context on demand / xml experiments / example styles / ...

context on demand / xml experiments / example styles / ...

[Hans Hagen]

Hi all, 

Last week we (that is, my collegue kees an me) have been setting up our
dedicated web server. Today I moved the ade site there, and i hope that it
went ok. Currently we run some experiments with the soon to be operational
site of a pragma offspring branch. I slightly changed the look and feel of
the pages and also added a toc [behind the first picture]. 

In 'links' you will find two new ones, these will bring you to the other
pragma site. 

Because we can setup multiple sites there, we have also a third one,
dedicated to some context activities. Later we will also set up a cvs
server [specially for tobias -)]. Because the machine is reasonably well
equipped [450 mhz with 256 k mem and some 20 Gig disk, connected by 100mb
to a pretty fast backbone in the usa, we can do some nice experiments with
on line publishing [we still are on isdn here, and it will take some time
till adsl will reach this village, which is a pain -)]. 

I have now set up several publishing on demand example workflows there: 

1 examns, generating tailor made examns from a 5000+ collection of items,
selection takes place in a pdf file, so this is purely pdf based [the pdf
docs involved are a nice example of what tex can do and they demonstrate
that one can (mis) use form fields for selecting pages in a fancy way] We
will use this database for other experiments as well. If your browser
supports byte serving, you can run the selections local, else you should
download the archives first.  

2 books, typesetting gutenberg project books, more on that later

3 qrc's, you can generate the latest qrc docs; this is a prelude to the
interface documentation project which some of you are involved in  

4 travel forms, generating complicated pdf forms 

More (advanced) examples will follow, like beginners manuals in many
languages [when ton has finished his updating, on which he's working right
now, those who have offered to translate [de, cz, ...] can go ahead]. I'm
also thinking of some mp related experiments (graphic paper things and so)
or interactive examns. When updated, the pdftex manual makes a good
candidate too. 

In these processes, users can set their preferences, like papersize, font
size, colors, indentation and more, when generating docs. 
This means that instead of keeping many versions around [think of A4/letter
* 10/11/12pt * 4 times whitespace * 4 times indent * 3 times color * n
times other things = many versions] we now have none but users can have
their personal preferences. [think of those who are still doomed to use
letter instead of a4] 

The second experiment, books, is kind of special, since it will be a
playground for context doing xml. The source code is xml, directly
processed by context, but the styles will be tex based, and usable for
other docs too. There will be multiple styles with many degrees of freedom,
so that you can see what the consequences of layout parameters are. Here we
can also play with protruding and hz. 

I want to make this books experiment (number 2) a 'context user' one, which
means that, as soon as we have the basic dtd kind of stable and some more
examples are converted from plain ascii to structered xml [this is not that
hard, since a perl script does the bulk, a few are done as an example,
while tobias is working on some german ones], we would like users to
contribute more converted files. There are some 3000 public books out there
in the gutenberg project, in multiple languages, so it would make a nice
big testbed for (1) styles and (2) examples and (3) xml processing. Of
course it needs some coordination, so if you want to contribute ... let us
know! I can also imagine that we support the ebook dtd [which is html
based, and therefore kind of crappy, but doable]. 

So, books will serve as sample style as well as xml site, but it's success
depends on the extend that users are willing to help convert the txt files
into xml ones. [i will do the styles, of which there is only a simple one
now -)]

Back to COD. The whole process of typesetting on demand is controlled by a
perl script that acts on simple project env setup files, while texexec does
the rest. Files are processed realtime in user dedicated space. This is
work in progress and parts of the toolkit may migrate in the context
distribution, although those of you who do this kind of things probably
have your own scripts, in which case you may benefit from extended texexec
functionality. There are several workflows possible, like pdf based, form
based, either or not directly, or in stages, etc.  

Hans

PS. In this context, there is a texexec feature that can be of use to some
of you:

  texexec --path=/original/path/of/tex/file --pdf somefile.tex 

combined with (the --use key is new)

  --env=listofenvironments and --use=listofmodules 

will set up context in such a wat that it will run on the current path but
pick up files from the path as given. [for this puspose a scratch file is
made]. This means that one does not have to copy files. Files like
somefile.xml are supported too, but the styles to handle that are not yet
public, since they are not stabelized yet (i want the freedom to redo
things in drastic ways).   

The basic xml handler will go into the distribution rather soon.

H. 

-------------------------------------------------------------------------
                                  Hans Hagen | PRAGMA ADE | pragma@wxs.nl
                      Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
 tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------