Replacing Docbook

Mike Evans mikee at saxicola.co.uk
Fri Aug 28 10:48:53 EDT 2015


On Fri, 28 Aug 2015 15:15:58 +0100
Mike Evans <mikee at saxicola.co.uk> wrote:

> On Fri, 28 Aug 2015 11:00:26 +0200
> Geert Janssens <geert.gnucash at kobaltwit.be> wrote:
> 
> > On Friday 28 August 2015 09:45:44 John Ralls wrote:
> > > > On Aug 28, 2015, at 9:03 AM, Geert Janssens
> > > > <geert.gnucash at kobaltwit.be> wrote:
> > > > 
> > > > Thanks for the heads up. That's certainly an interesting opportunity
> > > > to check out. On the other hand I wonder if markdown has enough
> > > > structure enforcement (for example to ensure contributors will
> > > > really use header markup instead of bold/underline where needed). I
> > > > do agree that docbook xml is a big hurdle for newcomers and even
> > > > not really appealing to more experienced people. So if we can find
> > > > a good middleground I'd be all for it.
> > > > 
> > > > Here's another option I have been pondering for a while, and just
> > > > now took the time to do some minimal research on:
> > > > http://blog.riemann.cc/2013/04/23/versioning-of-openoffice-libreoff
> > > > ice-documents-using-git/#comment-2209333934
> > > > <http://blog.riemann.cc/2013/04/23/versioning-of-openoffice-libreof
> > > > fice-documents-using-git/#comment-2209333934>
> > > > 
> > > > Move away from docbook completely and instead save our documents in
> > > > flat odt. Advantages: - This is a format that's easy to store and
> > > > manage in git.
> > > > - There is a free wysiwyg editor that's universally available:
> > > > libreoffice. Most people learn how to use it relatively quickly as
> > > > most of them have used word processors before. - libreoffice can
> > > > export to pdf. I even installed a plugin once to convert odt to
> > > > epub, which worked reasonably well. - libreoffice can also be used
> > > > headless for document conversion so it can be integrated in
> > > > automated build processes. - in theory libreoffice can even export
> > > > to html (though I have no idea of the quality).
> > > I’ve used libre/openoffice to create html. It works reasonably well.
> > > Calibre’s docs say it can ingest ODT, which will take care of the
> > > ebook and mobi outputs.
> > > 
> > > http://open.comsultia.com/docbook2odf/about
> > > <http://open.comsultia.com/docbook2odf/about> looks like the
> > > least-obsolete way to convert from DocBook to ODT. Its SVN repo was
> > > last updated in 2009 but our DocBook version is pretty old too so it
> > > will probably work OK.
> > > 
> > > Any disadvantages?
> > > 
> > The first one is mentioned in a comment of the blog post I referred to: libreoffice saves soft 
> > page breaks. These tend to all shift up and down when making edits. So the diffs will be 
> > slightly cluttered. We may be able to fix these by simply filtering them out before committing, 
> > however that would again mean a manual action for the contributor. We may also just decide 
> > to live with them.
> > 
> > That's the only disadvantage I'm currently aware of. A trial run will likely give us more insight 
> > and in worst case reveal potential show-stoppers.
> > 
> > Geert
> 
> I just tried docbook2odf with the guide and it does a reasonably good job.  Some imgaes need resizing and there's a "Generated by docbooktoodf" side bar on every page.  A little editing though and it should look OK.
> 
> Using GIT with LibreOffice works quite well after installing some python utils.
> 
> Mike E 
> 
Oh, dear.  More scrolling revealed loads of missing images.

Mike E


-- 
PGP key:
http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x00CDB13500D7AB53  



More information about the gnucash-devel mailing list