emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Apologia for bzr


From: Eli Zaretskii
Subject: Re: Apologia for bzr
Date: Fri, 03 Jan 2014 12:08:51 +0200

> Date: Fri, 3 Jan 2014 16:21:26 +0700
> From: Yuri Khan <address@hidden>
> Cc: Toby Cubitt <address@hidden>, 
>       Eric Raymond <address@hidden>, address@hidden, 
>       Emacs developers <address@hidden>
> 
> On Fri, Jan 3, 2014 at 3:57 PM, Eli Zaretskii <address@hidden> wrote:
> >
> > If you don't already know what is "staging", you will never understand
> > that this is one of the most important and useful options.  Also,
> > "haven't told Git about new files" fails to mention "git add".  Once
> > I've managed to grasp all that, I've made an alias for "commit -a",
> > because that's what I almost always want.  (And why isn't that the
> > default, dammit?)
> 
> Because staging is a key concept in git and it enables a whole lot of
> useful workflows.

I know that.  Don't take my questions as indications that I'm
unfamiliar with basic git terminology (I was once, but no more).

> E.g. you can work all day and half the next day on a
> feature, making small formatting changes and fix coding style
> violations on your way as you spot them, then fire up a commit tool
> and make three commits, one for trivial formatting changes, another
> for coding style fixes, and a third one with the feature you actually
> worked on.

I was talking about the defaults.  For the defaults, the important
question is what would J. R. Hacker want in most of his commits.  It
seems to me that occasional contributors to projects (as opposed to
their core developers) will seldom if ever get to the complicated
workflows you describe.  It's definitely true for me in a couple of
projects in which I'm involved (not Emacs, obviously).

> Without staging, you would have to look at the diff, back up and
> revert some changes so that the working directory looks the way you
> want for one commit, then the other, then the next one. Or you would
> hold off fixing small things until you have committed the feature, and
> risk forgetting to do them.

Well, let's begin by agreeing that what you described is just bad
habit: you should commit very frequently, and so seldom if ever get to
the point where you have hacked for a day and a half without a single
commit.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]