On 01/03/2012 11:09 PM, Bruno Haible wrote: > Stefano Lattarini wrote: >> report in the commit message of your patch the useful and extensive >> arguments, explanations and references you have laid out so excellently in >> your mail. > > Is it better to copy the arguments into a (long) commit message, > or add a reference to the mail(s) in the lists.gnu.org archive? > Ah, but these two courses of action are not mutually exclusive, are they? ;-)
> I tend to do the latter, because it allows future readers to recover > the entire context of the discussion thread, including arguments and > counter-arguments. But I'm open to either style. > You could report a "condensed" (but still complete) rationale in the git commit message, and at the same time keep reference to the original thread, so that interested readers can see how that rationale developed in practice. > Note: It may take an hour or two for a mail to appear in the lists.gnu.org > archives, but by the time a committer pushes a commit, he should have > waited a bit anyway. So I don't see this as a big problem. > I agree with you here. Regards, Stefano
