Return to BSD News archive
Path: sserve!manuel.anu.edu.au!munnari.oz.au!news.hawaii.edu!ames!haven.umd.edu!uunet!math.fu-berlin.de!unidui!du9ds3!veit From: veit@du9ds3 (Holger Veit) Newsgroups: comp.unix.bsd Subject: Re: Shared Libraries for 386BSD (long, w/source) Date: 28 Nov 92 10:05:54 GMT Organization: Uni-Duisburg FB9 Datenverarbeitung Lines: 117 Message-ID: <veit.722945154@du9ds3> References: <lohse.722861707@tech7> <JKH.92Nov27145531@whisker.lotus.ie> <veit.722878428@du9ds3> <1992Nov28.045223.1045@pool.info.sunyit.edu> Reply-To: veit@du9ds3.uni-duisburg.de NNTP-Posting-Host: du9ds3.fb9dv.uni-duisburg.de In <1992Nov28.045223.1045@pool.info.sunyit.edu> ujlh@pool.info.sunyit.edu (James Henrickson) writes: >In article <veit.722878428@du9ds3> veit@du9ds3.uni-duisburg.de writes: >> >>You may use this code, it might work, but you should expect a number of >>changes in 0.2 (I don't comment on this), which might cause a large number >>of things work in a different way, or no longer at all. This applies in >>particular to some (source) postings in the last few days or weeks, which are >>mainly results of "the unknown, lonely, uninformed hacker" and are often >>in no way acknowledged by THE WIZARD @ soda.berkeley.edu. [...] >I see nothing wrong with the latest posting of source code. I understand >that 0.2 might support shared libraries, but who is getting hurt by adding >shared libs to 0.1? Isn't that what research (and hacking) is all about? >I doubt this source code will have any effect on 0.2, and I doubt it will >have any effect on those people working on shared libs for 0.2. With all >of the anticipated changes in 0.2, wouldn't it make sense to just install >it from scratch? I seriously doubt it will be feasible to distribute >0.2 as a HUGE series of patches to 0.1, so I doubt this source code will >have any effect on 0.2. Read my first line. Everyone who can capture this code into a file and compile it can and may use it. The problem I see with uncoordinated work is that there are "hackers" and "users" in this group. The first group does not care whether they have to start the big compile overnight after changing something, or retrieving tons of code after a "weekend patch" (there is some other OS I am aiming at here). The other group, the users, want to see stability, i.e. not ten different interfaces, and they are likely the ones who try to force the developers into the portability direction which will really cripple improvements, because everyone who wants to develop a main piece of software (not simply a port of a relatively independent thing like for instance a newsreader) has to look at the present state and has to preserve it ad nauseam (you know the Intel hardware line, and the DOS-WINDOWS-OS/2-NT line?), just because somebody refuses to modify his software to a new standard (or because the original author is no longer interested or present). >You mention "people that mainly want to use 386bsd." If they are USERS >and not HACKERS, they shouldn't even try adding the source code. Wouldn't >it just be sufficient to append a "THIS IS FOR HACKERS ONLY" disclaimer >at the end of such an announcement? :-) I doubt this would help for such a long awaited and delicate area like shared libraries. Furthermore, I don't know how many people now start to improve the given pieces of codes, and add the missing parts independently from each others, and again throw half-cooked versions and bug-fixes on the market. Terry's patchkit was made to do some coordination on these patches floating-around, but it cannot deal with competing, incompatible sets of basically the same software. >Furthermore, you mention the threat of intervention by AT&T or USL or >somebody (don't remember for sure, I deleted that part already). If >this particular implementation of shared libs is kept separate from the >official 386bsd distribution, how would 386bsd be affected? I've seen >a lot of "questionable" add-on source code for operating systems floating >around, but I haven't seen the operating system people get dragged into >court over it. You miss the point here. This section does not address the existence or not-existance of some implementation in the offical distribution. The threat may or may not become true if some marketing person even believes in his stupidity that 386bsd may cut off any small piece of a marketing segment. Besides the backward-compatibility area that has been entered by the new (IMHO un-) improved pccons, which makes new research more complicated, an interface that more and more resembles one of the existing versions of UNIX, can make commercial vendors (value added resellers) think to bundle their high-tech product with a cheap 386bsd instead of the highly overpriced and in some areas more probalby buggy SVR4. USL's claim that BSDI uses proprietary code was only an anchor to hold them down, but the real warplace is the expected marketing segment. >I think that you made some good points, but I believe you are being >overprotective of the centralized development of 386bsd. Must there >be such a solid line drawn between users and official developers? >There are many of us that fall in the middle, and we are the ones you >seem to be speaking out against. I admit that I haven't done a lot >of kernel hacking, but what if I did something that I thought was >beneficial? Some of us would like to experiment with new features >NOW. Are you suggesting we use a different operating system, or that >we should start a "comp.os.386bsd.unofficial.hackers" newsgroup? I myself are not very happy about the policy of absolute secrets, but I understand Bill in that he wants to have a solid base of development and research and not evolutional growth and dozens of slightly different childrens of BSD. BSD4.3 has evolved from V6 to the point where it is now, by discipline and coordination of the developers, and not by experimenting into all possible directions. The latter appears to show fast results, but will undoubtedly cause large cuts and reorganisations in the long term. I think there shouldn't be a deep chasm between users and developers, and as you correctly state, most persons do not strictly belong to one of these groups. As I said I am asking for a bit more discipline and coordination, just as it is common in larger development or research projects. The point is not that "you think that your work is beneficial", but that the majority can live with it. You love your artwork, and it does work for for your own requirements, and you might not see the overall picture and vision the "WIZARDS" have. Nobody will forbid you to modify your own version of 386bsd to the degree that you might call it 386jlh, but if you do and post it be careful not to destroy the general line. You are no longer speaking for 386bsd then. 386bsd is "Jolix". If you cannot accept this, this newsgroup and this OS is not your OS. I accept this policy (at least for technical issues, my opinion remains my own opinion). >"This had to be said." :-) >-- >Jim H. >* >* James L. Henrickson | "I don't need a signature, I need a job!" >* ujlh@sunyit.edu | BSCS, December 1992 Holger -- | | / Dr. Holger Veit | INTERNET: veit@du9ds3.fb9dv.uni-duisburg.de |__| / University of Duisburg | "XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX | | / Dept. of Electr. Eng. | Sorry, the above really good fortune has | |/ Inst. f. Dataprocessing | been CENSORED because of obscenity"