On 6/1/06, Jason LaPier <[EMAIL PROTECTED]> wrote:
While we're on the subject of repositories - I was thinking about this last
night:
Does anyone know of a recommended path for using subversion to keep track of
actual documents (ie non-code)?

svn can be used as a backing store, start reading here
http://svnbook.red-bean.com/en/1.2/svn.developer.html#svn.developer.layerlib.ra

but look around to see if someone else is doing this already

I've heard of some people using XML-based
documents for this sort of thing, where they can check-in and check-out say
(for example) chapters in a book. The you have a revision history for actual
documentation, much like you'd have for code. I know the OpenDocument format
is XML based, but when OpenOffice saves an OD file, it's actually a zip file
containing several XML files in a couple of directories. I'm figuring it
would be possible so script the unzipping of the file as part of the
check-in to a repository (and re-zip on check-out - ugh), but when I think
about all that, it seems like you'd get too much extra stuff if you ever
wanted to compare revisions, and it may make little sense.

can you suppress the zipping step on export?

I'm willing to bet that for the average document most of the changes
happen in one file.

Right now, I'm experimenting with using HTML - written as a normal document
in OpenOffice writer, but saved as HTML, since _most_ formatting is
consistant, and I know when I look at diffs between versions it will be a
little more human-readable (thankfully OpenOffice's HTML is a lot cleaner
than MS Word's).

Any thoughts?

this outfit seems to do something like what you're talking about.
http://o3spaces.org/editions.html

but their source relese is slated for 2007

I haven't worked with oo.o much but it seems like this would be
relatively straightforward to implement as an export filter
_______________________________________________
EUGLUG mailing list
[email protected]
http://www.euglug.org/mailman/listinfo/euglug

Reply via email to