incubator-site/build/site/. Forrestbot has proved very handy for xml-site, which is now 155M and consequently takes ages to update over a modem connection. It also saves developers from having to deal with Forrest. As long as the docs validate (using any tool), they can be checked in and published through the webapp. Separation of content from presentation and all that.
Given the sites are getting so big (xml-site > 150M is a bit scary), is a cvs sync still the right way to do it? Storing binary files (PDF) in the cvs repository is going to get very big if any files get frequently changed (e.g. news in xml.apache.org), and it doesn't really seem necessary if we can regenerate any version of the pdf from the xml source of the same version.
Should we be looing at modifying the forrest portions of the site so that forrestbot pushes them into the web server in another way and thus take the generated docs out of CVS?
Or could we run a script on the web server that checks if the source has changed and tries to rebuild the site if it determines it has? (Effectively run forrestbot on the site - don't know what infrastructure@ would feel about that.)
Cheers,
Berin
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
