home | list info | list archive | date index | thread index

Re: [OCLUG-Tech] anyone out there using a production publishing toolchain?

  • Subject: Re: [OCLUG-Tech] anyone out there using a production publishing toolchain?
  • From: "Prof. John C Nash" <nashjc [ at ] uottawa [ dot ] ca>
  • Date: Wed, 16 Feb 2011 10:06:10 -0500
I recently completed a 313 page book using Latex. The learning cost is heavy, but the
rewards are too.

Latex is also used by a lot of technical / scientific publishers. I've a paper in process
at the moment. Style sheet supplied. Bibliographic support is strong, but again needs
learning.

There are also recent developments in what is termed "reproducible research" to allow code
to be embedded in Latex. We use this to write articles where the statistics (procedures
and data) may change. Running special scripts in Sweave and ODFweave allow latex and/or
OpenOffice docs to be processed to finished pdf. There are apparently other initiatives
like this, but they suggest a useful way to incorporate code and data into documents that
avoid copy and paste errors.

JN


On 02/16/2011 07:08 AM, Robert P. J. Day wrote:
> 
>   is anyone out there using a (preferably xml/docbook-based)
> publishing toolchain for publishing books, manuals or courseware that
> runs on Linux?
> 
>   i've got a number of old courses, and more i want to write, and i
> want to update my publishing toolchain.  that i want to use some form
> of xml/docbook is pretty much a no-brainer since that will give me the
> most flexibility in terms of output rendering, but there are a number
> of choices even there, so i was wondering if anyone on this list is
> using some type of toolchain like that for regular production use at
> their place.
> 
>   the first option is to use straight docbook, which will definitely
> work but is kind of verbose since you typically need to type in those
> overly long tags.
> 
>   one variation on that is that, way back when, i wrote a pre-docbook
> processing step that used an XSLT processor (xsltproc) that let me add
> a bunch of specific shorter tags (like <p> -> <para> and so on), then
> i ran *my* original content through what's called an "identity
> transform" that simply replaced my shortcuts with the proper docbook
> tags.  (and, no, it's not as simple as you think.)  so that gave me
> the ability to use proper docbook but save a lot of typing by sneaking
> in my handy little transform at the beginning.
> 
>   yet another option is publican -- https://fedorahosted.org/publican/
> -- which is a packaged docbook/xml toolchain that lets you do neat
> things like branding your documents.  so it hides a lot of the grunt
> work.  i'm just checking it out, and if anyone else is using it, i'd
> be interested in knowing what you think of it.
> 
>   the final option that i just ran across is "slacker's docbook,"
> http://slackerdoc.tigris.org/.  i found this only because i just
> finished pre-pub reviewing the next edition of this book:
> 
> http://www.amazon.ca/Introduction-Design-Patterns-Qt/dp/0131879057/
> 
> and noticed that the authors stated they'd written the whole thing
> using that reduced form of docbook, which conveniently adds a whole
> bunch of time-saving features that are normally a bit of a pain with
> regular docbook.  so that's now an option as well.
> 
>   so ... any opinions?
> 
> rday
> 
> p.s.  i could almost make a talk out of this, couldn't i?  :-)
>