*BSD News Article 56746


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!nntp.coast.net!lll-winken.llnl.gov!trib.apple.com!agate!boulder!rintintin.Colorado.EDU!vilhuber
From: nelsoni@rintintin.Colorado.EDU (Ian S. Nelson)
Newsgroups: comp.os.linux.advocacy,comp.unix.bsd.freebsd.misc,comp.unix.advocacy
Subject: Re: Linux vs FreeBSD
Date: 5 Dec 95 22:42:26 GMT
Organization: University of Colorado at Boulder
Lines: 52
Distribution: comp
Message-ID: <vilhuber.818203346@rintintin.Colorado.EDU>
References: <489kuu$rbo@pelican.cs.ucla.edu> <49smvs$8gd@josie.abo.fi> <49tban$978@times.tfs.com> <4a10kr$k06@mark.ucdavis.edu>
NNTP-Posting-Host: rintintin.colorado.edu
X-Newsreader: NN version 6.5.0 #8 (NOV)
Xref: euryale.cc.adfa.oz.au comp.os.linux.advocacy:30142 comp.unix.bsd.freebsd.misc:10336 comp.unix.advocacy:12131

obrien@cs.ucdavis.edu (David E. O'Brien) writes:

>>Am I correct to think that the FreeBSD "equivalent", this CVS or
>>whatever you called it, can't be _read_ except by a small core team?
>>Whatever for? Keeping people from making their own changes and writing
>>to it I can see, but...?

>I believe what we have here is basically a misunderstanding (or total lack
>of knowledge of source code revision control systems).  I guess it only
>makes sense since so many Linux and *BSD users have never programmed in a
>commercial environment (especially one working to DoD stds).

DoD standards call for CMM Level 3 and up.  That is far far beyond version...
Any programmer at all should be using RCS or CMVC or something.  DoD standards
or not, it is just a good idea.

>I would highly suggest doing a man on `SCCS', `RCS', or `CVS' on your
>favorite Unix box.  If you get no hits, try another.  Or ftp RCS from
>prep.ai.mit.edu:/pub/gnu/rcs* and read the paper written about it.

>A code revision system simply tracks the *CHANGES* to the source code.
>This allows you to get a copy of that file/system foo looked like last
>Tuesday.  This could be useful to see how a file changed, or to go back to
>an older version of a file because you make changes that were bad.  Often
>fixing one problems creates another.  So it useful to see what has changed
>from system X to system Y when something worked in X, but is now broken
>in/for Y.  These are the reasons for using SCCS/RCS/CVS.
>As stated so many times, the *latest* source is always available for *BSD,
>just like for Linux.  EXCEPT, that for *BSD it is one-stop-shopping.  For
>Linux I would have to visit WAY too many ftp sites for my tastes.

Not just that, but on a few occasions I have grabbed a new kernel to fix a few
defects I've been dealing with and the new kernel ends up breaking other 
things.  Version control systems make it much easier to pop back out the files
I need from older versions.  Moreover, it saves tons of space in most cases:
You could probably have then entire RCS tree of Linux for all versions in space
that isn't much bigger that 2x the size of the source now.  I would think that
a lot of people would be interested in that.  I've got CD-ROMs with 20 copies
of source and a lot of them just have minor changes. 


>> System'' that will forever prevent it from truly becoming a Big Success
>> (i.e. bigger than it is now among innumerable hobbyists), and these two
>> things are both tightly interrelated:
>> 
>>     1)  The lack of a central authority for the entire OS, and
>> 
>>     2)  the lack of any single party who is concerned about (and who
>> 	takes personal/corporate responsibility for) the level of
>> 	standard conformance (both ANSI and POSIX) for the entire OS.

Many linux users seem to think that those are pluses.