[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Emacs-diffs] trunk r117340: * lisp/calculator.el: Lots of revisions
From: |
Eli Barzilay |
Subject: |
Re: [Emacs-diffs] trunk r117340: * lisp/calculator.el: Lots of revisions |
Date: |
Wed, 18 Jun 2014 06:08:37 -0400 |
Yesterday, Stefan Monnier wrote:
> > * lisp/calculator.el: Lots of revisions
> > - Kill the calculator buffer after electric mode too.
> > - Make decimal mode have "," groups, so it's more fitting for
> > use in money calculations.
> > - Factorial works with non-integer inputs.
> > - Swallow less errors.
> > - Lots of other improvements, but no changes to custom
> > variables, or other user visible changes (except the above).
>
> The above description is the kind of thing that should go into
> etc/NEWS.
Shouldn't the NEWS entry list only user-visible changes? (That is,
only the first three items in the above.) I also have a small bug fix
to commit, which also seems like something that shouldn't be mentioned
in the NEWS --?
> But you forgot to include a ChangeLog entry for that commit.
Should I add it now? (I have a small bug fix in the new code, so
maybe I can add the previous text with this commit?)
> For the ChangeLog entry, you should instead include details about
> which parts of the code you changed (see
> https://www.gnu.org/prep/standards/html_node/Change-Logs.html). You
> can use "C-x 4 a" (both from the calculator.el and the *vc-diff*
> buffers) to help you write that.
(AFAICT, it doesn't help with the actual log message, right?)
> And the commit message should be a textual copy of the ChangeLog
> entry just with the leading TABs stripped (you can use C-c C-a from
> the *vc-log* buffer to let Emacs fetch that ChangeLog text for you).
(I should obviously follow whatever is the convention, but as a
sidenote, this requirement is implicitly saying "I'm obsolete", since
if it's always followed, then it is identical to the commit log, and
therefore the changelog file can be auto-generated. Similarly for the
log message contents -- which is specified rigidly to the point where
it can be done better with a VCS. But I'll repeat to avoid flaming
followups: I will try to follow whatever the convention are.)
--
((lambda (x) (x x)) (lambda (x) (x x))) Eli Barzilay:
http://barzilay.org/ Maze is Life!