Established Sites

I recently migrated Established Sites from an internally hosted subversion repository to a git repository hosted by github.com.

I'm now considering the idea of converting Established Sites from a Nexista (PHP) application to a Sinatra application (Ruby).

It actually wouldn't be too difficult as so much of the logic is in XML and the logic is in XSL. I've even made a couple of tiny commits in this direction.

I recently forked the yui-app-theme repository at github for use with Regdel. It has worked out splendidly, and since I'm committed to jQuery, I'm planning a jquery hybrid. :-)

Established Sites might just be the first app to use it! If so, the first task will be to combine droppy with the nice tabs that yui-app-theme supports.

See also: jQuery app theme

It turns out Git is a pretty terrific storage and revision system and is getting embraced by web developers as a storage system for websites.

Therefore, I'm considering switching from an RDBMS to Git for the storage of basic content. As far as the organization, metadata, and page relationships, an RDBMS will still be useful, no doubt.

There are a few projects using Git as a storage engine:

• Ikiwiki - my fork
• Djanjo Git CMS
• Git Wiki my fork

Established Sites is moving its code repository from an internally managed subversion repository to Github.com:

• http://github.com/docunext/established-sites

Enjoy!

The established-sites.com website is now "Powered by Movable Type". I held off on doing this for awhile, as I wanted it to be "Powered by Established Sites", but since I've recently decided to build off the MTOS data model, I figured it was fair middle ground.

So what does that mean - build off the MTOS data model? Just how it reads - Established Sites will use the same data model as MTOS.

Why? The MTOS data model, while imperfect, has superb support for multiple sites. Therefore, it *should* be reasonable to publish some with MTOS, and other with whatever else has access to the database.

I currently use XSL to render CSS, but its not a very elegant process. I have been looking at CSS parsers and CSS libraries and I'm now feeling like it will be easier to use something else.I'm not exactly sure what my final choice will be, but I'm going to try out the PHP PEAR HTML_CSS library. I've briefly tested it and it seems like it may do the trick. See my first attempt at a CSS filter to get an idea of what I'm trying to accomplish.