[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Circumstances in which ChangeLog format is no longer useful
From: |
John Darrington |
Subject: |
Re: Circumstances in which ChangeLog format is no longer useful |
Date: |
Sat, 29 Jul 2017 19:17:36 +0200 |
User-agent: |
Mutt/1.5.23 (2014-03-12) |
On Sat, Jul 29, 2017 at 06:21:27AM -0400, Alfred M. Szmidt wrote:
People keep saying that they are useless, but so far no example of why
they are useless have been presented or even the slightest argument
was to how one could extract similar information from VCS. neither
diff nor annotate provide this information, they cannot do so in a
generic way and the only way to do it is to write it manually.
You are asking me to prove a negative. This is something which is not
reasonable. Normally, the onus is on the proponents of a case to
provide evidence of its existence (Russel's Teapot).
You have attempted to do this with your example:
* sysdeps/unix/sysv/linux/powerpc/sys/ucontext.h (NFPREG)
[__WORDSIZE != 32]: Rename to __NFPREG and define NFPREG to
__NFPREG if [__USE_MISC].
...
but this ignores the fact that most (all?) VCSs provide "checkin comments",
"commit messages" or some other nomenclature, which provides the possibility
to provide *exactly* this sort of summary.
However like I said earlier, it is not particularly useful to know NFPREG has
been renamed to __NFPREG - that is redundant information. More useful is to
know *why* it has been so renamed.
You said that we already have a recommendation for providing a reason for
changes. But when I read section 6.8 of the GNU coding standards, I can't
find such a recommendation. I only see a recommendation for summarizing
the changes. To quote the opening paragraph:
"Keep a change log to describe all the changes made to program source files."
Well, a history of the changes (such as provided by RCS, CVS, GIT, Aegies, hist
...)
is the most comprehensive description one can possibly have. *If* such a system
is available then ChangeLog is simply a waste of bytes (not to mention developer
effort producing it).
I don't see any recommendation for providing a rationale for changes. But I
have
often looked at ChangeLogs (including my own) and thought "I wish I knew *why*
that
change was made".
J'
signature.asc
Description: Digital signature
- Circumstances in which ChangeLog format is no longer useful, Joseph Myers, 2017/07/28
- Re: Circumstances in which ChangeLog format is no longer useful, Alfred M. Szmidt, 2017/07/28
- Re: Circumstances in which ChangeLog format is no longer useful, Joseph Myers, 2017/07/28
- Re: Circumstances in which ChangeLog format is no longer useful, Alfred M. Szmidt, 2017/07/28
- Re: Circumstances in which ChangeLog format is no longer useful, John Darrington, 2017/07/29
- Re: Circumstances in which ChangeLog format is no longer useful, Alfred M. Szmidt, 2017/07/29
- Re: Circumstances in which ChangeLog format is no longer useful, Paul Smith, 2017/07/29
- Re: Circumstances in which ChangeLog format is no longer useful,
John Darrington <=
- Re: Circumstances in which ChangeLog format is no longer useful, Rical Jasan, 2017/07/29
- Re: Circumstances in which ChangeLog format is no longer useful, Joseph Myers, 2017/07/31
- Re: Circumstances in which ChangeLog format is no longer useful, Joseph Myers, 2017/07/31
- Re: Circumstances in which ChangeLog format is no longer useful, Joseph Myers, 2017/07/31