Since some of my SVN Repositories did not survive the last Update of the Database, and were really hard to recover, or some even needed to be recovered from older Backup, I decided to switch to Mercurial for all me Repository needs.
Looking around on the Web there a probably 100s of guides explaining the basics and setup, but actually None of them worked as a whole for me, so I decided to write down my experiences maybe they apply to somebody, and if not, at least it will be easier for me next time.
First of all thanks to the following Guides I got everything working:
Second I’m running on a Debian Lenny Server with all Updates until 5/5/2009, Mercurial is 1.0.1, Apache is Apache/2.2.9 (Debian).
Now a couple hints, make sure cgi is enabled
sudo a2enmod cgi
sudo a2enmod cgid
Now Setup a new Location for Mercurial like this one
ScriptAliasMatch ^/merc(.*) /MYREPOPATH/hgwebdir.cgi$1
<Location /merc>
Allow from all
Options ExecCGI
AuthType Digest
AuthName “REPOSNAME”
AuthDigestProvider file
AuthUserFile PATHTOPASSWORDFILE
Require valid-user
</Location>
Now copy the file hgwebdir.cgi found in /usr/share/hg to the Repository Path and Setup a hgweb.config file in the same location to point to your repositories
[collections]
MYREPOPATH = MYREPOPATH
Now just the repositories need to be initilized in this Path
mkdir testrepo
cd testrepo && hg init
and the Password file has to be created
mkdir testrepo
htdigest -c PASSWORDFILE ‘REPOSNAME’ USERNAME
If you now add allow_push = USERNAME to the hgrc file for the repository you can push your changes to the newly initilized repository after cloning it.
Hope those hints will help, if not comments are welcome.
Code from the sideline 