punia@uvm-cs.UUCP (David T. Punia) (05/03/85)
So why MUST these poor kids be FORCED to buy THAT micro? At least partly because of the power of the almighty buck. In order to obtain the magnitude of discount that was negotiated, AT&T, and any other vendor, I imagine, needed some guarantee of volume. This is the way things work in business. That the machines are available at the price they are makes me wonder why anyone would WANT to choose a different machine! OK, so what about the kids who already have their own micros? Sorry, courseware being developed for the AT&T is taking advantage of it's 640 X 400 graphics resolution. As far as I'm concerned, any high school kid who already owns a PC of this class is not hurting financially to begin with. Besides, he/she can probably sell it for what the AT&T will cost. It should certainly NOT be UVM's responsibility to function as a used equipment dealer. I'm a Rainbow fan myself (let's not talk about it), but it seems silly, given the tremendous amount of software that's going to be generated, not to have an AT&T. As of now, I don't know of any AT&T compatibles, so it's a non-issue. Anybody want to buy a used Rainbow? Why the AT&T? I was not on the advisory committee, so I can only speculate. It's faster and comes with far more standard equipment than the "standard" it emulates. It gives you seven compatible expansion slots that can be made into 16-bit slots. Main memory can be fully expanded to 640K without using any of the expansion slots. This leaves plenty of room for upgrades, including new processor boards, to help fend off obsolescence. And AT&T is not likely to go belly-up in the foreseeable future. Bell Labs is a pretty strong support group, and support has been guaranteed. -- REPLACE THIS LINE WITH YOUR FAVORITE DISCLAIMER David T. Punia, Dept. of Computer Science & Electrical Engineering, The University of Vermont, Burlington, VT 05405 802-656-3330 USENET --> ....!decvax!dartvax!uvm-gen!punia CSNET ---> punia@uvm
chenr@tilt.FUN (Ray Chen) (05/05/85)
I'm a CS major. At UVM, this would put into the category of students required to buy an AT&T 6300. Now, I'm sorry, but an AT&T 6300 would be pretty much useless to me as a personal computer. Given the type of projects I work on (screen editors, artificial intelligence, compilers, etc.) the 6300 simply doesn't have either the cycles, the disk, or the memory that I need. It would make a great word-processor and a nice terminal emulator -- and that's about it. For most of the mathematical calculations I have to do, I prefer my HP 34C. I have quite a few friends here (like my graphics-oriented friend who thrashes a Vax to death) who are in the same position. We all regard a well-tuned SUN-class machine as a nice personal computer. I would deeply resent any college that would make me blatantly waste $2000 no matter how well off I was financially. Ray Chen
bradley@ut-sally.UUCP (David K. Bradley) (05/05/85)
In article <380@uvm-cs.UUCP> punia@uvm-cs.UUCP (David T. Punia) writes: > > So why MUST these poor kids be FORCED to buy THAT micro? > > At least partly because of the power of the almighty buck. In >order to obtain the magnitude of discount that was negotiated, AT&T, and >any other vendor, I imagine, needed some guarantee of volume. This is >the way things work in business. That the machines are available at the >price they are makes me wonder why anyone would WANT to choose a different >machine! > For those of you that are interested, via a special deal with Apple students (and faculty/staff, I beleive) at the University of Texas at Austin can buy a Macintosh with 128K, 1 external drive (in addition to the internal drive), Macwrite, Macpaint, and PRINTER, for ~$1600. The same package with 512K is ~$2200. Admittedly this does not include any compilers and thus is not very useful to students in technical majors (without their shelling out more $$), however the machine has a lot of potential (68000 processor, good graphics, mouse, graphics printer, small enough to fit on a dorm desk :-)), and once some good and inexpensive software becomes available it will be hard to beat. David K. Bradley bradley@ut-sally.UUCP
punia@uvm-gen.UUCP (David T. Punia) (05/06/85)
>Now, I'm sorry, but an AT&T 6300 would be pretty much useless to me as >a personal computer. Given the type of projects I work on (screen >editors, artificial intelligence, compilers, etc.) the 6300 simply >doesn't have either the cycles, the disk, or the memory that I need. You don't sound like a freshman, or even a sophomore to me. Keep in mind that the basic system is adequate for all of the students for several years, at least. As far as doing AI and other more involved types of activities, remember two things: 1.) It can make a good workstation to communicate with the bigger resources on campus. 2.) Seven expansion slots. They may not be good for 20 MHz, but there's still a lot that can be done with them. The 6300 will certainly never offer the power of a SUN, but it is felt to be a suitable workstation for the vast majority of the applications that will be run by undergraduates, including CS majors. -- David T. Punia, Engineering Specialist, Univ. of Vermont CS/EE dept. USENET --> !decvax!dartvax!uvm-gen!punia CSNET ---> punia@uvm
chenr@tilt.FUN (Ray Chen) (05/08/85)
In article <239@uvm-gen.UUCP> punia@uvm-gen.UUCP (David T. Punia) writes: > > You don't sound like a freshman, or even a sophomore to me. >Keep in mind that the basic system is adequate for all of the students >for several years, at least. As far as doing AI and other more involved >types of activities, remember two things: > > 1.) It can make a good workstation to communicate with the bigger > resources on campus. > > 2.) Seven expansion slots. They may not be good for 20 MHz, but > there's still a lot that can be done with them. > >The 6300 will certainly never offer the power of a SUN, but it is felt to >be a suitable workstation for the vast majority of the applications that >will be run by undergraduates, including CS majors. Some background information. I'm a senior CS major. As a computer, a 6300 would have become useless as of the second half of my sophmore year. The same holds true for many of my friends. Now, I'm not arguing that a word-processor or terminal emulator/file transfer machine is a bad thing to have. However, I am claiming that as a computer, a 6300 will only be able to satisfy a limited set of needs. I think people should realize that a Business major, an Engineering major, a Math major, and a Computer Science major are all going to require different levels and types of computational power, and that no one machine is going to be cost-effective for all of them. For most Business majors, a 6300 should be adequate for at least 4 years. A CS major will be lucky if the 6300 is good for two. By his junior year at the latest, he'll be doing simulations, compiler hacking, kernel building, and who knows what else. So why is UVM forcing some people to spend money on an obsolete machine that within two years will be useless? If somebody wants to argue that the 6300 will be still useful as a file transfer machine I'll agree with him. I'll also ask him to rephrase the UVM pc announcement as requiring certain incoming freshman to purchase an AT&T terminal emulator/file transfer machine for $2000. (Please remember that without a printer you can't do word-processing unless at some point you transfer a file to a machine with a printer. As UVM doesn't require purchase of a printer, that makes the 6300 relatively useless for local word-processing unless the poor or soon-to-be-poor student pours more money into his machine.) Ray Chen princeton!tilt!chenr
jss@sjuvax.UUCP (J. Shapiro) (05/12/85)
> > >Now, I'm sorry, but an AT&T 6300 would be pretty much useless to me as > >a personal computer. Given the type of projects I work on (screen > >editors, artificial intelligence, compilers, etc.) the 6300 simply > >doesn't have either the cycles, the disk, or the memory that I need. > I sat back and agreed with this one, but seeing the response, I want to say something... > You don't sound like a freshman, or even a sophomore to me. I am a Junior at Haverford College, and I have thought the IBM PC was a joke as a micro since it was introduced. It is still unable to solve any problems which were then remotely interesting to me. Please note that at the time I was in *HIGH SCHOOL*, and I was far far far from the only person who recognized the silly limits of the box, and who counseled people against the purchase of that machine. Regrettably, many people did not realize those limits, and we are now stuck with this cruddy "standard" of nonfunctionality. Consider that the segment size limit is a real problem. Even Intel admits it. To me a minimally acceptable working system for my own day to day use would be a 68000/68020 or 32016/32032 with appropriate MMU and FPU, at least 1Meg of memory, and at least a 20Mbyte hard disk with an access time no higher than 30 ns. It should be single user, but multitasking. I own a Mac, and I am waiting for the Hard drive, soon to be available from General Computers. I deem it barely satisfactory, even with 512k, and highly overpriced. The lack of an MMU is a crucial piece of brain damage. > Keep in mind that the basic system is adequate for all of the students > for several years, at least. Bullshit. I can think of 10 people in my class of 246 who would drive that machine into the ground. On the other hand, those 10 can badly abuse VAXen too. Note please that Haverford is not a school much into CS. At any school with a CS department your percentage of CPU killers is likely to be far far higher. Also note that the machine you are having students buy doesn't even have enough real horsepower to do single user text formatting at a reasonable speed, and that for this reason, no one has even tried to come out with a decent text formatter for the 8086. Microsoft Word is very good, but also VERY SLOW. One of mny main reasons for buying a Mac is that it has enough flexibility for me to write a decent formatting package. > As far as doing AI and other more involved > types of activities, remember two things: > > 1.) It can make a good workstation to communicate with the bigger > resources on campus. So what. If I am doing personal AI work I certainly don't want to use Haverford's machinery, because this would render any of my results partially the property of the College. This policy at the University of Pennsylvania stops students from persuing a lot of good work. I sincerely believe that my ideas are worth something (so do my clients...), and I certainly would not wish to "give" them to a University when they are potentially worth a great deal to me commercially. For this reason I have been sitting on several ideas until I can purchase adequate hardware. > 2.) Seven expansion slots. They may not be good for 20 MHz, but > there's still a lot that can be done with them. > Not with the segment size of 64K. Most useful languages are stack based, and good bloody luck doing a Lisp implementation that runs acceptably on a small segment architecture. Consider also the memory addressing limitations of DOS. > The 6300 will certainly never offer the power of a SUN, but it is felt to > be a suitable workstation for the vast majority of the applications that > will be run by undergraduates, including CS majors. I agree that 95% of all people will find this arrangement entirely satisfactory. For the remainder you are wasting (by the time they have purchased enough hard disk and memory to avoid screaming every time they turn on the box) something like $8,000. This would really piss me off, and the lack of consideration it implies, and the lack of concern about my hard earned money would most certainly stop me from applying to your institution. I submit that the University of Vermont would do well to consider allowing students more flexibility in purchasing their machinery. You wouldn't insist that everyone drive a Porsche, and neither do I, but I also wouldn't dream of mandating that everyone purchase only a Deux Cheveux (ask your local French department - the Deux Cheveux is noted for the fact that it is cheaper to replace than to repair, and drives like it). The AT&T is not a bad box - in particular it does not claim to be anything more than it is. It is, however, inadequate for enough applications that it should not be the only choice available. Jonathan S. Shapiro Haverford College System Manager and Computing Consultant
herbie@watdcsu.UUCP (Herb Chong [DCS]) (05/13/85)
In article <1144@sjuvax.UUCP> jss@sjuvax.UUCP (J. Shapiro) writes: (while describing his ideal personal) >and at least a 20Mbyte hard disk with an access time no higher >than 30 ns. i could use a few that fast on our vax. surely you mean 30ms. as for CPU killers, i did some stuff for a 4th year project that required just over 2 CPU hours on an IBM 3081. using a few rough scalings, that's on the order of 50 CPU hours on a 11/780, for one run. i needed to do 3 correct ones. there are a few of us that will run any CPU made into the ground. i would not even dream of using anything less than a vax even for scaled down testing. i think that the UVM decision was based on the "average" student needs in the particular faculties and was as much political as technical. what were the technical qualifications of the people who made the recommendations and the decisions for this action? if this has already been discussed (i don't remember seeing it), then please send e-mail. Herb Chong... I'm user-friendly -- I don't byte, I nybble.... UUCP: {decvax|utzoo|ihnp4|allegra|clyde}!watmath!water!watdcsu!herbie CSNET: herbie%watdcsu@waterloo.csnet ARPA: herbie%watdcsu%waterloo.csnet@csnet-relay.arpa NETNORTH, BITNET, EARN: herbie@watdcs, herbie@watdcsu
jss@sjuvax.UUCP (J. Shapiro) (05/16/85)
> In article <1144@sjuvax.UUCP> jss@sjuvax.UUCP (J. Shapiro) writes: > (while describing his ideal personal) > > >and at least a 20Mbyte hard disk with an access time no higher > >than 30 ns. > i could use a few that fast on our vax. surely you mean 30ms. > Indeed I do mean 30ms, and thank you for the correction. Jon Shapiro
djb@gatech.CSNET (David J. Buechner) (05/17/85)
> OK, so what about the kids who already have their own micros? > > Sorry, courseware being developed for the AT&T is taking advantage >of it's 640 X 400 graphics resolution. As far as I'm concerned, any >high school kid who already owns a PC of this class is not hurting financially >to begin with. Besides, he/she can probably sell it for what the AT&T will >cost. It is ludicrous (sp?) to suggest that selling one's micro of choice is a method available to finance a required AT&T. I would very much resent Ga. Tech telling me that I bought a machine that won't do me any good and that they in their (mythical) wisdom have found an AT&T micro that is much better for me. If students are going to be required to purchase a micro they should be allowed to exercise their own intelligence and get a machine they feel fulfills their needs both for the classwork for which the machine is being required and for other work that they may be doing. The institute would then have to write coursware which can be used on a number of micros. If they cannot currently do this they should not require students to buy micros. (small flame on) It is also invalid to make such a sweeping over generalization of the state of affairs to assume that any student coming out of high school with a micro is financially anything, well off or otherwise. (small flame off) ------------------------------- (some disclaimer required) David Buechner UUCP : ...!{akgua,allegra,emory,rlgvax,sb1,ut-ngp,ut-sally}!gatech!djb ARPA : djb.gatech@CSNet-Relay CSNET : djb@gatech Federal snail : Ga. Tech P.O. Box 33336, Atlanta, Georgia 30332
roy@phri.UUCP (Roy Smith) (05/19/85)
> If students are going to be required to purchase a micro they should > be allowed to exercise their own intelligence and get a machine they > feel fulfills their needs both for the classwork for which the machine > is being required and for other work that they may be doing. One of the many things a college does for its students is prevent them from making bad decisions (or try to, anyway). Who is better able to judge what machine is more suitable for someone's college career; a kid fresh out of high school or the people who are planning his/her courses for the next 4 years? Replace "micro" with "textbook" in the above quote and see how absurd it sounds. Yes, I know that a PC costs 100 times what a textbook does, but you only buy 1 PC for your college career, while you might very well buy 100 textbooks. -- allegra!phri!roy (Roy Smith) System Administrator, Public Health Research Institute
chenr@tilt.FUN (Ray Chen) (05/21/85)
In article <235@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: >> If students are going to be required to purchase a micro they should >> be allowed to exercise their own intelligence and get a machine they >> feel fulfills their needs both for the classwork for which the machine >> is being required and for other work that they may be doing. > > One of the many things a college does for its students is prevent >them from making bad decisions (or try to, anyway). Who is better able to >judge what machine is more suitable for someone's college career; a kid >fresh out of high school or the people who are planning his/her courses for >the next 4 years? > > Replace "micro" with "textbook" in the above quote and see how >absurd it sounds. Yes, I know that a PC costs 100 times what a textbook >does, but you only buy 1 PC for your college career, while you might very >well buy 100 textbooks. For a rebuttal of Roy Smith's argument, replace the word "micro" with "calculator" in the first paragraph. Ray Chen princeton!tilt!chenr
desjardins@h-sc1.UUCP (marie desjardins) (05/22/85)
Roy Smith: > Who is better able to > judge what machine is more suitable for someone's college career; a kid > fresh out of high school or the people who are planning his/her courses for > the next 4 years? The kid fresh out of high school, probably. The bureaucrats making these decisions almost never know anything about technology. There's nothing wrong with making a recommendation, but every student has different needs, all of which may not be satisfied by the PC chosen by the administrators. marie desjardins
punia@uvm-gen.UUCP (David T. Punia) (05/22/85)
Mr. Chen's "rebuttal" to Mr. Smith overlooks some very important distinctions between the functions of calculators, textbooks and micros. A textbook is "programmed" uniquely with specific information in a specific order to go along with the "program" an educator desires to use in a given class. The calculator is a far more general purpose tool that operates at a much lower level than either the textbook or the programmed micro. The choice of calculator that a student uses in classes is a far less critical one than the choice of text. Imagine the chaos if students were to choose their own textbooks. Some of the same arguments apply to microcomputers. The courseware for a class will comprise both textbooks and microcomputer software specifically chosen to fit an instructor's program. To repeat an argument already made, it is very much a matter of practicality that students all use a micro that can run the same software. It would be tremendously impractical to expect the faculty who write the software, or the bookstore that must distribute it, to carry a multiplicity of versions of the same product to meet the needs of those students who feel they are the better judges of what equipment they need to learn about things they know very little about. Such is usually not the case with a calculator. Even if a student does go into an engineering class with a four function calculator, it is not usually beyond the financial means of the student to replace it with a more appropriate tool. Certainly that would not be the case for a student whose choice of a micro could not meet the needs of his curriculum. It has to be assumed that the people responsible for the curriculum know at least a little more than those they are planning to educate. The committee that recommended the micro to be used at UVM comprised 12 individuals drawn from each of the areas that will be requiring the micros, as well as a few administrative personnel. These were not people unfamiliar with micros to begin with, but people who have already been involved with microcomputing in an educational environment, or at the very least, well versed in computing in general. There has to be an element of trust in the expertise of these individuals, in their knowledge of their own curricula, in their acquired or existent expertise in microcomputing, and in their desire to select the machine that best met their educational needs, now and in the forseeable future. Not to mention their probably better ability to resist being sold a bill of goods by a particularly convincing salesman. And not to mention the fact that evaluation machines were available for several months from numerous vendors. Who is better able to select? A kid fresh out of high school? A very special kid, indeed! -- ***REPLACE THIS LINE WITH YOUR FAVORITE DISCLAIMER*** David T. Punia, Dept. of Computer Science & Electrical Engineering, The University of Vermont, Burlington, VT 05405 802-656-3330 USENET --> ....!decvax!dartvax!uvm-gen!punia CSNET ---> punia@uvm
csc@watmath.UUCP (Jan Gray) (05/24/85)
In article <240@uvm-gen.UUCP> punia@uvm-gen.UUCP (David T. Punia) writes: > . . . Who is better able to select? A kid fresh out of >high school? A very special kid, indeed! > The "very special kid"s will avoid your University because they do not wish to purchase a computer that doesn't suit them. I would. "I do and do and do for you kids, and *this* is the thanks I get!" (temporarily not) Jan Gray (jsgray@watmath.UUCP) University of Waterloo (519) 885-1211 x3870
cdshaw@watmum.UUCP (Chris Shaw) (05/24/85)
>>In article <240@uvm-gen.UUCP> punia@uvm-gen.UUCP (David T. Punia) writes: >> . . . Who is better able to select? A kid fresh out of >>high school? A very special kid, indeed! > >The "very special kid"s will avoid your University because they do not wish >to purchase a computer that doesn't suit them. I would. > Jan Gray Well, be that as it may, I have yet to meet anyone fresh out of high school who knew enough to make a good decision about what machine to buy. Sure, your average smart-ass kid has his (her) preferences, but such preferences are generally so narrow-minded as to defy rational belief. This is the typical result of exposure to only one system during high school. Of course, this is exactly what you will get if all your students use just one system throughout university: One-system hacks who can do anything on a 6300, but put them on a different box and they're dead in the water. On the other hand, people claiming to need lots of CPU are just playing around, I think. I asked one of these people what required all this power, and he told me that he once ran a simulation that took ~5 hours of 3033 time. I assume he was an undergrad at the time. It only later struck me that 5 hours of mainframe at the cheapest rate here will cost in excess of $500. It's obvious that this person's prof wasted considerable amounts of resources by assigning a ridiculous project. No doubt the person involved could have learned an equal amount by doing a simulation taking 10 minutes or less. I guess my main point is that there is really no reason that student projects for course credit should be of the same degree of complexity or of a size equal in hugeness to what one would find in real life. The problems one does need not be mere toys, but for a compiler course, it seems foolish to assign an ADA compiler when a Pascal subset will do for pedagogical purposes. Chris Shaw watmath!watmum!cdshaw or cdshaw@watmath University of Waterloo In doubt? Eat hot high-speed death -- the experts' choice in gastric vileness !
spaf@gatech.CSNET (Gene Spafford) (05/25/85)
What I think many of you are failing to question is the nature of the
programs at UVM -- what kind of courses do they teach? What kind of
major equipment do they have on campus? It is entirely possible that
they are like the majority of educational institutions in this country
with limited computational resources and a limited faculty. For the
kinds of courses they are able to offer and support, an AT&T 6300 may
be an entirely appropriate processor which will allow them to offer a
better education than they can right now.
Not a lot of schools have mutliple Vaxen and networks of Suns or
Apollos. May institutions are limited to a few 11/45's, a small DG or
Prime, and some cycles on the central administration machine. 10
student AT&T 6300s effectively doubles the campus computing power
available for instruction....
I'm not claiming that this is the case at UVM, but I know it is the
situation at a number of small colleges and universities, especially
the colleges that don't have the resources to obtain large research
grants to buy equipment. For many institutions, an AT&T 6300 has
plenty of power for an undergraduate throughout his/her degree work.
The school I did my undergraduate work at had a Prime 550-I as THE
computational research facility for the whole college. Despite that,
CS majors got a pretty good education (though it was short on
practical experience) due to the talent and imagination of the faculty.
With that kind of alternative, making every student have a 6300
is a great idea.
--
Gene "3 months and holding" Spafford
The Clouds Project, School of ICS, Georgia Tech, Atlanta GA 30332
CSNet: Spaf @ GATech ARPA: Spaf%GATech.CSNet @ CSNet-Relay.ARPA
uucp: ...!{akgua,allegra,hplabs,ihnp4,linus,seismo,ulysses}!gatech!spafpaul@osu-eddie.UUCP (Paul Placeway) (05/27/85)
[bug juice] I have a question on the subject... If the softawre needs to run on all the student machines, isn't it more rational to write the programs in something portable and port it to a group of machines rather than just the one? Isn't this the reason that Unix was developed to the extent that it is??? As a for instance: It is easy to make a nice, but machine independant (from the outside) screen and keyboard routine package for things like Apple ][s, IBM PCs & clones, Macs, etc. and write programs that use these as the basis for development. If you want to know more about portable software, I am sure that both AT&T Bell Labs, and Columbia U. (the kermit folks) can tell you all you want to know... A grad student friend of mine got an Apple //c to write her dissertation on and is very happy with it. I, on the other hand, find the Apple far too slow for my own work (start a compile on one, go out, have lunch, come back, wait... 8-) and would rather spend the extra $$$ for a better machine. The analogy that a micro is like a text book dosn't hold because I have bought text books that were not for a class that I will ever take, because the book is a good refrence. As for calculators, a Physics prof. that I had recomended TI 58s, but an EE prof recomended getting an HP (pref. 41cv or cx). I, by the way, own a 41cv... The correct solution to University courseware is NOT to force all the students to buy the same machine, but rather to develop portable programs and offer a (perhaps limited) choice of what to get. Why should I as a student be forced to buy an AT&T when I could get a Mac for the same ammount of money? Paul Placeway The Ohio State University, CIS Dept. ...!cbosgd!osu-eddie!paul paul@ohio-state (CSNet) Standard disclaimer: the opinions above are MINE.