<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Arne Babenhauserheide wrote:
<blockquote cite="mid:200901062342.17190.arne_bab@web.de" type="cite">
<pre wrap="">Am Montag 05 Januar 2009 20:46:08 schrieb Neal Becker:
</pre>
<blockquote type="cite">
<pre wrap=""><a class="moz-txt-link-freetext" href="http://blogs.gnome.org/newren/2009/01/03/gnome-dvcs-survey-results/">http://blogs.gnome.org/newren/2009/01/03/gnome-dvcs-survey-results/</a>
</pre>
</blockquote>
<pre wrap=""><!---->
This somehow bothers me, since it shows that some people use momentum as main
argument for switching to git. </pre>
</blockquote>
<br>
I think that momentum _is_ the main argument for switching to git from
a centralized VCS. You will find a git bias in the userbase for git
(just like you would find an hg bias in hg users and a bzr bias in bzr
users). <br>
<br>
<br>
The main advantage of git is that there are a lot of people who are
familiar with it as you get closer to projects that potentially share
developers with the kernel. Outside of this group, git support tends to
fall away because of many disadvantages:<br>
* non-uniformity in command arguments<br>
* too many core commands<br>
* relative lack of windows support<br>
<br>
(though someday git will probably solve both the first and last problem)<br>
I think overall that git suffers from a lack of direction. They got it
good enough to use for the kernel and then they began to suffer from
code sprawl problems. <br>
<br>
I don't think it is difficult to promote Mercurial at all compared to
git, so long as you can get away from the git bias. I think it is much
more difficult to promote hg over bzr. Bazaar has:<br>
* that dumb page full of misconceptions and inconsistancies that seems
to rally people who don't entirely understand what the differences
between them really are (with things like "bzr has built-in merge
support while hg doesn't; how can you have a dvcs without merge built
in?").<br>
* bound branches (which really is a killer feature for commercial users
used to any centralized vcs; you really don't need anything more than
your basic svn commandset to get to use bzr in this mode)<br>
* commercial backing<br>
* very minor merge improvements<br>
* a more efficient repo (space wise) for small numbers of changesets
(making for good demonstrations)<br>
<br>
Whereas Mercurial has:<br>
* still faster with nearly every command (though all 3 are fast enough
that this doesn't matter much)<br>
* more efficient repo for large numbers of changesets<br>
* a more extensible plugin architecture (with both hooks and extensions)<br>
* an easy to use quilt-like system<br>
* a more active community<br>
<br>
It seems to me that when it comes to the corporate world, bzr handily
wins out, but in the open source world the two systems are on nearly
equal footing (and due to the git bias in the linux environment both
lose to git).<br>
</body>
</html>