be accessible by a single keypress within the cs router; it displays the notices to the student, based on his signon course. (If his signon course is not known to csmsg1, it asks him what course he is interested in.) Lesson csmsg2 contains the same display functions as csmsg1, plus a graphical editor for creating the messages. (The editor does not yet include a few intented bells and whistles, but it's almost complete.)↓ (continued) ---------- response 1 10/27 12.06 friedman csa The purpose of this notice is (1) to invite inspection of csmsg1, and especially of csmsg2 (the editor); and (2) to suggest that a system facility of this general type is a desirable goal.↓ To facilitate inspection of the editor, two fictitional "course" entries have been created named courses "cstest" and "sample". "cstest" contains a few examples of what the editor is capable of. Interested persons are invited to edit messages for course "sample", as long as common holds out; its editing security code is "tryit".↓ Comments and suggestions are welcome; they may be sent by pnote to the above signon. ---------- note 287 usersin 10/27 12.35 friedman csa When writing a group of related lessons, especially lessons which share the same common, it would be highly desirable to have an answer to the question "how many users are currently in lesson _x_". Now, I know that the "usersin" system variable can tell me how many users there are in the lesson I am executing, but I need other lessons at the same time. Also, I know how to do this myself by putting a suitable counter in the shared common, but this is subject to "ghost" errors (e.g., if a user leaves the lesson by an execution error), and it does take up a fair bit of code--in each lesson. Could we have a system means of determining either the number of users in all lessons sharing a common, or the number of users in a lesson whose name is supplied--say as the tag of a "usersin" command? How about it? ---------- response 1 10/28 01.11 michael english Ditto, mostly for the -common- case. How about a *usercin* variable, or is such information even kept by the system? ---------- response 2 10/28 04.20 koning csstaff It must; otherwise, how could it return common when the last user of common leaves? ---------- response 3 11/05 08.52 friedman csa See note ¬$483 for continuation. ---------- note 289 conderrpag 10/27 13.05 fritz ames I just had my lesson deleted from under me from the condense error page. I hit NEXT immediately at the page (it didn't even finish plotting the errors), and got "now preparing..." etc. I thought this was fixed... ---------- response 1 10/27 16.17 phil mast p So did we. Evidently pauses are still occurring during the jumpout to your lesson from condense error page. I have no idea when, if ever, this will be fixed. Phil ---------- note 293 prooftwo 10/27 15.23 gelder cerl Thanks to those of you who have examined the lessons of PROOF--proof1,proof2,..etc,and who have made helpful comments and criticisms. These are being taken to heart and adjustments are being made to correct errors and to locate frustrating program problems. Too, at the suggestion of several, a proof0 will be written to better introduce the student to what is expected of hin/her and to explain the rubric in which the proofs are cast. Look for a future note asking you to look at the lessons again. gelder. ---------- note 299 ¬, error 10/27 18.25 campanini mfl The illegal (unfortunately) use of the ¬, symbol in the tag of an -exactc- command used to be flagged as a condense error. On today's prime time version (10/27), it no longer causes a condense error, but does cause the -exactc- to reject any and all answers. If it is not possible to make the ¬, legal for all conditional statements, It's use of should at least be flagged as a condense error, as this could be a difficult debugger for a new programmer. ---------- response 1 10/27 19.51 blomme s The special delimiter used for -writec- commands should have its name reduced to "writec delimiter" I feel. Checking for its mis-use does indeed seem necessary and appropriate (it does require some special coding to recognize since it consists of two 6 bit character codes, one of which by itself is the comma). As far as using that delimiter even in -writec-'s, I might note that using a semicolon or simply an end-of-line seems simpler and more reasonable. The only case requiring the use of the special delimiter is that you have entries containing semi-colons and also ones that take up more than one line (so that end-of-line cannot be used as end-of-entry). ---------- response 2 10/27 21.09 judy pso I respectfully beg to differ! I _always_ use ¬,'s because it drives me NUTS to have 35 tags nicely typed, and then realize that the 36th tag needs to use two lines or a comma or a semi-colon. ---------- response 3 10/27 21.17 judy pso And while we're on the subject of delimiters, this is my pet wish-it-were command: series 1¬,neg tag¬,zero tag¬,one tag¬,big tag Which would be used like this: writec (expression);series,1 $$ or -writes-? arrow 101 answer (expression);series,1 $$ -answers-? There are soooo many places where I have a long -writec- followed by an equally long -answerc- where the tags of the two commands are identical. ---------- response 4 10/27 23.17 michael english I must agree with Judy. The "¬," is the only character on the keyset that can reasonably be used as a delimiter in -writec- statements without having to worry about needing to use punctuation or more than one line. The language- oriented projects experience this problem the most. It is regretable that there is no wired-in 6-bit character expressly for this purpose, but delimiters are such an artificiality it is hard to remember which are legal in what commands. The idea of a "universal delimiter" is basically a sound one, and helpful to both new and ex- perienced authors. If it is not too difficult, it should be retained and even expanded. [If necessary, it would even be possible to switch keycode assignments with one of the lesser used 6-bit keys, perhaps the SHIFT/ASSIGN key, o70. I can here the yells now.....] ---------- response 5 10/28 02.02 midden p Question michael: six bit codes, o70? Wait, doesn't o70 mean shift? o7001="A". respectfully yours, Marshall Midden ---------- response 6 10/28 08.47 michael english The last line was a joke, Marshall..... ---------- response 7 10/28 10.44 rick mflu About the series answers-- This could be accomplished with useful (but still non- existant) commands such as -packc- and -answerv-: packc problem,length,answer¬,a horse¬,three dogs¬, nothing¬,a cat¬,¬, * * the delimiter between answers could also be * a comma or the end of the line, similar to -writec- * do question $$ place question on the screen arrow answerv answer,length * * if "length" were omitted, the first o00 would * be tagged as the end of the answer * no write ntries>2¬,The answer is "¬0a,answer,length¬1"¬,¬, ---------- response 8 10/28 12.26 friedman csa As a matter of fact, end-of-line is NOT a delimiter for -writec-, nor should it be; I have coded many -writec-s in which some or all of the responses were multiple line, with either commas or ¬,s as delimiters.↓ I agree with Judy about the use of ¬, in -writec-s: I use it habitually because I'm tired of going back and changing delimiters upon deciding that I now want a comma in one of the tags. Semicolon is no help; I may decide I want a semicolon in the middle of a tag.↓ I also agree with the desirablity of making ¬, a truly "universal" delimiter: it ought to be accepted by every conditional command, at least as an alternative to the comma or semicolon usually prescribed. For example, it's a damn nuisance to remember that -block- needs commas and -transfr- needs semicolons between the principle tags; if both accepted ¬,.... ---------- response 9 10/28 13.18 tenczar s 2c≤/...I do intend to fix answer judging to use ¬, as a delimiter...I just thought it more important to get the -answerc- command out soonest than worry about the ¬,. My current work with punctuation will fix the ¬, problem in answer judging as a fallout. By the way, there will still be problems in many commands even when all are fixed to use the "ununiveral" delimiter...because many commands need to use two levels of delimiters..e.g., the -answerc- command currently uses ";" for major cases and "," for separation of synonymous answers within a case! And the -draw- command uses ";" between endpoints and "," to separate the x,y coordinates. Etc. ---------- response 10 10/28 19.41 koning csstaff I suggest ¬, as major delimiter and "," as minor delimiter, with the provision that if there is only one level of delimiter, comma can be used. And perhaps ";" as legal alternative for ¬,. ¬↑ˇP¬≤↑ˇK ---------- response 11 10/28 20.19 blomme s Except for the special case of entries in a -writec- which may contain ";" (for which we have provided a special delimiter), I'm afraid I don't see what this fuss is all about. Are people really saying that they would prefer making two key inputs and producing a meaningless symbol for EVERY command in the language in order to be consistent with a SPECIAL case of the -writec-? P.S. end-of-line indeed definitely WILL function as a separator in -writec- if that is what follows the conditional expression (and of course this is not useful in the case where any of the entries take multiple lines). ---------- response 12 10/29 03.04 tenczar s I am afraid I agree with Blomme (for once)...the "¬," delimitor's use surely is limited to cases where the user might get into trouble because of the "freedom" allowed in the rest of the tag...namely in tags where the user has the freedom of inputting anything on the keyset...namely, the -write- (and its derivitives) and the -answer- (and its derivitives)...has anyone even considered using the "¬," in such tags as: spec nospell¬,nomark¬,allwords inhibit erase¬,arrow jkey help¬,back¬,lab abort common¬,records the tag format in most commands is entirely rigid...commas can mean only one thing...and the comma is easier to use and looks better and prints better! The "¬," is misnamed! Who has a better name. ---------- response 13 10/29 10.17 friedman csa Ok, I see the point about ¬, with the examples cited in the previous response, but I (for one) was thinking more of commands with conditional forms. I still advocate making ¬, a universal delimiter optionally replacing the highest level delimiter (; in -answerc-, comma in -writec-) in all conditional commands. One advantage is that one could then just change the command name to have, for example, a -writec- and an -answerc- with the same texts. The most obvious advantage is that one doesn't have to remember which delimiter applies to which command. [In that reagrd, -block- should accept at least semicolons, if not ¬,, in place of commas, to be like -transfer-.] Also, the language then acquires a symmetry that any well designed language ought to have. ---------- note 301 readr rost 10/27 20.04 blomme s A change to the use of "zreturn" for the case of -readr roster- is planned within the next couple weeks. It will be changed to contain a status return; this will complete the process of establishing a consistent format for status return information in "zreturn" (as opposed to the rather varying forms given by the "error" variable). Now that the student count for a course can be obtained from the -readset- command, that use of "zreturn" will no longer be necessary. The change to the system will not be made until all lessons that use -readr roster- have been changed; the help of the authors of the lessons involved will be appreciated. ---------- response 1 10/28 11.40 berger mfl What do you want those of us who use readr roster to do? ---------- response 2 10/28 12.28 friedman csa Convert to the new 3-tag form of -readset-; see new features note -$92. ---------- note 302 -remove- 10/27 20.55 judy pso Perhaps I am far, far behind the times, but I know that AIDS has never documented this: In a random drill, where items are being selected from a -setperm- list by a -randp-, one often wants to -remove- the item from the -modperm- list IF and ONLY IF the student answered correctly on the first try. My customary device is this: answer correct one do (ntries=1),removit,x no * unit removit remove item This format is certainly handier: answer correct one remove -(ntries=1)*item If (ntries¬=1) then the () are zero and the whole expression becomes zero and _nothing_ is done. For (ntries=1) the expression ¬[-(ntries=1)*item¬] just becomes (1)*item. ---------- response 1 10/28 11.08 maggs law This is a useful trick a lot of places. However the code suggested is somewhat obscure. I usually do this sort of thing by first defining: define junk if(x,then,y,else,z)=-x≠y-not(x)≠z nothing=0 item=....... *then I can code remove if(ntries=0,then,item,else,nothing) ---------- response 2 10/28 12.29 friedman csa Nice coding, Maggs! There's hope for Tutor yet! ---------- response 3 10/28 13.29 judy pso Your example would be handy in the case where both possibilities are non-zero, but what a lot of coding for a simple case like "remove -(ntries=1)*item " ! You still have to know that the true case is -1 and the false case is 0 (in order to write the define) so why not just apply that knowledge? ---------- response 4 10/28 18.45 jones mcl This brings to mind something that I suggested about 9 months ago, the embedable conditional. The suggested format was in keeping with the conventional TUTOR conditional form, and was suggested in response to a note from someone who wanted a conditional at; instead of: remove if(ntries=1,then,gunk,else,nothing) I suggested: remove ¬0ntries=1,gunk,0¬1 I am glad that there is a way of doing this with a define, thanks for publishing it! Jones ---------- response 5 10/28 19.45 koning csstaff The reason for defining the 'if' rather than simply 'applying the knowledge' is the same as for using any define whatsoever: readbility! ---------- note 310 Data Key 10/28 09.35 bl abbott mtc Something I found by pure accident: while in REPLACE mode editing a block, if you press the DATA key, Plato changes you into INSERT mode in the block where you were editing. This is something I would have liked to know long ago, but it wasn't in the help sequence for block editing. Why Not? ---------- response 1 10/28 09.43 nagel vu Look at the help sequences for a block. Enter the block, press -HELP- , part 3 (change contents) about the second entry mentions the use of DATA to change to insert mode. ---------- response 2 10/28 19.47 koning csstaff Is it, perhaps, a new and unannounced feature? ---------- response 3 10/29 08.22 mike b cornell It is neither new, nor unannounced. See response ¬$1. M¬ ike≤≤≤≤ˇB ---------- note 312 mrouter? 10/28 10.17 jim pso how do i change the "change code" on a new type instructor file? how do i name existing¬ (prior to the new system feature note) modules of a new type instructor file? jim ghesquiere ---------- response 1 10/28 10.25 judy pso When you first enter a new-type instructor file, it DEMANDS that you fill in its datapage, right? After that, to change the security code, 1) enter the course to which the file is attached 2) choose ¬$4 (curriculum design) 3) on the "curriculum options" page, again choose ¬$4 ---------- response 2 10/28 10.36 chabay s SHIFT-LAB while editing a module will let you change its name. (It is on the help page there, but perhaps that's a bit obscure....I'll try to put in some more help.) ---------- response 3 10/28 13.52 whisenhunt siu Is there any way to copy an instructor file from another file after the information has already been added? (I tried zeroing everything!) Gary Whisenhunt ---------- response 4 10/28 15.49 chabay s Currently the only time when you can copy another instructor file is when you first initialize you instructor file. This will be fixed -- in the meantime, you will have to destroy and re-create your instructor file to do this. ---------- response 5 10/28 22.25 whisenhunt siu OK thanks! GW ---------- note 314 freeze 10/28 10.26 judy pso I assume that the systems personnel know that terminals are "freezing" and then coming unstuck after a minute or two? ---------- response 1 10/28 10.55 maggs law I got a complaint about this on one of the law terminals -- complaint was apparently filed late yesterday. ---------- response 2 10/28 15.21 fritz ames This happens to us at Ames quite often after the system comes up (ptime OR non-ptime versions). It's not even surprising anymore... you just sit and wait for a few minutes for it to unstick. Ah well, it's better than the 7-hour freezes we had last summer! ---------- response 3 10/29 13.28 rader s I don't believe that we are aware of either of these phenomena--thank you for putting in the note, Judy. I'm not sure what to ask for--some statistics as to where and when and how frequently these things are happening would be a help. So--if you are experiencing such problems let me know about it--site/station numbers, times of day and extenuating circumstances are all appreciated. (We did, of course, have a serious problem with hanging terminals on Oct. 27th--but those didn't come unstuck, and were quite a different problem.) ---------- note 324 hangup 10/28 12.56 kienitz ed317 I was hung up twice on 13-? on the author page. I managed to sign on at another terminal and kick myself off of the first one, but it was unavailable for anyone else to use after that still. Will this be fixed soon? I don't like to have to leave while I'm still signed on ---------- response 1 10/29 10.14 fay rtv362 If it does occur and you cannot back out, but desire to leave, attempt to relay the info to an operator (via another signon or thru another user). We will then attempt to get it unhung or get systems personnel to do so. Yes, systems is working on the problem. It is known and every effort is being made to minimize its occurence. ---------- note 325 disks/fire 10/28 13.00 gilpin peer In the benighted part of the world where programs are still stored on cards. really valuable programs and data are dup- licated and the copies stored in separate places, against the possiblity of loss by fire. The thought crossed my mind today of what a horrendous catastrophe would be wrought if the PLATO disk packs were visited by fire and what options there are for protection against same. / Since there are now disk drives in two different rooms separated by two floors of relatively fireproof building, one simple measure would be to store backups and originals in differ- ent rooms. Quite possibly a precaution of this sort is already in effect. If not, is it something that would be feasible to do? ---------- response 1 10/28 13.23 judy pso Archival (sp?) disk copies of the PLATO system are stored in another _building_, in addition to the usual backups which are done frequently (daily?). I am not sure how often "archival" backups are made. ---------- response 2 10/28 14.09 celia pso It would seem to me that the most _recent_ backups should be stored in the other building. If the archival copies are more than a month or so old, they are probably next to useless. ---------- response 3 10/28 16.54 stan smith chem Gee I hope that lessons more than a month old are not next to useless. My class is running today on a lesson which I wrote in 1971. ---------- response 4 10/29 13.32 rader s Archives are stored in another building; we intend to store a weekly backup set in a separate room also. ---------- note 331 char many 10/28 13.50 me smithy mtc I purpose the following. (formal arn't i) 1.I tried to convert a 43 character charset to char commands and it told me that it was to long so i deleted a few than tried again, still to long. Went into aids to look at what the limit was, no such luck. Tried the help options on the charset option page, nothing but nerts. So after wearing my finger out trying to ask people after finely finding the right person (thanks to cnb) I thought it would be nice if aids and the "Sorry to long" could say something allittle more indepth. LIKE HOW MANY (limit) ---------- response 1 10/28 15.09 tenczar s a comment on a slightly perpendicular matter... did you know that the use of -char- commands is to be discouraged since the command uses lots of condensing time (all the numbers have to be constantly chewed up), lots more execution time and poor use of ECS...in comparison to -charsets- The exceptions... ...where you need only a few characters (few=less than 10) ...where you need to manage the characters that are loaded into the terminal because you need more than one charsets worth or you have your own editor or whatever... ---------- response 2 10/28 15.33 fritz ames Well, since the char commands produced by the charset editor are made up of uniform-length octal tags of 8 words each, it would seem logical to assume that the limit is int(322/8) or 40 chars (which is, indeed, the case). ---------- response 3 10/28 17.37 berger mfl But the users that tutor is intended for (dumb instructor types) don't know that... ---------- response 4 10/28 19.53 koning csstaff I have wanted to convert a charset with 122 characters to char commands once. True, it was an unusual application, but still... By the way, for those freaks who want to do this, wouldn't a charset->common come in handy? ---------- response 5 10/29 09.38 mcneil a uicc Yeah. I once wanted to underline a whole set of chars. If Id had common¬Dchars¬Dcommon conversion around it would have been a snap. ---------- note 340 comeditor 10/28 17.00 a appel uni For some reason, the common editor (not the plain old fashioned inspector) ignores font (o75). Why does it do this? Could it be fixed? Why not? ---------- response 1 10/28 17.33 warner iu The systems staff uses a command called -clean- to convert the alpha representations of common to a "more readable" form by removing fonts, accesses, backspaces, shifts, and the like. UNFORTUANTELY... it converts them all to spaces, so you still have to look in the octal to find out what the word really looks like. A while back, Andrew Appel (I think, or someone else in uni) had a nice routine that converted accesses, fonts, backspaces, etc, into nice little -char- pictures. It was used in conjunction with the fad for little faces O≤M≤I≤X O≤W≤I≤X≤A that suddenly appeared all over pad. Perhaps a systems version of this would be nice in the common editor? ---------- note 350 mail truck 10/28 20.27 saintey cs196 what ever happened to that little mail truck that would come on the screen whenever you got notes? i really thought that was neat, and many people agree that it should be brought back. any comments? ---------- response 1 10/28 20.49 dave fuller uimc It takes too much time, and screws up your already loaded charset. Uh uh... ---------- response 2 10/29 08.29 michael matha The current, simple "NOTES" will do nicely, thank you. J. M. Felty ---------- response 3 10/29 09.28 k mast p The truck is still alive and well and living in lesson 'holiday'. It will be brought back during "Be kind to your friendly Pnotemaster week". ---------- response 4 10/29 09.38 hinton ssu I almost hate to bring this up, but I thought everybody agreed not to send -ext-s to the slidelight any more.... ---------- response 5 10/29 09.41 friedman csa Please! Lay off loading chars...as stated, it screws up the loaded charset. Maybe briefly on special occasions, but not routinely! ---------- response 6 10/29 15.52 bl abbott mtc the little letter screws up charsets also if your notes course limit is too high ---------- response 7 10/29 19.44 k mast p That is intentional. The purpose of the little letter is to make you hate it so much that you want to get rid of it. ---------- note 351 termop 10/28 20.34 dowling ames I recall recently in g-notes discussion on a possible change of "term" to behave (possibly optionally) as an on-page help unit, but cannot locate this discussion. Could anyone inform me of (preferably) the status of the proposed change or (as a second choice) the location of the previous discussion on this? Thanks W¬≤↑ˇC¬≤↑ˇD ---------- response 1 10/29 08.33 mike b cornell The discussion is in Old Help Notes ¬$96. Briefly restated, the systems people said that the suggested change corresponded with their own thinking and that it would be done--within a year or so. M¬ ike≤≤≤≤ˇB ---------- note 353 comload 10/28 20.53 john r reading It is a known feature of TUTOR that the common work variables are not zeroed except at the beginning of a time- slice. Because of this it is possible to write reasonably conservative code that fails only under certain system load conditions. An illustration of this is given in the first response to this note. I suggest that -comload- should always zero the common work variables before executing. In my opinion, the overhead would be more than justified because it would insure that to-be-published lessons which function now would not fail under some as-yet-unknown timeslicing conventions in the future. ---------- response 1 10/28 21.01 john r reading * example for previous page comload nc1,1,19 $$ table of tables nc20,nc2,nc3 $$ desired table of numbers calc n1‎nc(nc20) $$ ok because table loaded * * ¬X this is buried deep in a -do-ne unit comload nc1,300,10 $$ some other table * ¬W * at this point the second part of the first comload may * or may not be 'in effect'. the common vars beyond nc19 * will retain their old values until the end of the current * timeslice. an attempted calc n1‎nc(nc20) * may or may not work depending on current length of time- * slice - and that seems to vary depending on system load. ---------- response 2 10/29 00.12 koning csstaff I would suggest that work variables NEVER be zeroed, in order to save 150 ¬ms per timeslice..... ---------- response 3 10/29 17.39 b sherwood s Interesting point, John. Hmmmmm and ugghhh. As to NEVER zeroing the work variables (to save 150¬ms), that is not possible for security reasons -- that would let people read other people's common and storage. ---------- response 4 10/29 17.55 koning csstaff How about zeroing it only if the next person uses nc/vc variables? That way, ;essons that don't use common at all, and therefore could be no security threat, are not given this extra overhead. ---------- response 5 10/30 10.54 b sherwood phys That is possible. ---------- response 6 10/30 17.57 dave infe But you don't know whose common and storage you're reading; who would object to allowing their stuff to be read anonymously? At any rate, the cost of -comload- is high enough already; those who do a lot of them would hate to see the situation made worse... ---------- response 7 10/31 09.10 white p Inspect "anotes6", block 1-g if you have any doubts about the value of the system zeroing *nc* variables every timeslice. ---------- note 355 0 ¬= 0 10/28 21.35 parrello uimatha Put this in your TERM calc and smoke it (but not if there are police in the vicinity): '0;'≠0 = '0;' Moral: be careful about what kinds of garbage you have in variables that are to be "ignored" by multiplying by 0. I know a lot of mathematicians who are going to be upset by this new development in the field (of real numbers). ---------- response 1 10/28 23.24 clark lawyer Also, note that typing "3≠4;anything", results in 12 and not unrecognized variables name. This works not only in term calc, but in TUTOR lessons. calc v1‎3≠4;anything Results in condense error: calc v1‎3≠_4_;anything non-recognized end of line System error (calce1, error ¬$46) and not undefined var error! Does a semi cause PLATO to stop processing the calc? Systems? ---------- response 2 10/29 09.36 hinton ssu Yet further: CALC these versions of RS's equation: 'o;' or 'o;'≠0 and see what nice numbers you get.... ---------- response 3 10/29 09.45 elston rhrc Of course if you show it in a lesson you get 0/0 and as we all know In infinity all things are possible...... 2c≤/ ---------- response 4 10/29 13.06 curly iu Alos note 1/'o;'=1/0 ---------- response 5 10/29 17.56 koning csstaff The point is that 'o;' puts garbage in the exponent..... ---------- response 6 10/29 18.22 parrello uimatha The point is that I _can't_ believe that 0/0 is an integer value, and yet when it is in n1 it gets treated as such. RS ---------- note 358 hidn chars 10/28 23.15 dowling ames The present error display notes "hidden characters", and upon request displays them. To achieve this display, several font characters are loaded. Anyone using this feature will find his font characters modified upon return to the editor, and will most probably have to wait through a character reload upon the next execution of the lesson. Would it be possible to replace these font chars with composite chars from the standard character set? some suggestions are: <≤>, ¬0≤¬1, H≤I, ↑≠≤↑≡, ¬I≤=, <≤X, ¬A≤¬D, ¬A≤¬,, m≤w≤o≤a≤=. If this were implemented, I would also propose that this expansion be made available in the editor, and that while expanded, the line be editable, with each expanded code treated as a single character. This would GREATLY simplify the greatest editing headaches, lines containing hidden characters in which erase eats up more than desired.¬↑ˇ≤≤W¬≤↑ˇO¬≤ˇD ---------- response 1 10/29 08.37 jim pso looking into it.... ---------- response 2 10/29 18.13 broadus css I think you could just make it in the following way and fake every body into thinking that it is still the old way. ¬I≤ˆ¬I≤¬ˆˇ_ Bob Broadus ---------- note 375 searchstop 10/29 10.40 walter physio It would save user time if the STOP key were able to halt searching in the X-search accessible from the block page. --------- response 1 10/29 11.21 battenberg css hERE----hERE-----I agree----this is an excelent idea!!!!! ---BRAVO!!----ΔΔ---BRAVO!!!!-----this would save alot of time------------the king of hearts----rex ---------- response 2 10/29 11.59 michael english Talk about jumping on the bandwagon..... STOP already does terminate the search, turkey..... ---------- response 3 10/29 13.44 kravitz cerl It didn't for me must a few moments ago. Stop1 did, but it left me on the author mode page. Pressing next resulted in no action. ---------- response 4 10/29 14.10 steinberg mathw stop does stop the search, but it usually takes a while to be noticed. ---------- response 5 10/29 16.04 berger mfl Are you guys talking about the same thing? I think the original note referred to lesson "search" where STOP does not work, rather than the editor. ---------- response 6 10/29 16.14 michael english STOP terminates↓ 1) The X option from the author mode page. 2) The X option from the block display page. 3) The inter-block X option from the line display page.↓ 1 ¬+ 2 drop the key in a long, fruitless search. ---------- note 376 job avail 10/29 10.43 stull med JOB AVAILABLE↑ˇ Description: Full-time, 12 mo. appt. for Specialist in Auto- mated Ed., Computer-based Ed. Proj., School of Basic Med Sciences at Urbana-Champaign↑ˇ Duties: PLATO computer programming of health science lessons↑ˇ Qual.: Bachelor's degree required, Master's preferred. Two years' experience on PLATO is recommended. Must be Tutor expert ¬+ familiar w/ one other language.↑ˇ Salary: Competitive↑ˇ Position Avail.: Immediately↑ˇ Applications may be picked up at 1115 1/2 W. Oregon and must be returned no later than 4 pm Monday November 3, l975. For additional information address correspondence to stull/med via p-notes or call Dr. Richard Stull, Medical Author Group, 333-2507.↑ˇ The U. of I. is an Affirmative Action - Equal Opportunity Employer. ---------- response 1 10/30 11.08 mosley cstate Good deal but this note should have been entered in lesson "wantad". ---------- note 377 courseware 10/29 10.53 golden s The courseware notesfile, (lesson courseware), is now full. I shall leave it for users to read or reread, but I don't think it will be necessary to carry on additional public discussions until some new action by the University of CDC occurs. Thank you for your contributions. Additional thoughts are always welcome via personal notes. ---------- note 378 moor bye 10/29 10.58 berger mfl *** SPECIAL ANNOUNCEMENT *** Dave Moor, a long-time Language Laboratory employee, is leaving for a new position in California. The Language Laboratory is having a party for him on Friday, October 31, at 3:00 P.M. in room G-69 FLB. We hope that those that have known him will attend. Please bring goodies for everyone to eat (optional of course). No reservations necessary. See you Friday! The MFL recreation committee ******** ---------- response 1 10/29 18.12 wood medsiuc This is a moornful occassion. ---------- response 2 10/29 18.22 berger mfl Let's not have any moor responses like that! ---------- response 3 10/29 19.45 koning csstaff ***shiver*** ---------- response 4 10/30 08.26 nate iumusic let's see...now when we want to clear this general note, I will have to receive communication telling me to delete this note, then koning will have to delete his note, then berger, then koning, and finally berger....hmmmmm.... Any more people joining on the chain of archival notes? (that may get requested to delete his response) nice going! Nate ---------- note 397 Allerton 10/29 18.53 carter comm Our Second Annual Allerton Bike Ride will take place Saturday, November 1st--weather permitting. We will meet on the corner of Wright and Green at 9am. Bring a sack lunch. Distance is about 60 miles. Anyone that wishes to come is invited. ---------- note 405 -contest- 10/30 12.34 fay o ---RE-ANNOUNCING THE PLATO DOWN TAPES CONTEST--- *** Just a gentle reminder, folks. You have about 6 days left to get your favorite drawing/animation entered into the PLATO down tapes display contest. See lesson "tapedisp" in the author mode (preferably inspect mode) for rules and prizes. Tim ---------- response 1 10/31 08.45 olson ced I had an entry in "tapedisp" , but when I went in to make some changes, it was gone....does this mean it was already judged and rejected/accepted? ---------- response 2 10/31 09.19 fay o No. It means we have moved the entries to a codeworded file a) to make room for new entries and b) to avoid the possibility of some unbalanced soul re-initializing the file for us. The winning entries _will_ be made available to all users (so you don't have to wait until we go down to see them!). ---------- response 3 10/31 09.19 errol kka The "judgers" have moved the entries to another file to make room for more entries. I'm sure yours is ok. ---------- note 406 delete,my 10/30 13.47 powers ames What's the deal? I thought you weren't supposed to get deleted from library--I know ECS is tight, but give a rookie a chance! ≤≤≤≤Kit Powers ---------- note 407 job 10/30 14.22 j predmore mtc Plato programmer needed Chanute needs a full-time, experienced, plato programmer. Pay is about $10,000 per anum. Contact j predmore of mtc for more information. ---------- response 1 10/31 10.25 weible german Query for the day: What is the nominative form of "anum"? ---------- response 2 11/01 14.13 koning csstaff That depends on the gender... ---------- note 408 fish+slide 10/30 16.01 george tech An informal meeting of microfiche users and the CERL microfiche staff was held on Monday, October 27↑ˆth↑ˇ. Slide selectors and microfiche were two of several topics discussed at the meeting. The meeting has been outlined in the microfiche area of the lesson "peripheral". All interested parties are invited to respond in lesson "peripheral". george traynor ---------- note 409 signoninfo 10/30 16.37 meers wright An instructor has no way of seeing when his records were last used and cannot tell if someone might possibly be using those records. Could instructors get the same type of display that authors get the first time they see the author mode page? ---------- response 1 10/30 17.34 chabay s An instructor, like a student, can see information on use of his records by pressing LAB (for additional info) off the password page. It seems desirable to keep the signon displays for instructors as simple as possible, and not clutter displays with extraneous information which may be confusing to novices. ---------- response 2 10/30 18.47 meers wright Thanks.. ---------- response 3 10/30 20.09 whisenhunt siu Can a student see his records? I just tried a student signon and all it did was to change the password. It might be a good idea to include that somewhere for students, since the records they use may determine their grade in a course. GW ---------- response 4 10/30 21.58 chabay s Hmmmmm... I thought a student could see info on his records, but I guess not. Do you really think that it is important? (Perhaps it's better the way it is -- students get confused if they are given too many options.) ---------- response 5 10/31 08.01 fay o I'll second that, Ruth, and invite anyone who thinks it is not true that students get very confused about "computer stuff" to come and sit in the operator's office for a day (preferably a weekday between, say, 10 am and 4 pm). ***"I can't get into my course" ***"How do you turn off the machine" ***"It won't take my answer" ***"Who am I and why am I here" ***"It says I'm not registered" ***"It says my memory is excessive" ***"What is the meaning of life" ---to quote but a few of the "confusions". ---------- response 6 10/31 09.26 meers wright People do use othe people's signons without the first persons knowledge you know.. The problem has been largely eliminated for authors by simply displaying the 'last time on' info the first time they see the main author mode page. Instead of a confusing option, it would be written out for instructors (maybe students also, but at least instructors) at the bottom of their first display. ---------- response 7 10/31 10.19 berger mfl For the archives: Fay forgot the classic question that we hear all the time from students: "What is my password?" ---------- response 8 10/31 14.44 whisenhunt siu AMEN! ---------- response 9 10/31 18.54 parrello uimatha You missed the most common, frustrating, annoying, and painful to answer one: "What's the password to get into the games?" ---------- note 410 loop bug? 10/30 17.17 bonnie matha Just a minute ago I was Uing and shift-Uing and shift- minusing in the editor, when my terminal got in a loop. It would display 5 lines, full screen erase, display the same 5 lines, erase..... I pressed keys, but even shift- stop was ignored. Then a key got in I guess and I had control again. Several days ago during moonwar (a demo, of course) the Big Board display did the same thing. Soon everyone left the lesson. Shift-stop got me out. I assumed it was a bug in the lesson until it happened today in the editor. ----------- response 1 10/30 17.26 kemp uicc The same type of thing happened to me, except i was in Help Notes, one line cycled about 30-35 times before it stop. Is Plato have a nervous breakdown!? ---------- response 2 10/30 17.28 midden p see response to next note. Hopefully things like this won't occur except on rare occasions... ---------- note 411 edit loop 10/30 17.18 ka lutz usc something strange just happened while i was editing. I was in replace mode and was using the copy key, when suddenly the display went into a continuous loop. The code was plotted, then the arrow in the usual way for replace mode but then the cycle would repeat. It went though about 30 cycles. I hit back a few times then shift stop (on about the 15th repition). The lesson finally condensed and then i was able to hit shift stop again and go back to editing. What happened? kathy ---------- response 1 10/30 17.27 midden p Sorry, a background job on the system took the cpu away from plato. This symptom is one of not enough cpu time. Hopefully only on rare instances will this happen. (like today's!) Marshall Midden ---------- response 2 10/30 20.47 daleske ames This problem has been constantly occuring here at ames. It has, however, occured usually about 5-10 minutes before a system "crash", thus, the crash has been suspected as the main cause. Do you want us to further report this situation? The problem of the post-crash stickeys is seemingly along the same lines; input is suspended for 1-10 minutes usually within one minute of a system re-load. Before the crash, input is suspended, but the last output is re-sent continuously. Should this be further reported also? John ---------- response 3 10/31 12.37 rader s I believe we understand why you get repeated output as PLATO goes down--the same output buffers happen to get sent again and again. I don't know of any reason for you not to be able to get in when PLATO comes back up, however--I would appreciate info such as site/station, time and duration of occurrences. Thanks. ---------- note 421 concepfast 10/31 00.35 koning csstaff Is -concept- really much faster now? Aids seems to work at unprecedented speed! If this is not my imagination, congratulations to the sysprogs... ---------- response 1 10/31 00.56 tenczar s no, concept doesn't work any faster (can't get much faster than 1-2 milliseconds for 1000+ vocabulary plus 100's of concepts!) other things have been changed in the system ---------- response 2 10/31 08.58 jim pso over the last few months, many things were changed within AIDS. few, if any, are noticeable "on the surface". all the changes are aimed at making AIDS faster and having better cross references to other parts of AIDS. this work is almost complete (for this round). ---------- note 424 problems 10/31 08.09 straayer telcomc Various notes say things like "this is a known system problem to be fixed some day." Is there, or could there be a reference lesson of command problems with stats of fixes, like "will be working 12-1-75", "will be fixed in 2 months", "will never ever work correctly", and so on. Example problems include loops when cpu usage is high, lack of good control for multiple simultaneous users of a single common, "+" and such keys working variously as both word characters and punctuation. A list of such items would therefore be a summary of problems and requests found in "help" notes that elicit system programmer action. ---------- response 1 10/31 09.10 golden s A list of known problems is a good idea. I doubt that we could attach dates for changes. ---------- response 2 11/01 14.17 koning csstaff You could split the list in two parts: existing problems and fixed problems, and move things from one list to the other as appropriate. ---------- note 434 -rcircle- 10/31 10.41 white p Those interested in proposed changes to the -circle- command and the proposed addition of an -rcircle- command may participate in a discussion in lesson "rcircle". ---------- note 435 printer 10/31 11.03 maggs law I would like to raise here in a more general forum a question I raised earlier at a PLATO Users' Committee Meeting. This is the question of providing at CERL a high-speed high quality printer capable of printing the upper and lower case letters, the numbers, and normal punctuation. This would allow the vast majority of us who do not have Terminet or other expensive gadgets to print out nice reports. I do not have in mind the combination of low quality and high flexibility that we get on Varian printouts; I would tolerate low flexibility in exchange for high quality. I would hope that either CDC has a stock upper and lower case printer train, or a link could be established with the IBM computer at CSO to enable use of its upper and lower case train. ---------- response 1 10/31 13.11 white p I doubt if it is worth it for CERL to buy a good quality printer with upper and lower case just for the printing of reports. I have used Plato for entering 3 or 4 reports that were later printed on the local 360 in upper and lower case. However, the mechanism for transfering the files was poor enough, and the DECWRITER IIs that CSO has are good enough, that I am now using a terminal in DCL for entering reports. There is an experimental link between Plato and the DEC10 at CSO, you might want to check on the status of that project. ---------- response 2 10/31 13.19 rader s CDC supplies chain printers similar to the IBM 1403, called a model 580. The price is $61K; there are some six different chains to choose from--and the option of custom made chains. The chains cost about $3500. For under $10K, we could get a dot matrix printer that will do listings (with full charsets, etc) and other chores. We have not felt that we could justify the expense of the chain printer. We could fund development of a replacement copier for the Varian (attachable to any terminal), a system printer/plotter (a replacement perhaps for the line printer and certainly for the Varian plotter), and a typewriter-at- your-terminal using a real typewriter, with rather less than that amount of money. If we had it. ---------- response 3 10/31 13.21 frankel p We have in fact been investigating printers for this purpose on and off for quite a while. We were thinking of going another step further in terms of print quality...IBM Selectric or something on that order. We had an IBM rep visit last spring with one of their word-processing machines; it cost $10,000 and was rather more than we needed. There are, however, other machines available, such as the Xerox/Diablo Hi-Type I (and II) for less money. We are hoping to purchase such a machine when funds become available. Our goal is to be able to edit reports, etc. on line, and print them out (perhaps slowly) in a form ready for the printers. Hopefully some action will be taken before the end of the year (1975). I can probably give you more info if you want it. ---------- response 4 10/31 16.10 maggs law I was thinking of replacing the regular printer you use now for printouts with an upper and lower case chain printer. This would give almost everybody better printouts. ---------- response 5 10/31 16.39 dave fuller uimc Maybe everyone could chip in and get that heap of ... that CDC so fondly calls a printer a good sendoff-- 61K, Bob, now, that's just $64.21 per terminal... ---------- response 6 10/31 23.37 straayer telcomc Isn't there also a need for software to support printers if and/or when a standard plato offered printer becomes available to all? Are there any efforts in this direction? It seems for now the "system" policy is to let this area lay undeveloped in favor of other projects. ---------- response 7 11/01 02.25 hody med hey let's be reasonable! they can't do everything!! ---------- response 8 11/01 10.23 berger mfl Besides, how are you going to develop usable software when the hardware isn't even on the drawing board! ---------- response 9 11/02 02.13 al mfl If we could wait until 1980, that $61K would only be about $.061 per terminal! ---------- note 436 new course 10/31 11.14 pat tech As of October 27^t^h several authors in course cerl were transferred to course tech. Some of those authors are: dennis s george pat streff They are all now registered in course tech. Microfiche users and persons interested in microfiche should now contact george of tech and/or pat of tech for information on PLATO microfiche. ---------- response 1 10/31 11.25 jim pso a list of CERL people (including sign-ons) is maintained in the author resources section of AIDS. ---------- note 439 readr info 10/31 12.40 michael english It would be quite useful if the the restart information was added to the list of things returned by a -readr stats-. -------- response 1 11/01 09.25 b sherwood s Sounds logical. ---------- note 443 vocabs 10/31 13.58 mcneil a uicc In the things to do that never get done dept- When one creates a vocab block it is put at the end of the lesson. Vocabs have to be at the top of the lesson (before concept commands). Please don't put the vocabs at the end of the lesson. AL9000 ---------- response 1 10/31 14.14 woolley p You're right, this is something that definitely needs to be done and probably would not be very hard to do. Therefore, you should not expect any action on it within your lifetime. ---------- response 2 10/31 14.48 michael english Please. instead finish the plans so you can invoke a -vocabs- like a -use- without charging for a -use-. Or better yet, so -vocabs- could be done like a -micro-, as many as are needed in a lesson. Maybe by 2001? ---------- response 3 10/31 15.19 hinton ssu Well, whichever priority you-all choose or don't choose, please do something about the problem of putting a Vocabs block last, then not being able to use it in a lesson ! ndh ---------- response 4 10/31 20.10 silver ve The vocabs block cant be -used-, anyhow, unless it is converted back to a regular block. So it is a nice editing convenience, mostly. ---------- response 5 10/31 22.51 iezek ames Will something ever be done about this or will systems just keep making jokes on it? Mark Iezek ---------- response 6 11/01 10.38 little i research yes.... ---------- note 456 no DI? 11/01 09.40 todd mflu good day.... for all those who did not receive DI's this cheery morning....sorry.....I could not find several of those elusive engineering buildings....being a biology major myself and never wandering north of Green 'cept to gaze at the wondrous plasma panel.....I could not find the Civil Engineering Research Laboratory, and the Electrical Engineering Research Laboratory was locked.... ......if you did not receive your DI call 3-7411 and tell them your name and address (where the paper should appear) and they will bring you the paper this afternoon or monday.....(also tell them where the building is!!!) again...sorry.....todd/mflu ---------- note 460 what pause 11/01 14.15 crooks siu If anyone is interested in another systems mistake condense lesson 'spanish2'...you should have to press NEXT to pass the pause, you say...NOT SO!!... Observe: at 1010 write Please enter lesson Spanish2 through lesson Spanish1 pause press stop1 What the ____ is going on?? magician ---------- response 1 11/01 14.34 white p You -press next- in the ieu, which takes you through the -pause-. Next time you have a problem like this, calling a consultant would solve your problem in a minute or two, or at least blank out your inspect code so that the many knowledgeable authors that read notes could help you. ---------- note 464 X-srh err 11/02 09.46 brant ames I have a unit in one of my lessons which is not found by the complete lesson search from a block. i.e. U(somename) then press NEXT will find the unit if it is in the block I am in, but NEXT1 will not find the unit no matter what block I am in. Lesson name and unit name are available to systems people upon request. George Brant ---------- response 1 11/02 11.10 nate iumusic Is the block in which the unit name is concensed in condensed? If not, I think that has affects the lesson- wide X search. THINK that's the way it runs, Nate ---------- response 2 11/02 12.02 brant ames Thanks for the suggestion but the block is NOT partialed out. p.s. systems people, I have another unit in the same lesson which also cannot be found by the whole lesson U-search from within a block. George Brant ---------- response 3 11/02 13.01 frankel p Leave the lesson name and unit name in a pnote to me. ---------- response 4 11/02 21.18 warner iu One possiblility: If the unit you are searching for just _happens_ to be on line 1 of your display, then neither the NEXT-type search or the NEXT1-type lesson search will find it. Maybe for this case, what is needed is a message similar to "not found" which flashes "look, dumdum!" ---------- note 468 reada? 11/02 14.40 stone pfw will the data that is available to account directors in lesson "account" be available with something similiar to a readd and readr command soon? or is it already...i would like to know... Scott Stone ---------- response 1 11/03 13.59 b sherwood s It isn't available now. Won't likely happen for a while. I know of some other kinds of displays that would be nice. What kinds of needs had you noticed? ---------- note 469 edit tabs 11/02 18.19 parrello uimatha Why, for the love of TUTOR, were the tabs changed in the editor? They are totally incompatible with the old ones, thus those of us who have not finished one or two of our lessons yet find that our comments won't line up. (This may make no difference to most of you but some of us actually DO like to make our coding comprehensible to others, especially if we've ever had to debug programs that the idiot authors didn't bother to document.) I don't mind the addition of a close tab in the tag that much, but would appreciate it if the third and fourth tabs lined up with the old thrid and fourth tabs. ---------- response 1 11/02 18.24 char mathw thrid ? ......... ---------- response 2 11/02 18.32 berger mfl I thought it wasn't my fault that things weren't lining up... I'd rather have them compatible with the old ones as well. ---------- response 3 11/02 18.35 parrello uimatha "third", blast it! ---------- response 4 11/02 19.54 keith s matha Looks like a tabset option in the editor would be useful (so that an author might specify any set of tabs). ---------- response 5 11/02 20.23 brand matha well personally i like the new tabs a lot better; it makes lining up comments easier than before, since there are no longer those big gaps..... ---------- response 6 11/02 21.35 little t research I also am in favor of a variable tab setting option that would work like T option now only you enter your own tabs. Todd ---------- response 7 11/02 22.48 parrello uimatha You can close the gaps all you want. We would just like the new tabs to include the old ones. ---------- response 8 11/03 02.42 sellers arizona Tenczar is experimenting with something to do with tabs. Come to sunny Arizona to get NEWS about the system. ---------- response 9 11/03 04.41 tenczar s not quite right...indenting ---------- response 10 11/03 10.13 stone pfw unforunately, i for once must agree with berger and parrello.. (yick) what am i saying? h e l p . . . ---------- response 11 11/03 11.14 hinton ssu Today, at least, the tabs are set 8 spaces apart, which seems to be a nice series of tabsets. But if the tab settings are going to change from what they were, can the System reset old tabs, the way it changes old commands ? Or would this REALLY mess up things ? Could tabs be reset in old lessons on request ? And while we're at it, I would like to vote AGAIN for tab-setting ability in the editor.... ---------- response 12 11/03 12.08 halvorsen forestry I doubt if the system could reset the old tabs to the new ones since there is no way of telling where a tab is, given a source line. A tab simply adds spaces until it reaches that column. Therefore, the system couldn't tell whether the line had spaces or tabs entered. Another vote to have tab setting ability. ---------- response 13 11/03 12.53 fumento pso Another possibility would be to put in a TO option (in addition to the current TT, TC, and TF) to specify old Tutor tabs. ---------- response 14 11/03 19.12 koning csstaff And document the T command, of course... ---------- response 15 11/04 02.04 sellers arizona Sorry, I stand corrected. Indenting ---------- response 16 11/04 11.38 parrello uimatha Ok. I can see that if we are really going to have indenting that enforcing these 8-space tabs all the way across is necessary, and perhaps this even explains the bizarre behaviour of MICRO i. I must admit, however, that when I originally wrote the note, all I knew was that I was having a terrible time keeping my comments lined up with the old ones. ---------- note 480 downer 11/03 08.37 parrello mfl About the most recent down tapes: Over here at FLB, they are almost readable on stations 13-0 and 13-1,. but elsewhere all they do is flash the red lights and exercise the slide projectors. Although some of my colleagues at this place enjoy the light display and the symphonious clinks, it does tend to get crowded around the two terminals that work. Does anyone else have this problem? ---------- response 1 11/03 09.00 frankel p Mike Williams says he's going to work on that...if I can just get him in here before the system comes up. ---------- response 2 11/03 17.49 b parrello med I think the system's already up, Dave.. ---------- note 482 response 11/03 09.26 friedman csa Could I please have a system response to the suggestion in note ¬$288? Thank you. ---------- response 1 11/04 15.25 b sherwood phys Presumably it should be a reserved word like "usersin"? What name is reasonable? (It seems possible and useful.) ---------- response 2 11/04 21.22 koning csstaff commonin ---------- response 3 11/05 08.51 friedman csa userscom?↓ I won't quibble about the mnemonics of the name, if only the feature would be created!↓ Thanks for the response, Bruce. ---------- note 483 stolen 11/03 10.07 friedman csa Anyone out there seen a stray keyset, serial number 859? (Please turn over your keyset--what numbers is written there in big black numerals?) We have been ripped off. I put this note here on the off chance that it was removed as a prank. We now have a perfectly good terminal with no keyset!↓ By the way, we have resorted to bolting our keysets down to the table to prevent further thefts. It restricts movement and convenience somewhat, but at least the thing's there when you need it! ---------- response 1 11/03 11.06 michael english George, I know how angry that can make you; I had one swiped from the Ed Bldg 2 years ago. But you might be interested to know that one of the things that makes PLATO usable to the physically handicapped is that the keyset can be moved. Personally, I doubt I could handle a bolted-down PLATO keyset without a lot more difficulty. ---------- response 2 11/03 11.58 sellers arizona steel cable? ---------- response 3 11/03 12.08 berger mfl Yes, they stole the cable with the keyset. Seriously, two thefts in two years ins't bad, considering security as it is, but if this continues, the convienience of the removable keyset may soon disappear. ---------- response 4 11/03 17.08 friedman csa I understand it's more like 10 thefts so far. We simply can't afford to replace the things.↓ Gary, thanks for the reminder. Sorry our room won't be suitable for you, though I know you seldom get over here.↓ Seriously, it's a real problem! But it seems to be a fact of life that the 99; must suffer because of 1; (or less) inconsiderate clods. ---------- response 5 11/04 08.03 mont csa With the budget situation so bad, departments cannot afford the $500 for new keysets. We'd have to give up pens and pencils for a long time. ---------- response 6 11/04 13.11 frye mfl I noticed the article in the DI, and was surprised to find that those keysets were worth that much! How much cheaper are the newest (and supposedly cheapest) keysets? A question -- does anybody have a theory on WHY the keyset was stolen? ---------- response 7 11/04 19.49 alan reading A homemade PLATO system? ---------- response 8 11/04 21.24 koning csstaff Probably ripped off by some ignorant who didn't realize these keyboards don't put out ASCII... That also explains why they are expensive: small quantity production. Besides, they are hi-quality keyboards. ---------- response 9 11/05 08.49 friedman csa I expect the thing is either dismantled for parts, or hanging on someone's wall as a souvenir! ---------- note 486 forms 11/03 11.14 davis uicem Since the inception of -notes- a minor, but nagging, problem has existed: what is the past tense and adjectival form of a TUTOR command? The only convention seems to be that of pointing out the problem. [E.g. In the "do-ed" (sp?!) (yech!!) unit, I want to...) May I propose a solution. 1. For commands not ending is _e_, add "ˆ.ed" (ˆ. = sup,period) Thus, -calcˆ.ed-; -ansuˆ.ed- ; -copyˆ.ed- . 2. For commands ending in _e_, add "ˆ.d" . Thus -defineˆ.ed- ; -useˆ.d- . 3. For stylistic reasons, let the following exceptions be used. -doˆ.ne- -put- (not -putˆ.ed-) -jumpˆ.edˆ.out- -doˆ.neˆ.to- -set- (not -setˆ.ed-) -showˆ.n- -goˆ.neˆ.to- -specs-(not -specsˆ.ed-) -storenˆ.d- -writ(e)ˆ.ten- (But perhaps -writˆ.ten- and -f(i)ˆ.ouˆ.nd- -fˆ.ouˆ.nd- would be better.) Dan Davis ---------- response 1 11/03 11.34 michael english UGH↑ˆ (2↑ˆ59-1↑ˇ)↑ˆ ---------- response 2 11/03 11.46 frye mfl Four thousand, six hundred and thirty-eight votes... against! Dave ---------- response 3 11/03 12.46 warner iu One more vote against.... ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬ ¬¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬ ¬¬ ¬ ¬↑ˆ¬↑ˆ¬↑ˆ¬↑ˆ¬↑ˆ that's a bad -doˆ.ed- ↓ (aaarrrrgggghhhhh on your own time, please.)↓↓ Silas ---------- response 4 11/03 13.12 s zweig iumed ahhh.......grammar!! ---------- response 5 11/03 15.46 davis uicem I guess the moral is, if you have a lousy idea don't suggest it on a Monday. Dan Davis ---------- response 6 11/03 18.50 dave fuller uimc I can't believe this... ---------- response 7 11/04 07.51 john r reading Aw, the idea isn't so bad, really. It's just that most of us have trouble enough expressing an idea in ANY form. let alone following a set of arbitrary conventions. The people who would remember the conventions are those who are sufficiently articulate that their own invented conventions would be equally understandable. And that's _usage_, not _grammar_. ---------- response 8 11/04 08.33 judy pso I think it's a very creative idea. Particularly, -doˆ.ne-, -calcˆ.ed-, and -doˆ.ne-. Superscript, period is awkward to type. How about -do'ne- or -do¬,ne- ? -fˆ.ouˆ.nd- is pretty yecchy, I agree, but don't knock down a complete idea just because some element isn't suitable! ---------- response 9 11/04 11.42 buz phar by the way, wouldn't it be -defineˆ.d- rather than -defineˆed.- if your going to propose a convention, stick to it. ---------- response 10 11/05 12.45 iezek ames YUCK!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! M. Iezek ---------- note 487 terminets 11/03 11.14 flood rhrc Terminet users: Please note that the 300 msec delay following linefeed characters can be reduced to 67 msec while performing multiple linefeeds (without sending any printing characters). The delay following a linefeed which is preceded by printing characters should still be at least 300 msec, however. General Electric recommends 350 msec for the first linefeed of a group if it follows printing char- acters, but 300 msec appears to suffice without loss of data. As an example: line 1[300 msec] line 2[300 msec] [67 msec] [67 msec] . . . [67 msec] is the recommended sequence. This information should prove useful in speeding up the "software" formfeed and vertical tab procedures. Please feel free to contact me if you need additional information. ---------- note 489 Seminar 11/03 11.19 whansen csa Seminar R. G. Montanelli Evaluations of Two Courses Taught with PLATO aid. 1:00 PM CST 115 DCL Tuesday, Nov 4, 1975 (Urbana campus) PLATO has been used for part of CS103 (50 students) and CS105 (600 students). Experiments were carried out to evaluate student performance. This talk will report the preliminary results of these experiments. ---------- response 1 11/03 12.28 walter physio For those of us who will be unable to attend, will you enter a short summary of results in NOTES? ---------- response 2 11/03 17.26 whansen csa Montanelli has written a paper and can give you a copy ---------- response 3 11/04 08.11 mont csa Actually, I have 1 paper which discusses the early results and a second one in progress which will include all data discussed in this talk. For copies of the earlier report, write or call Mrs. Erna Wright, DCS mailroom, 222DCL. For the later one, drop me a pnote in a month or so. Rich Montanelli ---------- note 496 job 11/03 12.30 rader s We are looking for a student hourly to work as a PLATO IV computer operator. The hours are 15-20 hours per week, evening and weekend shifts. We would like the person to start as soon as possible. Send p-notes to me (rader/s), and arrange for an interview. Decisions will be made by Wednesday, Nov. 5th. ---------- note 500 SD error 11/03 13.21 silver ve In Show Display mode, the following code: size 2 at 505 write Faults size 0 gives: FAults in size 2. When the write tag is "fAults" the display is done correctly. It is repeatable. John ---------- response 1 11/03 13.31 weible german I've encountered the same error recently. ---------- response 2 11/03 14.43 roper siu Also try copying out block funeral from lesson tapedisp... The SD option refuses to show the text on the arch properly, losing spaces and such. When condense it shows up fine. Quetzal ---------- response 3 11/03 17.48 iezek ames I have also encountered such an error in SD. Mark Iezek ---------- response 4 11/03 17.50 b parrello med This is known. and was mentioned in "notes" a short while ago, I think. ---------- response 5 11/05 12.43 iezek ames Has anything been done on it though? Mark Iezek ---------- note 502 DOWN 11/03 13.53 rader s We are expecting to install some additional computer equipment in late December-early January, and it is necessary for PLATO to be down for a long enough period for workmen to install larger motor-generator sets, get more chilled water capacity connected up, and remove a wall. To accomplish this, we expect PLATO to be down from 6 am December 22nd until 8 am January 10th. The additional equipment includes a second computer to permit development of multiple-executor systems, and additional memory to permit us to run a standard CDC KRONOS (NOS) system. The arrival of the equipment is still conditional; the site preparation is definite, however. ---------- response 1 11/03 14.01 wood medsiuc AHHHHHHHHHHHHH!!!!!!!!!! ---------- response 2 11/03 14.04 david hebrew Thank you very much for the detailed information WELL IN ADVANCE of the proposed dates. This is the kind of thing that restores my faith in humanity. ---------- response 3 11/03 16.22 mcneil a uicc Unfortunately, several PLATO users in the Chicago area will want to start winter sessions on Jan. 5th. ---------- response 4 11/03 17.03 michael cornell Also unfortunately, it cancels my plans for a NEAT project which I can't _possibly_ get done except if I work on it all intersession. Oh, well... I hope the new equipment will not be immediately overloaded by adding a whole flock more terminals! Mike Oltz ---------- response 5 11/03 17.23 eades propman We must welcome the progress; however, the long down time will cause considerable problem. Anything that can be done to lessen or interrupt the down time would be most appreciated. ---------- response 6 11/03 18.35 meers wright thats almost 20 days... Dave better have lots of copies of those tapes of his handy when they wear out... ---------- response 7 11/03 19.34 mike b cornell Hmmm.... Is there any documentation available on the system about the KRONOS (NOS) system? M¬ ike≤≤≤≤ˇB ---------- response 8 11/03 22.31 rader s It will not be likely that we will come back up for a short while; we will of course hope to come back earlier than the final date. It is not possible to make any promises at this time. We will keep you informed about our plans up to the time we go down (and after?). No, we have no documentation of Kronos (Nos) on the system--we do not expect this change in the underlying structure to be visible or of any practical importance to you. ---------- response 9 11/04 08.17 mont csa I sure hope this doesn't bring a repeat of the absolute chaos we had during the first 2 weeks, last spring. At that time things were so bad that we actually considered dumping PLATO for 600 students and reorganizing the entire course on on the fly. Isn't there any change to plan these things for summer or sometime when a 2 week late delivery won't give month long headaches to hundreds of people and ruin experiments planned to assess the supposed 'advantages of using PLATO'! ---------- response 10 11/05 09.56 siegel peer Is it the intention of the system staff that PLATO will be up by 8 am, January 10th, without regard to the com- pletion of the job? That is, can we plan for Jan. 10th or may this date be extended as well? ---------- response 11 11/05 14.06 railing iu I remember how it was last winter when the new ECS was put in, so anything that can be done to reasure PLATO users about exactly how long it will be down and what date it will come up will be very welcome. Maybe you could have a tape that says "PLATO will be down (x) days" or have the completion date on the tapes, or both. We miss the system when it goes down! Malkin ---------- note 504 embed,yuck 11/03 14.01 wood medsiuc I tried the following code: calcs n1,n30‎'Q','Q≤ˆˆ-','Q≤ ' . mode rewrite . write ¬0m,w¬1 ¬0a,n30¬1. . Unfortunately, the embeded mode write command did not work...it erased the top half of "Q" when it wrote the "-" above the letter. I changed the code to something a little less glamorous (namely using a -mode write- statement followed by a -mode rewrite-) and obtained the desired results of leaving the terminal in rewrite mode except for that one character. Why did the embed feature not work in this case? ---------- response 1 11/03 14.04 little t research As I understand it, the embeded mode feature works only till the next embeded command in the same write statement. Todd ---------- response 2 11/03 14.17 michael english I know it's too late to change this, but would someone please tell me the rationale for having the embedded mode operate in this fashion? What flexibility does this way offer over the "common sense" approach? Or is it simply technically infeasible? ---------- response 3 11/05 14.19 mcneil a uicc The present scheme seems plain old dumb to me. ---------- note 509 editcom? 11/03 14.40 elston rhrc I encountered a strange quirk in the common editor this morning. In the OCTAL entry mode if you erase the 'o' and entry -0 you seem to get the first ten chars of the copy buffer or some such thing stuffed in the word instead of 077777 77777 77777 77777 as one would expect. Entering -o0 gives the correct result. Is this the way it is suppose to work. 2c≤/ ---------- response 1 11/03 16.23 white p Thanks, fixed now. Did you know that "-0" is matched by an -answer (blank)-, and that "-0+-0" is too, but "0" isn't and "-0-+0" isn't? ---------- response 2 11/03 21.57 b sherwood s Sigh. ---------- note 511 copy-block 11/03 15.11 marty smith mtc When using the "COPY-a-block" option in the editor, is the ANY possiblity of letting the author know which blocks are not set to condense? If not, why? your-inquisitive-SCHMUCK! ---------- response 1 11/03 21.58 b sherwood s Plotting speed, maybe? Or that nobody thought of it? ---------- response 2 11/03 22.32 rader s Nobody thought/bothered. I think it would be a good idea... ---------- response 3 11/04 16.45 marty smith mtc gee, thanks. will it ever be implemented? (seems to me plotting speed decrease would hardly be noticeable) your-would-love-to-see-it-implemented-SCHMUCK! ---------- note 515 endcontest 11/03 16.46 fay o Just a reminder: the PLATO down tapes display contest ends this Wed., Nov 5, at midnight (2400 or 0000 hrs, which- ever). Just 55 ¬+ 1/4 hrs of entry time left! Hurry--hurry--hurry... ---------- note 516 student id 11/03 16.47 michael english We have a lesson which reads the rosters of approximately 45 courses and automatically allocates student data storage space in datasets. The case has arisen twice already where a students signon name was changed through course records by the instructor of the course. The allocation program, noting the absence of an old name and the presence of a new one takes the only logical action -- it deletes the old record and creates a new one, destroying all the data collected under the old. It is infeasible to require instructors to alter the name in student records AND in our system. The simplest solu- tion, for us, would be for the system to assign a unique number to each record in a course, this number staying con- stant over name changes. This number could form the basis of any allocation routine, and would undoubtably have many other uses. ---------- response 1 11/03 16.57 marty smith mtc Which gets back to the question of when course directors will be able to -writecs- records into their own courses without going into the -records- routine. (that is, when are we going to be able to use our won routines to edit courses!?) ---------- response 2 11/03 17.07 michael cornell When you know as much about course records as a sysprog does.... ---------- response 3 11/03 17.11 marty smith mtc which is? ---------- response 4 11/03 18.42 dave fuller uimc writecs? ---------- response 5 11/03 19.10 meers wright Everyone does have a unique number, and it used to be shown next to your name on a lesson's 'last edited by' page.. The only way to see the number now is by looking at a course info page. If you were the last editor it will show your number next to your course... I did a lot of testing, and if you are number 38 in a course right now, and all names in your course get deleted except you, you will still be number 38 even though you are the only person left in the course... The number does not change if people are added, deleted, or names are changed. I don't know what happens when the course gets extended or shortened. If you delete number 3, and then create another signon later, it will take the first available number - in this case 3. Now that you have a unique number, you gotta figure out how to read it........al ---------- response 6 11/03 20.22 k mast p Not so! That 'unique' number will change without notice -- it is only used as a pointer to where in the course the record is. I symphathize with you, Gary. I encounter the same problem in Personal Notes and in Student Comments, and while it IS possible to solve for my cases, it is not easy and would be impossible without special system commands. I have proposed a unique number scheme which would assign a unique number to every individual and every course. Two commands would then be implemented such that given a number. it would return the name and course (or -1 if no such person) and given name it would return number (like- wise -1 if no such person). Such a system is still being discussed, and I have no idea when if ever it will be implemented (at even a system level), at it is very expensive disk access wise. (continued....) ---------- response 7 11/03 20.27 k mast p Releasing it in general poses another problem -- how to we keep people from finding out names and courses merely by writing routines to go through and try all numbers? The other major point against such a scheme is that in storing numbers instead of names, if anything anywhere (whether it be at at a systems level or users level) gets messed up. even by so much as a few bits, then it becomes virtually impossible to try and put it back together and figure out what happened. Perhaps someone can come up with a better solution (or modifications to this one to counter the points against it.) If so, I would be very happy to listen. ---------- response 8 11/03 20.41 michael english The proposed "give name, return number" command would solve my problem, as I could store both name and number. Any additions would be first checked to see if pending deletions have the same ID number, thus signalling a name change. I don't see how this particular command, if implemented as a -readr- option, could hurt anything. I should think that the disk accessing would not be exorbitant in this case. ---------- response 9 11/03 22.01 b sherwood s Work is under way to let you specify what options your instructors can perform. In which case you could rule out name changes..... ---------- response 10 11/03 22.38 koning csstaff That would not be a real solution... The record number would be, and for this application you don't need the (costly) ability to search for a record by record number. All you want is that a -readr- returns the record number as well as the name so that the 'allocation program' can search its records for the record number rather that for the record name. ¬↑ˇP¬≤↑ˇK ---------- response 11 11/04 08.26 mont csa Ruling out name changes is not really feasible, as typos should be corrected. Students ( and everyone else for that matter ) do not take kindly to having to spend the whole semester mistyping their names on PLATO. ---------- response 12 11/04 09.33 hinton ssu In a lesson under development which will be required of most of our undergraduates, we store the student's own school ID ¬$ in the 1st 27 bits of a var, and sent it to the datafile along with his/her last neme: I don't think it unduly clut- ters up the file.... ndh ---------- response 13 11/04 11.54 friedman csa The number Meers mentions is response 5 WILL change if the course file is either lengthened or shortened. ---------- note 533 packc 11/03 20.10 david hebrew Consider the following syntax for the much-discussed packc: ::= ::= ::= ::= ::= > ::= ¬I ::= ¬I packc (The first symbol chosen from is the one that is used in every subsequent reference to ) Examples: packc n1=5,n100,n2,True,False packc n1;n100;n2;negative;zero;one;two;many,many,more! packc n1;n100;n2;negative;zero;one;nomore; packc n1¬,n100¬,n2¬,Comma,negative¬,Semicolon;zero¬,ETC. ---------- response 1 11/04 13.10 b sherwood s Thanks for the suggestion. I think one of the reasons we never did put in a conditional form may have been not only the lack of a specific suggestion for the syntax, but also because we were probably thinking of using the name -pack-, rather than -packc-, which would vastly simplify the condensing. ---------- note 537 camera 11/03 21.12 blue cerl hello...anyone. I'm looking for a Leica M-3 or M-4 range finder camera. If you know of one available at a reasonable price, leave a note for blue , cerl. Thanks ---------- response 1 11/03 21.46 michael english Try lesson "wantad", not here..... ---------- response 2 11/04 16.04 broadus css there is a cannon in logotest ---------- note 545 newMopt? 11/03 23.18 s zweig iumed How about an option in the editor which would permit one to Modify all occurences of a given string with just one keypress? The present M option is great, but perhaps a shift-SOMETHING to change _all_ instances of a string would certainly be handy at times. Whatcha think, y'all? ---------- response 1 11/03 23.36 blomme s The option was finally put in only because it was limited in scope--and the valid point made at the time was that one nearly always wants to see the lines selected for change (and after change)--so it is very unlikely that such a feature would be provided in the foreseeable future. ---------- response 2 11/03 23.44 s zweig iumed I can also see valid reasons for wanting to do the whole schmeer at one shot. What I had in mind was not eliminating the present single replacement feature, but rather adding another keypress which would provide replacement of ALL occurences of the string. Perhaps this clarification will help. ---------- response 3 11/03 23.56 lieber m When replacing with the "m" option I often find things that I didn't want replaced especially when part of a longer string matches the "old string" It is possible to catch these if doing them one by one. And anything that works across block boundries makes you lose that GREAT option... "OUT"! ---------- response 4 11/04 07.58 avner s Speaking from long, hard-won experience (scars displayed upon request). it is almost always the case that the unintended changes that occur as a byproduct of "blind" alterations take more time to locate and correct than is saved by not having to do a quick verification on each. This is particularly true in TUTOR code where a relatively unconstrained format prevents the cross-checks possible in languages with more rigid formats (and no text!). Al Avner ---------- response 5 11/04 17.10 koning csstaff Well, if you are afraid of getting more scars, just don't use a feature like the one proposed. But I would like to have a replace-all for those (admittedly not too numerous) cases where I know for sure that I want to replace all 60 occurrences of a particular string. I would suggest that such a replace-all feature would be limited to one block at a time, to make sure that OUT still works. ---------- response 6 11/04 19.14 john r reading People who request such a feature will find that, in purgatory, they are consigned to the role of computer operator - with the sole function of retrieving undamaged lessons from backup. ---------- response 7 11/04 19.17 little t research Aww come on fellows. Lets get more serious. Most editors on other systems have such an option that works on the present page only. Now I agree because most other systems have this isn't a very good reason for having it, I still would be much in favor of having it. As to people who really mess things up and then press BACK, those type of people could also do D9999 HELP1 and save themselves the trouble of realizing what they just did. Todd ---------- response 8 11/04 22.30 blomme s Yes, let's get serious...and get some of the few hundred more important things done! ---------- response 9 11/05 05.19 ian cs425 I've seen this option in other editors. Invariably, the reason given for not (at least) typing out everything that has been changed turns out to be 'But a teletype is so slooooooowwwwwww' Somehow, that doesn't sould like a very good reason for putting it on PLATO. ---------- response 10 11/05 17.15 koning csstaff Well, another reason would be that I don't see why I should look at all occurrences of a string separately if I know for sure that I want to change _all_... *** end of notes ***