<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:wfw="http://wellformedweb.org/CommentAPI/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
	xmlns:slash="http://purl.org/rss/1.0/modules/slash/"
	>

<channel>
	<title>Jeremy&#039;s Thoughts &#187; git</title>
	<atom:link href="http://velohacker.com/tag/git/feed/" rel="self" type="application/rss+xml" />
	<link>http://velohacker.com</link>
	<description>Ramblings of a Cyclist Hacker</description>
	<lastBuildDate>Thu, 19 Apr 2012 15:50:17 +0000</lastBuildDate>
	<language>en</language>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.3.1</generator>
		<item>
		<title>Publishing git trees</title>
		<link>http://velohacker.com/2009/02/23/publishing-git-trees/</link>
		<comments>http://velohacker.com/2009/02/23/publishing-git-trees/#comments</comments>
		<pubDate>Mon, 23 Feb 2009 18:30:43 +0000</pubDate>
		<dc:creator>Jeremy</dc:creator>
				<category><![CDATA[Fedora]]></category>
		<category><![CDATA[git]]></category>

		<guid isPermaLink="false">http://velohacker.com/?p=3768</guid>
		<description><![CDATA[Scott@Ubuntu has posted a few times over the past week or so about some of the problems he&#8217;s hit while using git and the more I think about them and read them, I think that a lot of it comes &#8230; <a href="http://velohacker.com/2009/02/23/publishing-git-trees/">Continue reading <span class="meta-nav">&#8594;</span></a>]]></description>
			<content:encoded><![CDATA[<p><a href="http://www.netsplit.com">Scott@Ubuntu</a> has posted a few times over the past week or so about some of the problems he&#8217;s hit while using git and the more I think about them and read them, I think that a lot of it comes down to expectations of a very different mode of working than the git developers <i>&quot;push&quot;</i>.</p>
<p>The git way seems to largely be that the most common way for someone new to contribute code is that they write a patch (or series of patches), commit them locally, and then send them via email for review and eventual merging.  These are made easier through the existence of commands like <i>git format-patch</i>, <i>git send-email</i> and <i>git am</i> (apply-mailbox).  These really are wonderful tools if your workflow is around using email for patch review and merging as is done in the kernel and many other places.  There&#8217;s certainly an argument to be made that this sort of mailing list review ends up improving code quality.  But it&#8217;s not the only way.  And since git is about there being multiple ways to do things, maybe some of the other ways need to be made easier too&#8230;</p>
<p>One of the ones that seems to pop up commonly is that people want to have their trees available for others to pull from/merge in.  This is doable today, but as Scott notes, it&#8217;s not entirely straight-forward and requires some knowledge of how to set things up.  What if, instead, there was a <i>git publish</i> command that was able to essentially push your working branch to a remote location (via ssh/scp) and do the things necessary so that it&#8217;s clonable via http out of the box and not require any initial login and creating a repo, etc.  The question that comes to mind is &#8220;if I run git publish again, how does it differ from a git push&#8221; and the answer is probably not much, but I&#8217;m open to other opinions</p>
<p>The obvious second step after that would be to add easy-to-use support for publishing to somewhere like <a href="http://gitorious.com">gitorious</a>.  That would help for people who don&#8217;t have web hosting or a number of other things.  And that is a gap that bzr ends up filling by allowing people to use launchpad for arbitrary bzr hosting.  </p>
<p>The other workflow that might be interesting to better support would be something like <a href="http://www.review-board.org">Review Board</a>.  I keep wanting to set up an instance to play with and possibly even have people use for things like livecd-tools or anaconda patches.  It seems to combine some of the upsides of mailing list review (easy for all to see, easy to annotate changes, &#8230;) without some of the downsides (mailing lists suck, a topic for another day)  There&#8217;s some movement underway to try to get an install going for Fedora Infrastructure, so if you&#8217;re interested, <a href="https://fedorahosted.org/fedora-infrastructure/ticket/1196">help is wanted!</a>.  </p>
<p>What do people think?  Would a git publish command as described above (minus the aside of the Review Board stuff) be useful/interesting?</p>
]]></content:encoded>
			<wfw:commentRss>http://velohacker.com/2009/02/23/publishing-git-trees/feed/</wfw:commentRss>
		<slash:comments>3</slash:comments>
		</item>
	</channel>
</rss>

