wert.rice@rand-relay@sri-unix.UUCP (09/22/83)
From: Scott Comer <wert.rice@rand-relay> Hear now, what are you complaining about? Don't you enjoy using a system that is buggy, flaky, and downright unreliable? Doesn't your heart just go pitty-pat at the prospect of digging into the kernel code for the answer to yet another annoying "problem" (read that as Undocumented Feature and/or Restriction)? I find it strange that Unix is the computer scientists system, but its design philosophy flys in the face of most of the time-honored rules of software design, implementation, and documentation. We daily commit the sin of not practicing what we preach, by telling our beginning computer science students about the virtues of structured design and the importance of good data structures, while forcing them to use a system that is held together only by the considerable efforts of 3 systems programmers and lots of bubble gum and bailing wire. I realize this is heresy to say things like this. You'll probably point out that I wouldn't have even be able to read your message if it wasn't for fine Unix software. On the other hand, I will point out that many systems are capable of Unix-like functionality, except that they are lacking a certain sexy user interface. On the other hand, they are reliable, well documented, maintainable, and supported. By a responsive and responsible group of people. THEY answer their telephone, which is more than I can say for Berkeley. How long is this community going to put up with this utter piece of software trash?
mrose.uci%rand-relay@sri-unix.UUCP (09/23/83)
From: Marshall Rose <mrose.uci@rand-relay>
Grrr -
Your message was so hot that my terminal almost caught on fire. As
it was, it took me 5 minutes to decide if I should reply to you,
and another 10 to cool-off before doing so. Here goes.
I'd rather not bog down Unix-Wizards at this point in time with a
lengthy emotion-packed discussion of why operating system X is
better/worse that UNIX. Although I'm interested in comparing UNIX
to other systems, it seems lately that Unix-Wizards has been
getting a stream of messages about things which have been discussed
time and time again. Unfortunately, no one is saying anything new.
I find this pretty irritating, because this list is an invaluable
source of information, and I don't like it when my threshing yields
mostly chaff with little grain (even when I'm using a powerful mail
handler to grind my mail).
We're running 4.1a BSD UNIX on a couple of 750s, and I think its
great. Sure we have problems on occasion, and yes UCB doesn't
answer their phone (and UCI is a UC!). Despite this, (in my
opinion) 4.1a leaves most other operating systems in the dust. If
these other systems, which lack "a certain sexy user interface",
are so hot, then let them give users more than UNIX gives them. As
far as I'm concerned, put up or shut up! If someone has something
better, then people will stop using UNIX and starting using
whatever it is that's better.
Yes, BSD UNIX has its problems, but frankly, I'm grateful. I can't
imagine what computing would be (for me at least) like without
UNIX.
/mtrron%brl-vgr@sri-unix.UUCP (09/23/83)
From: Ron Natalie <ron@brl-vgr> Good god. UNIX holds no corner on the market. There is a difference between software production and computer science research. Most of the favorite hot shot operating systems used by the pioneering computer researchers fall into the same class (UNIX, MULTICS, ITS, etc.). The user community "puts up with this piece of trash" because they like it. If you want something in a nice package, feel free to look elsewhere. Those who have spent these past years getting their environment into something that is comfortable, convenient, and productive for them don't need people like you telling them that they are wrong. -Ron
dan@sri-tsc@sri-unix.UUCP (09/23/83)
That's funny, in the 12 years I have been in the "computer biz", I have NEVER EVEN SEEN a system that is "reliable, well documented, maintainable, and supported." Certainly not from Honeywell, DEC, IBM, or Univac. Where have you been hiding this marvel? -Dan Chernikoff
dan%bbncd@sri-unix.UUCP (09/27/83)
From: Dan Franklin <dan@bbncd> Multics--in particular, MIT-Multics, which was both well-documented, because it was a product, and well supported, because our bug reports went straight to the developers at CISL some two blocks away. (It's probably still that way, but I haven't been there for awhile.) It was also quite reliable, despite frequent new releases. It also maintained REAL compatibility; when I was there, there was at least one program which still ran, even though it was written on the original GE 645 hardware, running a (by now) truly ancient version of the software, and the sources had been lost long ago. It's too bad the hardware architecture is so archaic (though it has many special-purpose registers, it's only a one-accumulator machine) and Honeywell so inept about selling computers. The only useful thing UNIX has that Multics doesn't is pipes. Dan
bob%ucla-locus@sri-unix.UUCP (09/28/83)
From: Bob English <bob@ucla-locus> If the "Unix community" were willing to accept binary-only systems, and pay reasonable fees for both original software and continuing support, Unix would not be a "piece of software trash". Nor would most of us be using it. --bob--
cak%purdue@sri-unix.UUCP (09/29/83)
From: Christopher A Kent <cak@purdue> Name one such system. chris
gwyn%brl-vld@sri-unix.UUCP (09/29/83)
From: Doug Gwyn (VLD/VMB) <gwyn@brl-vld> We will put up with this "software trash", as you call it, as long as we like it better overall than the available alternatives.
edhall%rand-unix@sri-unix.UUCP (09/29/83)
I quote from your letter: >> I realize this is heresy to say things like this. You'll probably point >> out that I wouldn't have even be able to read your message if it wasn't >> for fine Unix software. On the other hand, I will point out that many >> systems are capable of Unix-like functionality, except that they are >> lacking a certain sexy user interface. On the other hand, they are >> reliable, well documented, maintainable, and supported. By a responsive >> and responsible group of people. THEY answer their telephone, which is >> more than I can say for Berkeley. >> >> How long is this community going to put up with this utter piece of >> software trash? My first reaction is to ask you why in the world you wish to make such complaints in `Unix-Wizards'? Seems a bit out of place, no? It's a bit late, and your miasmal letter just rubbed me the wrong way. I'm still trying to figure out if you merely want to vent your spleen over something--anything--or if you really do have some valid criticisms. In the former case, I guess it's better than kicking the dog. But if you really feel that UNIX is so very, very bad that it mystifies you as to how any sentient being can tolerate it, let's hear some specifics. Here is a list of questions for you: 1) Is the UNIX software really that unreliable compared to other systems? 1a) Which systems are better? (Your experience here seems atypical--I've seen 4.1BSD systems run for weeks, even months, continuously without breakdown or problem.) 1b) What are some of these bugs that have bitten you so bad? 2) Is UNIX *that* unstructured? 2a) What do you mean by `structured'? 2b) Which systems are better structured? 3) Is the UNIX documentation so incredibly worse than other systems? 3a) Which other systems? 4) Did you really expect Berkeley to provide free consulting? 4a) What vendors/systems provide such free consulting? (If you have a System III/V -flavored UNIX you can pay for full support from AT&T. Other consulting firms support BSD UNIX. Berkeley is a University, not a software vendor/consultant.) ...and so on and so forth. Constructive criticism is appreciated more than the other kind. I'll leave it to others to answer you point-by- point, if they find anything more than innuendo. I don't have the time. -Ed Hall edhall@rand-unix decvax!randvax!edhall
RSanders.Pascalx@denver@sri-unix.UUCP (10/02/83)
Multics.