<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
You should definately be using CVS. You can keep track of your changes
that way, to undo or redo them as needed, and also keep all developers
in sync. (CVS allows you to both edit the same file at the same time
and will incorporate both sets of changes. It can also tell you who
did what.)
<p>For a straightforward site, you can set up your web server to use the
CVS client and use your CVS tree as your doc tree (using update or export).
You can arrange so that only tested changes (you define 'tested') show
up on the web site. You can also use a separate build process to
manipulate all the files in CVS into the structure your web server wants
(this can be arbitrarily complex - I use it for code compilation and database
data model changes). All depends how much control you want over what
shows up on the 'production' system. My advice is to start simple
and grow as you need to. But the first step is to use CVS.
<p>Here are pointers to some docs:
<br> <a href="http://cvsbook.red-bean.com/">Open Source Development
with CVS</a> by Karl Fogel (the CVS chapters are free online; the OSD chapters
you have to buy; read the free stuff before you buy)
<p> <a href="http://www.cvshome.org/">CVS Home</a> (docs, code, the
whole shebang)
<p>Dave
<p>Faster-Laster wrote:
<blockquote TYPE=CITE>On Fri, Jun 01, 2001 at 01:07:19PM -0600, Kevin Cullis
wrote:
<p>> Second, would CVS be useful here?
<br> Yes, I know of someone local
to here (Nashville) that has set up his
<br>web server as a cvs tree and pages are checkout/in and worked on. cvs
<br>works great because these files are almost only text (html/php). I
<br>believe he had a development server that rsynced the stable tree of
this
<br>cvs to his production server. I seem to remember that he even had a
<br>windows cvs client for some of the developers that worked on win boxes.
<br>He spoke quite highly of the system.
<br>--</blockquote>
</html>