Re: In support of Jonas Bernoulli's Magit

[Marcin Borkowski]

On 2017-07-08, at 19:42, raman <address@hidden> wrote:

> For simple vc tasks -- and that typically equates to many things one
> traditionally did in all VC systems, vc-git is what I use --- examples:

FWIW, I sometimes use VC with Mecurial, and the experience is so bad
that the _sole reason_ I switched to Git was the existence of Magit.

The main annoyance is that if I edit _more than one file_, and press C-x
v v, instead of committing my all changes (or at least asking me about
it), it only commits the changes in the currently edited file.  Several
times I made an incomplete commit because of that.  I tried to find an
option to override this, but to no avail.  _That alone_ renders VC
totally unusable, annoying, broken and actively harmful for me.

Now that I think of it, I could have written a bug report (or even try
to implement this myself).  But you know what?  I just switched to Git
and started to use Magit, because it is so much better.

Now, vc-dir seems similar to magit-status.  But it has one _huge_
drawback: it requires me to press RET each time after I invoke it.  (Why
would I want to invoke vc-dir for some other directory than the one I'm
in anyway?  And in such rare cases, C-u would handle that much better.)

Now that I think of it, I could write a trivial wrapper for that (and
I might actually do it, for the sake of legacy Mercurial projects
I have).  But you know what?  If you use Git, then why even bother when
Magit exists?

I guess that Emacs' VC with distributed VCSs is fundamentally broken,
since it was really designed with RCS and its likes in mind, and
imposing that onto DVCSs is awkward.  I suspect that situation (with VC)
is better now than a few years ago (e.g., AFAIK pushing/pulling is now
implemented), but I haven't looked into it too much.

Sorry if that sounds harsh, but I consider VC's utility (with DVCSs) to
be zero _at best_.

Best,

--
Marcin Borkowski