Return to BSD News archive
Path: sserve!manuel!munnari.oz.au!uunet!spool.mu.edu!uwm.edu!psuvax1!rutgers!spcvxb!terry From: terry@spcvxb.spc.edu (Terry Kennedy, Operations Mgr.) Newsgroups: comp.unix.bsd Subject: Re: Catch What They're Saying About Us... Message-ID: <1992Sep25.054120.3966@spcvxb.spc.edu> Date: 25 Sep 92 09:41:20 GMT References: <19oe23INNqh0@agate.berkeley.edu> <VIXIE.92Sep23102423@cognition.pa.dec.com> <19ta0nINNj2q@agate.berkeley.edu> Organization: St. Peter's College, US Lines: 170 [I just _know_ I'm going to regret this. Oh well, what the heck...] In article <19ta0nINNj2q@agate.berkeley.edu>, wjolitz@soda.berkeley.edu (William F. Jolitz) writes: [Oddly enough, I can't find the referenced article by Paul Vixie on my system.] > I am puzzled as to why this has occurred, since I have always viewed > BSDi as irrelevent to the freely available software discussions. Well, I don't know of many folks who consider the OS to be the only thing of interest. Personally, my interests lie in developing / porting / using appli- cations and OS extensions. Thus, I develop under/for various operating systems, including 4BSD in various flavors, VAX/VMS, MS-DOS, RSTS/E, etc. One of the platforms I develop "freely available software" for is BSDI's product. How does the fact that the operating system is sold, not given away, influence either the quantity or quality of the freely available software available under it? > While I have known for some time that Mr. Vixie is very close to Mr's > Bostic, Karels, McKusick, and Adams, and tends to agree with their > views, I have been dissuaded by others from responding to his earlier > attacks in netnews, with the reason given that he is going through > severe personal problems. I felt sorry for him, and I thought he would > right himself eventually. Oh dear. Now I suppose I'll be found guilty of the same crimes. When will my sentence be announced? > I apologize to readers for what has occurred, and I ask the indulgence > of those who are impatient with the ugly turn this thread has taken to > please remember that both Bill and I want people to feel good about the > work their doing with 386BSD, and that while we sometimes cannot let > attacks go unanswered, we prefer to spend our time helping others > achieve their goals. From my point of view, the original discussion was a valid, if heated, tech- nical discussion until *someone* stated that if someone found 386BSD slower than Linux that there was something wrong with their reasoning processes, and then tried rabble-rousing in comp.unix.bsd with the "Catch what they're saying about us" topic. It seems that the vast majority of folks here are level-headed enough to not go storming off to comp.os.linux screaming for blood. Sorry to disappoint you... > There has been extensive trade press writings about the USL/UCB suit, > since USL (nee AT&T) and UCB (via CSRG/BSDi) are very large entities, > and have an enormous impact on a large number of people. Since we are > not a party to this suit, we are not of "topical" interest to the > press. You repeatedly attempt to disassociate yourselves from this suit. I wonder why? Surely you realize that if a verdict is reached that Net-2 contains USL proprietary material, you will have to cease distributing 386BSD? > BSDi is a commercial system, with a small user base, and is > proprietary, licensed, and expensive. BSDI *is* a commercial system. I am a BSDI customer. I pay for technical support. I report problems. I receive bug fixes in a prompt manner, similar to they way I receive support for other commercial software I own. [As a matter of fact, I find BSDI much better than most commercial vendors]. I neither expect nor demand that of 386BSD. However, as I'd like to get on with my application work, I'd like a system that was stable on my hardware. With my Unix background, on my hardware, the BSDI product worked fine, while I had problems wuth 386BSD. As always, people's mileage may vary. Proprietary? Depends on what you mean. If you mean that I can't make copies of the distribution tapes and give them away, then you're correct. If you're trying to imply that it is not compatible with other 4.3BSD-derived systems, I disagree. I've ported several large programs to it without any trouble. I expect that of all the available 386 Unix products, it will be the most compat- ible with 4.4BSD, which I plan on running on my VAX systems (as soon as the tape I ordered from CSRG gets here). Licensed? Yup. That's necessary in today's legal environment if you're in the commercial market. Expensive? I don't think so. It's probably beyond the average hobbyist's budget, but that isn't the market they're trying to compete in. By the way, I tried 386BSD first. I found a number of problems (which I've contributed to the bug+fix list). One of these was a major flaw [in my opin- ion] in the 0.1 scheduler which would likely explain why Linux users would think 386BSD was slow when doing background compiles. I decided that since my interest was in applications, that I'd save myself some money by buying the BSDI product, so I could get to work with my applications, rather than fixing OS things. > 386BSD was developed by Bill over four years and contributed to UCB > with the express purpose that it be made available to the BSD user > base, in order to allow people to use BSD on inexpensive platforms and > continue to work on new ideas. The system was ready for users in 1989, > but it's release was consistently put-off by CSRG. If it was ready in 1989, then why the long delay before 0.0? Assuming that that was entirely because of events outside of your control, let's progress to 0.1. If it was "ready for users" at 0.0, then 0.1 would just be a few quick bugfixes. However, we're well past 0.1's release and folks are still reporting some major problems. I assume you're working on these as fast as you can [no irony intended, I know what it's like to get bug reports converted to fixes] and that you will have the next release ready as soon as you can. However, that's a long way from 1989. That's the nature of software development. It's not your fault, it's not CSRG's fault, and it's not the user's fault. I'm just pointing out that while you had a workable system in 1989 [see, I'm taking your word for it], that there was a lot more to be done to have it work consistently well for the maj- ority of users who try it. Brooks has a bit to say about the difference be- tween a "program", a "programming system", and a "programming system product" in "The Mythical Man-Month". Fascinating reading. > BSDi, on the other hand, is probably the most adversely impacted by > both 386BSD software activities and Careware charity activities. As I said above, I believe you're playing to two different audiences. I'm at a loss to understand the relation that would have BSDI (or anyone else) "impacted by ... Careware charity activities". > By taking 386BSD and twisting it their own proprietary and costly > product, BSDi cannot easily match the rapid growth and learning curve > which a freely available version allows. Also, since the systems have > diverged (and will continue to diverge), they cannot easily take > advantage of new work in 386BSD, unless they decide to appropriate > whole portions of the system. In this case, their much-touted > leading-edge firm appears like the tail wagging the dog. In addition, > even with a small paid staff, a single small company cannot compete > with the hundreds of thousands of users making changes and adding new > things to the system every day. Now you're mis-stating the facts. From my point of view, BSDI looks like a high-quality port of Net-2 to the 386 architecture. They are comitted to tracking future changes in the base CSRG code. From examining 386BSD code, there appear to be major pieces where Bill decided to implement something different from the way CSRG did. I'm not saying that his way is better or worse, but it is different. Now, tell me again which product is incompatible? > But this is again, irrelevent, as BSDi is not a freely available > system. They do not put their changes on the net, but instead hold > them to themselves closely. They have received many changes from CSRG > (one and the same with BSDi), but are not distributing them. They have > even received Chris Torek's sparc code, paid for by the taxpayers, > which has not been made available to others despite repeated requests, > and are currently attempting to make money off of it in the same manner > in which they tried to appropriate 386BSD. Which way do you want it? Earlier you imply that the "relationship" you feel exists between CSRG and BSDI was being exploited to leak unreleased parts of future CSRG releases to BSDI. Now you're complaining that they aren't releasing them? I'd assume that there is some sort of agreement be- tween CSRG and BSDI, even if it's the standard CSRG license. Perhaps BSDI is constrained from releasing the code? If you have a complaint with Chris Torek's code not being available, take it up with Chris. Why do you expect BSDI to be the distribution point? > If we hadn't spent the time making sure 386BSD was completed and > distributed, it still would not be available either. And the best thing you can do is to continue your efforts. The world needs a diversity of software. Your effort should speak for itself, and you need not stoop to name-calling and allegations. > As I look over again this entire message, I can only assume that this > diatribe against me and 386BSD is merely the desperate and jealous > rantings of a person who resents deeply the success that 386BSD has > achieved, as well as our personal honesty and commitment to this > project. Why do I get the same feeling I have when rock stars start trying to impart their political views at concerts? Can we just stick to the code, please? Terry Kennedy Operations Manager, Academic Computing terry@spcvxa.bitnet St. Peter's College, Jersey City, NJ USA terry@spcvxa.spc.edu +1 201 915 9381