riks@athena.UUCP (Rik Smoody) (11/05/84)
We are in (dubiously) good company in teaching UNIX to undergrads: I refer to the teaching of law. Folks who study law here in Oregon have a much easier time with Oregon's bar exam than those who study out of state. It's not that general principles are different, but that some FACTS are important too. A master carpenter/contractor knows much more than the physics of wood and iron. Go to your local steel mill and ask for pieces of iron 3.54 inches long and 7/64 inch diameter. They will work for joining pieces of wood which are each 1.77 inches thick (which you carefully calculated as the thickness needed to support your load) I sure would not hire you to build for me. I want a carpenter who buys 2x4's and 16d sinkers (vinyl coated nails). Economics is an important part of most jobs. Compare also the study of medicine: the system is set up such that people are encouraged (forced, usually) to get their internship training in a different place from their med school. But, luckily, bodies work and fail similarly all over the country. And a large part of medical training is pragmatics. There is plenty of room for specialists, but I'd still think twice about a physician who's never delivered a baby nor sutured a cut. And right now, with the computer industry as it is, I'd "wonder" about an undergraduate program which did not include any study of UNIX or closely related system. I'd discount the program until they convinced me that they really did have a good replacement. Rik Smoody
jona@clyde.UUCP (Jon Allingham) (11/08/84)
> ... > There is plenty of room for specialists, but I'd still think twice about > a physician who's never delivered a baby nor sutured a cut. And right > now, with the computer industry as it is, I'd "wonder" about an > undergraduate program which did not include any study of UNIX or closely > related system. I'd discount the program until they convinced me that > they really did have a good replacement. I think this is a totally unreasonable expectation. As has been mentioned in previous articles, universities are in the business of teaching concepts, problem solving, and ( for CS programs ) structured programming which should enable any graduate to learn how to use most any other language/OS/computer when it is needed as opposed to trying to learn every one of the popular, fad operating systems and languages that comes along. Also, due to the lack of funding, most schools cannot run out and buy a new computer system at whim. Most computers that support Un*x are for the most part small and aren't able to support the load that the average university wants to put on them. Try running some extremely large number crunching problems on a Un*x machine sometime. I'll take a large CDC, Cray, or IBM any day. See what happens when more than 25 people want to use EMACS at once. Another problem is faculty "hogging" of resources. The popularity of Un*x is such that faculty members want to do all their pet research projects on the VAX ( or whatever ) and severely restrict access by students to the machine. When you start talking pragmatics, you should look at both sides of the problem. -- Jon M. Allingham (201) 386-3466 AT&T Bell Laboratories Rm 2A-110 Whippany, NJ
herbie@watdcsu.UUCP (Herb Chong, Computing Services) (11/08/84)
Knowing about Un*x and learning Un*x are two different things. I strongly believe that Un*x is, and will remain, an influential operating system in its design concepts. Any program which does not do even a casual study of it in an operating systems course is suspect. However, sitting down and actually using Un*x is not required, or even desired, in some cases. Also, the operating systems course should cover OS/360 and its successors, if only to learn how things should not be done. Sitting down at a terminal and being able to write and run a program in C is a poor criteria for the worth of a computer science program. Herb Chong... I'm user-friendly -- I don't byte, I nybble.... UUCP: {decvax|utzoo|ihnp4|allegra|clyde}!watmath!watdcsu!herbie CSNET: herbie%watdcsu@waterloo.csnet ARPA: herbie%watdcsu%waterloo.csnet@csnet-relay.arpa NETNORTH, BITNET: herbie@watdcs, herbie@watdcsu
dmmartindale@watcgl.UUCP (Dave Martindale) (11/08/84)
> I think this is a totally unreasonable expectation. As has been mentioned > in previous articles, universities are in the business of teaching > concepts, problem solving, and ( for CS programs ) structured programming > which should enable any graduate to learn how to use most any other > language/OS/computer when it is needed as opposed to trying to learn > every one of the popular, fad operating systems and languages that comes > along. Are you suggesting that UNIX is a "fad" operating system, while VMS, VM/CMS and friends are "real" operating systems? > Also, due to the lack of funding, most schools cannot run out and buy > a new computer system at whim. Most computers that support Un*x are > for the most part small and aren't able to support the load that the > average university wants to put on them. Try running some extremely > large number crunching problems on a Un*x machine sometime. I'll take > a large CDC, Cray, or IBM any day. See what happens when more than > 25 people want to use EMACS at once. I read this as saying that a "large CDC, Cray, or IBM" machine can support a large number of users because it prevents anyone from using one of the more powerful tools of modern computer science. Quite true. I consider that an argument for, not against, UNIX. Also, consider just how many VAXes you can buy for the price of one large CDC machine, or a Cray. And there are UNIX machines that give better performance/price than a VAX. > Another problem is faculty "hogging" of resources. The popularity of Un*x > is such that faculty members want to do all their pet research projects > on the VAX ( or whatever ) and severely restrict access by students > to the machine. Doesn't this indicate that either you need more VAXes, or someone needs to allocate the existing VAX time more fairly? Again, you seem to be arguing that a non-UNIX system is better than UNIX because nobody really wants to use it, so there won't be much competition for its resources! > When you start talking pragmatics, you should look at both sides of > the problem. I try to do that. At Waterloo, most of the undergraduate CS teaching is NOT done on UNIX. But the reasons are a bit different than those you postulate above. Certain 4th year and graduate courses are taught on UNIX because they need some of the facilities that UNIX offers. These pretty well saturate the available capacity on the teaching VAXes. There are a number of research VAXes that cannot be used for teaching because they were bought with research grant funds or donated by the manufacturer on the condition that they be used for research. Meanwhile, we have a Honeywell mainframe that has about the raw CPU power of a 780 but much higher I/O bandwidth and a sufficiently dumb front end that you cannot run a screen editor. It will support something like twice as many users as a 780. It gets used heavily for teaching, not because it is necessarily better than the VAXes, but because it is there and it works. Much teaching is also done on a group of IBM 4300-series machines that are run by the university's computing centre. The computing centre has a long history of commitment to IBM, for their own reasons. The hardware is there, it runs, the computing centre has a large staff to provide support for users. This mass of hardware is capable of handling very large numbers of users, and is the most suitable machine for most large long-running number-crunching jobs, so it gets used for this. Note that the choices here are primarily ones of which hardware can best handle which courses, and only secondarily which operating system people would like to use. There are certainly a number of people who do avoid UNIX because of the difficulty of understanding the system when all you have to go on is the standard manual set. The computing centre has user consultants and teaches quite a variety of courses periodically, making it much easier for an occasional computer user to learn how to use the IBM systems. The organization that runs the Honeywell and VAXes simply doesn't have the resources for this. The computing centre now has a VAX running UNIX, though, so this situation may change. (Now if only they'd run UTS on the IBM processors...)
dmmartindale@watcgl.UUCP (Dave Martindale) (11/08/84)
> Knowing about Un*x and learning Un*x are two different things. I strongly > believe that Un*x is, and will remain, an influential operating system > in its design concepts. Any program which does not do even a casual study > of it in an operating systems course is suspect. However, sitting down > and actually using Un*x is not required, or even desired, in some cases. > Also, the operating systems course should cover OS/360 and its successors, > if only to learn how things should not be done. Sitting down at a terminal > and being able to write and run a program in C is a poor criteria for the > worth of a computer science program. > > Herb Chong... > > I'm user-friendly -- I don't byte, I nybble.... > > UUCP: {decvax|utzoo|ihnp4|allegra|clyde}!watmath!watdcsu!herbie > CSNET: herbie%watdcsu@waterloo.csnet > ARPA: herbie%watdcsu%waterloo.csnet@csnet-relay.arpa > NETNORTH, BITNET: herbie@watdcs, herbie@watdcsu *** REPLACE THIS LINE WITH YOUR MESSAGE *** It is one thing to study an operating system for a few lectures in your operating systems course, and quite another to use it for long enough to develop a feel for what you like and don't like about it. And, I believe that the instinctive "feel" that a student develops for what methods are "good" and what ones are "bad" is likely to be reflected in the software that he/she writes once employed. If you do all of your programming in Fortran, you are likely to think about handling problems in a considerably different way than if you have been exposed to, and USED, the different approaches taken by C or APL or Lisp or Snobol or Prolog or even Pascal. I believe that the same applies to systems. If you have actually USED UNIX and VM/CMS and RSX and a Macintosh, you will have a much better idea of how systems can be designed (and mis-designed!) and what choices are best for the system (or perhaps just user interface) you are building today. So, I don't believe that exposure to UNIX or C, per se, are necessary. But they are good choices as examples of the present technology of computer science. So the question to ask is not "why aren't students exposed to UNIX?", but "if students aren't exposed to UNIX, what better system are they exposed to instead?".
eugene@ames.UUCP (Eugene Miya) (11/10/84)
> We are in (dubiously) good company in teaching UNIX to undergrads: . . . > now, with the computer industry as it is, I'd "wonder" about an > undergraduate program which did not include any study of UNIX or closely > related system. I'd discount the program until they convinced me that > they really did have a good replacement. > > Rik Smoody Hi Rik, long time no hear, no see. Friends from Caltech and SB say hi. I wonder this, too. As many know, California has a three tiered high education system: community colleges, state universities (name change care of Ron Raygun, without any addition of funds), and the UC system. I have been consulting recently with members of the Cal State system. None of the local CSUs are running UNIX. They can't afford to, nor do the local community colleges in Silicon Valley. I think those of us in know about UNIX are shielded. At the CSUs, the CSUs can't pay the money for a full-time UNIX person, so they have to use RSTS-E with DEC support. Two local CCs have donated HP and Tandem machines and they don't have the support either. Both have C programming language classes, but can you imagine learning C without a compiler at your school? They are (maybe if you have a home PC as many Si Vallees do, you can play with C. Similarly with a local Ada class. Most of these people attending the lower tiers are more concerned about BASIC, COBOL, some Ada, MS/DOS [or PC/DOS] (read: get a job) than about UNIX. Many of the others are running MVT/MVS, EXEC*1100 and so on. Ideally, [certainly?], the principles of operating systems should be independent and transferable between different OSes, but when you have to grovel in the assembly language, forget! I suspect some of these universities are in bed with IBM and Univac or their profs were involved with development, or they are unwilling to learn UNIX. --eugene miya NASA Ames Res. Ctr. {hplabs,ihnp4,dual,hao,vortex}!ames!aurora!eugene emiya@ames-vmsb.ARPA
chris@umcp-cs.UUCP (Chris Torek) (11/11/84)
There are really two issues here. First of all (as someone else has already pointed out) it is far cheaper to support hordes of students on one Univac 1100 running EXEC-8 than it is to support said hordes on VAXen or equivalents running Unix or some other "modern" operating system (VMS is considerably better than EXEC-8, for example). Many universities just can't afford the CPU crunchons it takes to teach Unix to undergraduates. The second is that if someone really knows what s/he (it?) is doing, knowing Unix well (from the outside; I'm not talking heavy kernel here) is not particularly important. With the right concepts, everything else just falls into place. To take a reverse example, I recently had to write a short SNOBOL program on our 1182, running EXEC-8. Even working on an unfamiliar machine (I avoid the Univac as much as possible) and in an unfamiliar language (I've never so much as *looked* at SNOBOL before), it didn't take me long to write it -- because I understood, in general, what was going on. Sure, I needed lots of reference manuals nearby, but if I were to use the machine or the language more often, I could soon hack at a reasonable rate. (I'm ashamed to admit that it took probably 3 hours to do such a tiny program: I spent at least an hour and a half before discovering that you can't use NE(foo) on user-defined types; gotta use \IDENT(foo). Sigh.) If someone understands the basic concepts in operating systems, then knowing Unix isn't so important. Pipes should be nothing more than a pleasant surprise (``gee, you mean I don't have to store 30 megabyte temporary files?''), the tool philosophy should be familiar, learning the editor shouldn't be too difficult (hey, if I can learn to use EDT in an hour, anyone should be able to learn Emacs in 20 minutes! :-)), and so forth. [Note, however, that being able to *use* something and being a wizard at it are two entirely different things.] If you need a Unix wizard, *then* Unix experience is a must. But if all you need is someone to write programs, then find someone who knows language design concepts and can write algorithms! Experience is not the most important thing in this case. -- (This line accidently left nonblank.) In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (301) 454-7690 UUCP: {seismo,allegra,brl-bmd}!umcp-cs!chris CSNet: chris@umcp-cs ARPA: chris@maryland
ems@amdahl.UUCP (Edward Michael Smith) (11/12/84)
> > ... > > now, with the computer industry as it is, I'd "wonder" about an > > undergraduate program which did not include any study of UNIX ... > > Rik Smoody > > None of the local CSUs are > running UNIX. They can't afford to, nor do the local community colleges > in Silicon Valley. I think those of us in know about UNIX are shielded. > > Most of these people attending the lower tiers are more concerned about > BASIC, COBOL, some Ada, MS/DOS [or PC/DOS] (read: get a job) than about > UNIX. > > --eugene miya It is an unfortunate fact that most of the software in the world of business is written in COBOL and runs on IBM type machines. The schools are just providing the skills to work in the real world. UNIX (tm) has caught on lately, but is still a gnat in comparison with the IBM/COBOL market. (Though with a lever long enough and a place to program in C ...) While it is sad that the schools look more at the quantity of use of a language/OS than at the quality of them, would all those applications programmers be better served by C and UNIX or by COBOL/VM/MVS? I hope that the schools are training for the past market and that C will lead to the end of COBOL. If this is true, we should see a change taking place in the schools starting in a few years... -- E. Michael Smith ...!{hplabs,ihnp4,amd,nsc}!amdahl!ems The opinions expressed by me are not necessarily those of anyone. (How can a company have an opinion, anyway...)
eugene@ames.UUCP (Eugene Miya) (11/12/84)
> > I think this is a totally unreasonable expectation. As has been mentioned > > in previous articles, universities are in the business of teaching > > concepts, problem solving, and ( for CS programs ) structured programming > > which should enable any graduate to learn how to use most any other > > language/OS/computer when it is needed as opposed to trying to learn > > every one of the popular, fad operating systems and languages that comes > > along. > > Are you suggesting that UNIX is a "fad" operating system, while VMS, > VM/CMS and friends are "real" operating systems? The original article [I thought] started out on a good thread. Yes, ideally, universities are in that business and should be unbiased..... I don't think you will find many who will disagree, but maybe "real men" should start eating quiche just to rebuke that popular book [and real programmers should start using Pascal, just to rebuke that popular arcticle]. As for fads, I guess studying electricity was a `fad' 200 years ago, too. > > > average university wants to put on them. Try running some extremely > > large number crunching problems on a Un*x machine sometime. I'll take > > a large CDC, Cray, or IBM any day. See what happens when more than > > 25 people want to use EMACS at once. > > I read this as saying that a "large CDC, Cray, or IBM" machine can support > a large number of users because it prevents anyone from using one of the > more powerful tools of modern computer science. Quite true. I consider > that an argument for, not against, UNIX. Also, consider just how many > VAXes you can buy for the price of one large CDC machine, or a Cray. > And there are UNIX machines that give better performance/price than a VAX. Working with people at LLNL and ANL, I must conclude that Crays (for this type of work are more cost effective) at this point in time. Obviously this is an oversimplification: if a Cray goes down everybody is out [They are quite reliable], but some number of VAX might be down and some work is being done somewhere. It depends on the work deing done. Avoid text editing small files. However, it would sure be nice to have some good software tools on the Cray. What makes an Apple II more powerful than the old ENIAC is the software, not just the level of electronics inside. > > Another problem is faculty "hogging" of resources. This is a political problem. Politics creates interesting problems, but running some other system (say CMS) is not going to solve this problem. We have people who want to add a batch control facility to SV. They are in a position of power, but their computer exposure is with IBMs, CDCs and Crays: give a person a hammer and everything turns into a nail. > > When you start talking pragmatics, you should look at both sides of > > the problem. This sounds kind of "holier-than-thou." Looking at a problem `pragmatically' is not necessarily the job of a unversity [perhaps "looking at both (or more) sides of a problem" is]. The discussion between the gentlemen from AT&T and the U of Waterloo sounds like the ongoing discussion between the older, CENTRALIZED computer concept and the newer distributed workstation (read `time-share') argument. It reflects the very structures of the two systems (AT&T and a university) where each of the men are coming from [Fred Brooks are you listening?]. Perhaps computing is undergoing the kind of social revolution which countries in Europe did went changing from monarchy to democracy. The computing center will still exist providing `super' services but it needs to adapt. --eugene miya NASA Ames Res. Ctr. {hplabs,ihnp4,dual,hao,vortex}!ames!aurora!eugene emiya@ames-vmsb.ARPA
ems@amdahl.UUCP (Edward Michael Smith) (11/12/84)
> > > Also, due to the lack of funding, most schools cannot run out and buy > > a new computer system at whim. Most computers that support Un*x are > > for the most part small and aren't able to support the load that the > > average university wants to put on them. Try running some extremely > > large number crunching problems on a Un*x machine sometime. I'll take > > a large CDC, Cray, or IBM any day. See what happens when more than > > 25 people want to use EMACS at once. Don't forget that AMDAHL has a UNIX (tm) port named UTS that runs on those nice big IBM style machines!! (UNIX is a trade mark of Bell Labs...) > > I read this as saying that a "large CDC, Cray, or IBM" machine can support > a large number of users because it prevents anyone from using one of the > more powerful tools of modern computer science. Quite true. Not true! See above comment... > > the resources for this. The computing centre now has a VAX running > UNIX, though, so this situation may change. (Now if only they'd run > UTS on the IBM processors...) Yes, yes!!! (Please forgive the 'commercial' aspect of this posting. I do work for AMDAHL, but I am also a part time instructor at a J.C. This posting is made in the spirit of an instructor thinking of the new students who could be reached *without buying new hardware*) -- E. Michael Smith ...!{hplabs,ihnp4,amd,nsc}!amdahl!ems The opinions expressed by me are not necessarily those of anyone. (How can a company have an opinion, anyway...)
herbie@watdcsu.UUCP (Herb Chong, Computing Services) (11/14/84)
I might point out that UTS implementation of Unix is not a complete one. I got this message recently from hlj@amdahl and am quoting without permission. I was asking about using vi on UTS. I had thought it was not possible at the time. Try to imagine using most of Unix without the facilties mentioned (i.e. using dumb in tset). Running UTS as a replacement for Unix on IBM and IBM-compatible systems is not practical unless you are completely aware of the limitations and are willing to work with them. I might also mention that the University of Waterloo was offered UTS on a one year trial basis and declined, partly for this reason. For the moment, VAXen and other such beasts are the only place to run a reasonably complete Unix system. IBM's VM/IX is still more rumor than fact, and Cray's Unix-like system for the Cray-2 is in limbo along with the Cray-2. A university supporting the teaching workload the size of the University of Waterloo must either run a LOT of VAXen, or a fewer number of bigger machines (read IBM). In any case, the amount of exposure to Unix is primarily at the 4th year level because of the limited resources. You only have time to use and understand the basic concepts. Herb Chong... I'm user-friendly -- I don't byte, I nybble.... UUCP: {decvax|utzoo|ihnp4|allegra|clyde}!watmath!watdcsu!herbie CSNET: herbie%watdcsu@waterloo.csnet ARPA: herbie%watdcsu%waterloo.csnet@csnet-relay.arpa NETNORTH, BITNET: herbie@watdcs, herbie@watdcsu POST: Department of Computing Services University of Waterloo Waterloo, ON N2L 3G1 (519)886-4733 x3524 --------------------------------------------------------------- The UTS 2.x product does not support vi, as it supports only 3270's and half duplex ASCII terminals. A development version of UTS that we use in the lab supports full duplex terminals, and thus vi and uucp, through special software modifications in the 4705 communications control (frontend).
dmmartindale@watcgl.UUCP (Dave Martindale) (11/14/84)
I DO wish that the department of computing services had taken Amdahl up on its offer and put up UTS for a trial period. If it is unusable because of features that are missing, the users will soon figure that out for themselves. Personally, I consider CMS almost unuseable because of the way unsolicited terminal input is handled. I would have liked to compare CMS to UTS. With respect to vi, does VM/CMS have a good screen editor that runs on ordinary ASCII terminals? It didn't seem to last time I used it. By deciding in advance that UTS was unsuitable, DCS has prevented any possibility of users deciding that they like it. I have heard very negative comments about UTS from people who have used it so I'm willing to believe it may have many problems but, dammit, what's wrong with letting users decide for themselves?
herbie@watdcsu.UUCP (Herb Chong, Computing Services) (11/14/84)
Good screen editor depends on what you mean by good. XEDIT is available provided you use the SIM3278 or Yale 3270 emulator packages. The feature that I miss most in vi is the ease of writing editor macros to do special things. On the other hand vi and company have regular expressions. IBM's apporach is not as flexible but comes close. Herb Chong... I'm user-friendly -- I don't byte, I nybble.... UUCP: {decvax|utzoo|ihnp4|allegra|clyde}!watmath!watdcsu!herbie CSNET: herbie%watdcsu@waterloo.csnet ARPA: herbie%watdcsu%waterloo.csnet@csnet-relay.arpa NETNORTH, BITNET: herbie@watdcs, herbie@watdcsu POST: Department of Computing Services University of Waterloo Waterloo, ON N2L 3G1 (519)886-4733 x3524
honey@down.FUN (11/15/84)
/***** down:net.cse / watdcsu!herbie / 11:28 pm Nov 14, 1984*/ ... I got this message recently from hlj@amdahl and am quoting without permission. ... /* ---------- */ thanks for the warning. rest assured that's the last confidential tidbit you will ever receive. peter honeyman
herbie@watdcsu.UUCP (Herb Chong, Computing Services) (11/15/84)
The quote was only substantiating what has been an open secret that has been well discussed in Datamation at least for the past year. If I though it was in the slightest bit confidential, I wouldn't have published it, and I doubt I would have received it, especially since I got it because of a net posting. But this all doesn't change the fact that at the moment, UTS does not adequately support a full duplex system for terminal I/O, so very large unix systems are not going to magically appear on mainframes of university campuses for a while yet. Herb Chong... I'm user-friendly -- I don't byte, I nybble.... UUCP: {decvax|utzoo|ihnp4|allegra|clyde}!watmath!watdcsu!herbie CSNET: herbie%watdcsu@waterloo.csnet ARPA: herbie%watdcsu%waterloo.csnet@csnet-relay.arpa NETNORTH, BITNET: herbie@watdcs, herbie@watdcsu POST: Department of Computing Services University of Waterloo Waterloo, ON N2L 3G1 (519)886-4733 x3524
wcs@ho95b.UUCP (Bill Stewart) (11/16/84)
> I DO wish that the department of computing services had taken Amdahl up on > its offer and put up UTS for a trial period. If it is unusable because of > features that are missing, the users will soon figure that out for themselves. There is a decent, relatively complete Unix System V for 370's - it's from AT&T. My building has one in the basement, and it was real nice being on a 370 with 40 users instead of my department's 80-user VAX. (We've gotten a few more vaxen and an AT&T 3B-20, so our machines are no longer totally overloaded.) The machine supports about 250 or 300 users; it was fairly new when I was using it. I don't know anything about educational licensing, but I think it's commercially available. I haven't done any work on the 370 in a while, but as of a year ago, there wasn't a reliable Fortran (hence no "S" package), and it went down more than our VAX, but these problems were actively being fixed, and I think have gone away. The operating system is not native, but runs on top of VM. This makes it somewhat slower than UTS, but is still fast. The front-end processor was a Series-1; it liked odd parity, and couldn't handle "stty echoe", but the system does vi fine, (and I assume also emacs?). The system was paged, but the average user was limited (and defaulted) to 8 Meg processes. Bill Stewart, AT&T Bell Labs, Holmdel NJ {ihnp4, allegra, vax135}!ho95c!wcs -- Bill Stewart AT&T Bell Labs, Holmdel NJ 1-201-949-0705 ...!ihnp4!ho95b!wcs
gam@amdahl.UUCP (Gordon A. Moffett) (11/20/84)
Herb Chong writes: > I might point out that UTS implementation of Unix is not a complete > one. It would be nice if you made clear *which* UTS you are refering to. In the letter from amdahl!hlj you quoted: > The UTS 2.x product does not support vi, as it supports only 3270's > and half duplex ASCII terminals. A development version of UTS that > we use in the lab supports full duplex terminals, and thus vi and uucp, > through special software modifications in the 4705 communications > control (frontend). Thus while UTS 2.x does not have full-duplex support, its successor will. Also UTS 2.x has full-screen editors for 3270-type terminals (which is the most common terminal used with 370-type CPUs), as well as other such full-screen utilities, so your analogy to "tset dumb" is not fair to that implimentation. -- Gordon A. Moffett ...!{ihnp4,hplabs,amd,nsc}!amdahl!gam 37 22'50" N / 122 59'12" W [ This is just me talking. ]
herbie@watdcsu.UUCP (Herb Chong, Computing Services) (11/21/84)
will and is are two different things. we were discussing is, if i remember correctly. I applaud the future enhancements, because they are needed, but they aren't here now. The truism that "what you really want is in the next release of software" sort of applies here. when that release gets shipped, then the opportunity for large IBM and IBM-compatible systems to run a unix environment for teaching is greatly improved. Then, you may see a further increase in the teaching of computer science directly in such an environment. For the moment, we will have to settle for greatly overloaded VAXen. Herb Chong...
dmmartindale@watcgl.UUCP (Dave Martindale) (11/21/84)
> when that release gets shipped, > then the opportunity for large IBM and IBM-compatible systems to run a unix > environment for teaching is greatly improved. Then, you may see a further > increase in the teaching of computer science directly in such an environment. > For the moment, we will have to settle for greatly overloaded VAXen. > Herb Chong... The problem is that the computer science department doesn't have enough VAXen to do any more than a small amount of its teaching on. And, unless arrangements have changed in the last few years, the math faculty doesn't have complete control over its computing budget anyway - some of it is in the form of pseudo-money that HAS to be spent on computing centre resources. We couldn't stop using the computing centre for teaching courses even if we wanted to. Given that, I am concerned about the possibility of making the computing centre a nicer environment for teaching.