Shipping hg-git by default?

[Erik Huelsmann]

On Fri, Sep 25, 2015 at 4:35 PM, Augie Fackler <lists at durin42.com> wrote:

> On Fri, Sep 25, 2015 at 10:28 AM, anatoly techtonik <techtonik at gmail.com>
> wrote:
> > Hi,
> >
> > Looks like https://bitbucket.org/ ditched Mercurial:
> >
> >     Bitbucket is the Git solution for professional teams
> >
> > So for me personally there is no incentive to use HG
> > anymore (even though I really miss 'hg inc`), especially
> > when I setup new virtual machine on Cloud9 or similar
> > platform to quickly patch and test one of my projects.
> >
> > But.. if installation of Mercurial already included hg-git
> > support, I could still use it to work with GitHub right of
> > the box without an additional installation hassle, which
> > is, of course, putting me away from that at the moment.
> >
> > What do you say?
>
> hg-git needs a ton of work to be viable for inclusion like this. It'd
> need things like better handling of history rewriting (incl. some
> modicum of phases support, as well as coping with rewrites on the git
> side more gracefully), and probably also a fair amount of UI spit and
> polish.
>

I think you're unnecessarily harsh on yourselves here. I use hg-git every
day, have done so for more than a year, and it works great! The only issue
that I have with it has to do with a very specific situation where HG
commit hashes are incorrectly (not) translated to Git hashes. Other than
that, no problems.

You might argue that I ought to switch to Git. I don't want to though: from
a user's perspective, Hg feels much more polished and tuned to *my*
processes than Git. Git feels like I'm handed the Lego bricks and I need to
make my own composed commands and processes out of it. Hg just offers me a
complete and comprehensible workflow. So, as long as I don't *have* to
switch to Git, I'm really happy to keep using Mercurial, with a mix of
repositories, both Hg and Git. (Unfortunately, with more of he latter than
the former.)

Even then, I'm not sure it makes complete sense, but it's pretty
> academic unless someone wants to invest that kind of time into hg-git
> (I don't personally have the time or motivation.)
>

>From the reactions on this thread, it looks like it makes complete sense to
a number of people :-)

What does the above remark about time and motivation mean in the longer
term though? I mean, do you still have time to update your repository if
you should get pull requests that are up to your standards? If not, would
there be ways to organize future development nonetheless? You seem to imply
that one person should be picking up the development, however, if there are
small enough actionable items, maybe many people can make a difference
together?

Thanks for your work on the module so far!

Regards,

-- 
Bye,

Erik.

http://efficito.com -- Hosted accounting and ERP.
Robust and Flexible. No vendor lock-in.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mercurial-scm.org/pipermail/mercurial/attachments/20150926/d2c02986/attachment-0002.html>