"Oliver Elphick" <[email protected]> writes: > <DDP_ROOT> > | > --------------------------------- > | | > webpages manuals > | | > __________________________ ___________________________________ > | | | | | | | | | > index.html [structure as now] [directory for each manual] general > | | > ___________________________ __________ > | | | | | | | | > manual_name.sgml [chapter_name.sgml...] [sundry items]
This looks good. Try to keep it pretty flat, the structure that is. Marcus has a pretty good Debian Docuemntation Heirarchy which you can probably use if you like for organization (i.e., replacing the current major categories in http://fatman.mathematik.tu-muenchen.de/~schwarz/debian-doc/manuals.html). Ask me and I can shoot you a copy, but Marcus says it's not ready yet and will be ready later this week or so. > Generated web pages will not be checked in with CVS but will exist in > a separate tree, with links to it from the CVS-managed pages. The managed > pages will be copied or linked to the web server This is very unclear to me. I suspect you may be intending to use CVS the wrong way. Yes yes and yes! the web pages should be in cvs, that's great. But why the links? I would just use 'cvs co' then 'cvs update -d' nightly (i.e., from cron), and maybe hide the CVS subdirs from Apache serving them up. But are you intending to serve up the manuals from CVS? Many manuals are already in cvs elsewhere, and/or part of larger packages, such as emacsen-common, doc-base, etc. Can you have the directory for each manual just be a checked out version of the manual from CVS if possible? Even this might be tricky because it might not be trivial to check out only the manual and nothing else. So please tell us more about the technical details of how you intend to accomplish this. -- .....A. P. [EMAIL PROTECTED]<URL:http://www.onShore.com/> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
