[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Using cvs for websites:
From: |
Maarten de Boer |
Subject: |
Re: Using cvs for websites: |
Date: |
Wed, 23 May 2001 11:02:01 +0200 |
> I imported the document root of the website.
> Then removed that copy and checked out a fresh copy from the repository like
> I usually do.
> I then checked out a copy of the site to my home location. When I do a
> 'cvs commit' at home I have to do a 'cvs update' in the document root to get
> the changes merged across. This isn't too much trouble (It would be nice to
> get rid of the two step process if anyone has done this type of setup).
I have a very simple perl script "cvsview" that I use as a webserver to access
webpages stored in CVS. This script is in /usr/bin/cvsview and made executable.
Note that this has been a quick hack, but it works fine for me. It doesn't do
CGI's though, but it does use /etc/mimetypes.
I attach the script. In order to use it, you'll have to do the following:
Create a user cvsview. The script is executed through inetd, with this user:
>From /etc/inetd.conf:
cvsview stream tcp nowait cvsview /usr/sbin/tcpd /usr/bin/cvsview
>From /etc/services:
cvsview XXXX/tcp # cvsview
where XXXX is the port you want to use. You can perfectly use 80, but
make sure you don't have another webserver running on that port already.
I have both apache and cvsview running at the same time, so I use a
different port.
The users homedir /home/cvsview is were the working copies are stored.
It has the following permissions:
drwxr-s--- 6 cvsview cvsview 4096 Mar 7 17:33 /home/cvsview/
I can control which CVS directories can be accessed by the user cvsview
with standard permissions.
> But my major question is. How can I update my document root now without
> having 'CVS' directories everywhere. Can I centralize the location of all
> the CVS database information?
Why do the CVS directories bother you?
Maarten
cvsview
Description: Binary data