Re: ChangeLogs and generated files

[Karl Berry]

Hi again Joseph,

    When should a ChangeLog entry mention the regeneration of a generated 
    file, in the form "* configure: Regenerate."?  

My inclination is that it can/should be left up to the individual
package.

In practice, I don't remember ever seeing such entries outside of GCC,
e.g., Emacs doesn't.

I wouldn't want to start mentioning regenerated-and-checked-in files in
my packages, because the reasons it's useful for GCC don't apply to
Texinfo -- there is no dedicated testing the way there is for GCC, and
there are only a couple committers and we don't need reminding about the
few generated files.  So it would just feel like noise in the changelog.
But if other packages find it useful to do so, that's certainly fine.

Anyway, I know you don't want my individual opinion :).  Here are the
choices I see:

1) You or I could draft a little change for standards.texi saying that
it can be useful to mention regenerated files in changelogs, for the
reasons you wrote, and it's fine (but not required) to do so.  I can't
see any reason why rms would reject that.

2) If you want to argue more strongly for it being a GNU-wide practice,
though, we/you can present your arguments to rms.

Wdyt?

Thanks,
Karl

P.S. On the copyright range thing, Donald and I are exchanging email,
no firm conclusion yet ...