Re: Should we restore manually maintained ChangeLogs

[Eli Zaretskii <eliz@gnu.org>]

> Cc: emacs-devel@gnu.org
> From: Paul Eggert <eggert@cs.ucla.edu>
> Date: Tue, 8 Mar 2016 08:34:04 -0800
> 
> Eli Zaretskii wrote:
> > You (and some others) say the format and the content in the log
> > messages are important, and I agree.  But if we do care about them,
> > how can we NOT clean them up?
> 
> We can care about them, but agree to fix them up only until they become part of 
> history. Once they're history we don't worry about trying to change history; 
> they're just old mistakes that are part of the log but are not otherwise part of 
> the current Emacs. If managed well, this can help motivate contributors to write 
> good commit messages the first time. This approach is not perfect, but it works 
> reasonably well in other projects and it is way easier to explain and to 
> maintain than what we're doing now, or what we did a year ago.

AFAIK, those other projects get commits from veteran GNU developers
such as yourself, which have the ChangeLog format and good log
messages burnt into their finger memories.

Emacs is different.  We have to educate the newcomers to become such
veterans.  And education simply doesn't work without the need to fix
your beginner's mistakes.  By giving up on the need to fix them, we
inadvertently send a very loud and clear message to the newcomers
saying that good, correct, and accurate log messages are not
important.  As result, they will never acquire the skills that you and
other veterans here have since long ago.

When I was a newcomer, I had the privilege of getting comments and
requests to fix my log messages from Richard and others.  Had they not
insist on making my errors clear to me, had they not asked for me to
fix them, I wouldn't be able to write the kind of log messages and
other short descriptions I can do today.

So let's stop thinking about ourselves -- we don't need these fixes
anyway.  Let's think about the new generation, the ones we must
educate as long as we are around.  They _need_ us to point them to
their flops, and they need to learn to fix them.  This cannot be
learned in theory, only by doing.  It is an important part of their
apprenticeship.  This isn't about perfectionism, this is about
teaching the newcomers to be better Emacs developers.

> So, for example, we should strive to get the "tiny change" stuff right the first 
> time in commit messages; but if we make mistakes in that area it OK -- the sky 
> will not fall down, and software archaeologists of the future will still be able 
> to figure things out well enough.

The skies will not fall down if Emacs ceases to exist, either, or
becomes a much less clean and orderly project than it is now.  But
we'd like to avoid that, one hopes.