home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.linux
- Path: sparky!uunet!stanford.edu!snorkelwacker.mit.edu!bloom-picayune.mit.edu!root
- From: tytso@ATHENA.MIT.EDU (Theodore Ts'o)
- Subject: Re: header woes with 2.2.2d
- Message-ID: <1992Aug19.201415.12339@athena.mit.edu>
- Sender: root@athena.mit.edu (System PRIVILEGED Account)
- Reply-To: tytso@ATHENA.MIT.EDU (Theodore Ts'o)
- Organization: The Internet
- Date: Wed, 19 Aug 1992 20:14:15 GMT
- Lines: 53
-
- From: davidsen@ariel.crd.GE.COM (william E Davidsen)
- Date: 19 Aug 92 12:46:49 GMT
- Reply-To: davidsen@crd.ge.com (bill davidsen)
-
- Having done a good bit of distributed development over the years has
- convinced me that this is one place where a little good software
- engineering practice is important. We are talking about three people (or
- groups) promptly sharing changes to the header file. People who are in
- fast communication via network.
-
- Ah, the difference is that's what I think is happening right now. Keep
- in mind that most of the incompatibilities happen because of new features
- in the kernel.
-
- So when Linus releases changes to the kernel, he's doing exactly what
- you ask. He generally does so quite promptly (especially if you compare
- what he does against the release cycles of companies like Microsoft),
- and he does so publically. Then, everyone who does work on these
- changes (like the maintainers of ps, GCC, and possibly others that we
- don't know about) can update their programs so that they work again.
-
- The problem is that everybody *else* takes the new patches right away,
- and complains when things break. I don't know.... when *I* took
- patchlevel 1 to 0.97, I took a couple of minutes glancing through the
- changes, and then I was able to quickly hack ps to make it work. It
- didn't slow me down much. But if it is going to slow you down; if
- you're only interested in doing applications development on Linux (and
- that *is* an honorable thing to do), I would suggest that you wait for a
- few weeks until everything settles down.
-
- Keep in mind that when the next set of patches to Linux comes out, Linus
- has already said that the virtual memory code has been rewritten, and ps
- is going to break in major ways. I'd rather not have him hold back his
- release while other people scurry around and fix ps, and the other
- programs.
-
- I would rather the kernel changes get released, and people who don't
- mind living on the bleeding edge, or who don't mind living without "ps"
- can play with it right away, and people who want all of the comforts of
- home SIMPLY NOT TAKE THE KERNEL CHANGES RIGHT AWAY. Is that such a hard
- thing?
-
- If people don't have enough self control to keep their paws off of the
- latest stuff (or who could take the latest stuff and then not bitch when
- things broke), one alternative is to have a fascist alpha-testing
- program, as you suggested --- but when you have to do this in the
- freeware domain, what you are doing is forcibly excluding people from
- taking the "testing release" when they should know better. Perhaps it
- is a stupid hope, but I was hoping that people would be adult enough and
- intelligent enough and have enough self-control so that sort of thing
- wouldn't be necessary.
-
- - Ted
-