<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	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/"
		>
<channel>
	<title>Comments on: Finally putting my personal configuration files under version control</title>
	<atom:link href="http://mnemonikk.org/2009/03/23/finally-putting-my-personal-configuration-files-under-version-control/feed/" rel="self" type="application/rss+xml" />
	<link>http://mnemonikk.org/2009/03/23/finally-putting-my-personal-configuration-files-under-version-control/</link>
	<description>My life with the machine</description>
	<lastBuildDate>Tue, 20 Jul 2010 22:45:35 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.0</generator>
	<item>
		<title>By: pille</title>
		<link>http://mnemonikk.org/2009/03/23/finally-putting-my-personal-configuration-files-under-version-control/comment-page-1/#comment-3489</link>
		<dc:creator>pille</dc:creator>
		<pubDate>Tue, 20 Jul 2010 22:45:35 +0000</pubDate>
		<guid isPermaLink="false">http://mnemonikk.org/?p=22#comment-3489</guid>
		<description>managing multiple repositories is the only scenario for such a script. 

i&#039;ve rewritten this tool to support this feature. 
it can be found along with an explanation at http://struction.de/projects/dotVC

until my need for that feature using the suggested .htignore hack worked fine.</description>
		<content:encoded><![CDATA[<p>managing multiple repositories is the only scenario for such a script. </p>
<p>i&#8217;ve rewritten this tool to support this feature.<br />
it can be found along with an explanation at <a href="http://struction.de/projects/dotVC" rel="nofollow">http://struction.de/projects/dotVC</a></p>
<p>until my need for that feature using the suggested .htignore hack worked fine.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: admin</title>
		<link>http://mnemonikk.org/2009/03/23/finally-putting-my-personal-configuration-files-under-version-control/comment-page-1/#comment-222</link>
		<dc:creator>admin</dc:creator>
		<pubDate>Mon, 23 Mar 2009 23:42:31 +0000</pubDate>
		<guid isPermaLink="false">http://mnemonikk.org/?p=22#comment-222</guid>
		<description>A valid argument for the symlinking-approach: you can easily opt-out if you want to use a differing config file on some host for some odd reason, so you don&#039;t have to worry about unintentionally updating some file if you don&#039;t want to.</description>
		<content:encoded><![CDATA[<p>A valid argument for the symlinking-approach: you can easily opt-out if you want to use a differing config file on some host for some odd reason, so you don&#8217;t have to worry about unintentionally updating some file if you don&#8217;t want to.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: admin</title>
		<link>http://mnemonikk.org/2009/03/23/finally-putting-my-personal-configuration-files-under-version-control/comment-page-1/#comment-220</link>
		<dc:creator>admin</dc:creator>
		<pubDate>Mon, 23 Mar 2009 23:33:00 +0000</pubDate>
		<guid isPermaLink="false">http://mnemonikk.org/?p=22#comment-220</guid>
		<description>Good point, I haven&#039;t thought of that. But the symlinking-approach still has some potential advantages: it&#039;s immediately apparent what files are under VC; I can have several repositories, e.g. a repository that I want to make public, and one for private data. Fortunately, emacs sees that the symlinks lead to version controlled files, so this is only mildly inconvenient. But still, good point.</description>
		<content:encoded><![CDATA[<p>Good point, I haven&#8217;t thought of that. But the symlinking-approach still has some potential advantages: it&#8217;s immediately apparent what files are under VC; I can have several repositories, e.g. a repository that I want to make public, and one for private data. Fortunately, emacs sees that the symlinks lead to version controlled files, so this is only mildly inconvenient. But still, good point.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Giorgos Keramidas</title>
		<link>http://mnemonikk.org/2009/03/23/finally-putting-my-personal-configuration-files-under-version-control/comment-page-1/#comment-219</link>
		<dc:creator>Giorgos Keramidas</dc:creator>
		<pubDate>Mon, 23 Mar 2009 23:21:14 +0000</pubDate>
		<guid isPermaLink="false">http://mnemonikk.org/?p=22#comment-219</guid>
		<description>Ignore files help a lot when you want to avoid crawling huge and largely irrelevant subdirectories of $HOME.

My first commit at $HOME/.hg/ about a year ago was:

$ hg log -r0
changeset:   0:51a27c91da6a
user:        Giorgos Keramidas 
date:        Sun Feb 10 22:34:41 2008 +0200
summary:     Initial revision: ignore everything by default in $HOME dirs.

This was just an .hgignore file that ignores &#039;**&#039;, the pattern that matches any file in any subdirectory.

Then I &quot;hg add&quot;-ed only the files that I want to track.  A short script that prints pairs of {rev, number of files} pairs shows that I have been slowly increasing the number of tracked files since last year.  I started with only one file (the .hgignore file) and have reached 400 files now:

$ hg log --template &#039;{rev}\n&#039; -r 0:tip &#124; \
  while read rev ; do \
    echo &quot;${rev}&quot; $(hg manifest &quot;${rev}&quot; &#124; \
                    wc -l &#124; awk &#039;{print $1}&#039;) ; \
  done &#124; \
  awk &#039;BEGIN {ofiles=0;}
       { if ($2 != ofiles) {
             printf &quot; {%d %d},&quot;, $1, $2;
             ofiles=$2;
             fflush();
          }
        }&#039;
=&gt; {0 1}, {1 2}, {2 8}, ... {497 399}, {500 400},

Running &quot;hg stat&quot; is almost instantaneous, even though my entire $HOME is quite a few GB now:

$ time hg stat
M .newsrc
M .newsrc.eld

real    0m0.796s
user    0m0.407s
sys     0m0.381s
$

This is for a different version control system, but my intuition says not traversing ignored subdirs should be &#039;fixable&#039; in other VCSs too.</description>
		<content:encoded><![CDATA[<p>Ignore files help a lot when you want to avoid crawling huge and largely irrelevant subdirectories of $HOME.</p>
<p>My first commit at $HOME/.hg/ about a year ago was:</p>
<p>$ hg log -r0<br />
changeset:   0:51a27c91da6a<br />
user:        Giorgos Keramidas<br />
date:        Sun Feb 10 22:34:41 2008 +0200<br />
summary:     Initial revision: ignore everything by default in $HOME dirs.</p>
<p>This was just an .hgignore file that ignores &#8216;**&#8217;, the pattern that matches any file in any subdirectory.</p>
<p>Then I &#8220;hg add&#8221;-ed only the files that I want to track.  A short script that prints pairs of {rev, number of files} pairs shows that I have been slowly increasing the number of tracked files since last year.  I started with only one file (the .hgignore file) and have reached 400 files now:</p>
<p>$ hg log &#8211;template &#8216;{rev}\n&#8217; -r 0:tip | \<br />
  while read rev ; do \<br />
    echo &#8220;${rev}&#8221; $(hg manifest &#8220;${rev}&#8221; | \<br />
                    wc -l | awk &#8216;{print $1}&#8217;) ; \<br />
  done | \<br />
  awk &#8216;BEGIN {ofiles=0;}<br />
       { if ($2 != ofiles) {<br />
             printf &#8221; {%d %d},&#8221;, $1, $2;<br />
             ofiles=$2;<br />
             fflush();<br />
          }<br />
        }&#8217;<br />
=&gt; {0 1}, {1 2}, {2 8}, &#8230; {497 399}, {500 400},</p>
<p>Running &#8220;hg stat&#8221; is almost instantaneous, even though my entire $HOME is quite a few GB now:</p>
<p>$ time hg stat<br />
M .newsrc<br />
M .newsrc.eld</p>
<p>real    0m0.796s<br />
user    0m0.407s<br />
sys     0m0.381s<br />
$</p>
<p>This is for a different version control system, but my intuition says not traversing ignored subdirs should be &#8216;fixable&#8217; in other VCSs too.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
