Gentle reader, 

At PRAGMA we are using ConTeXt. With ConTeXt we are able to 
compose complex documents, for both paper and screen. Here 
are some features. 

(1)  It's quite complete. It offers a lot of functionality 
     in an integrated package. This functionality is packed 
     into a few hundred highly structured commands. 

(2)  The package is parameter driven, which means that one 
     can just specify what one needs. One does not need to 
     be a TeX expert to get things done. 

(3)  The package is more or less object oriented, which
     means that one can have many registers, many kinds of
     tables (of contents), many types of whatever. 

(4)  Of course there are facilities like blockmove and reuse
     of texts, where location specific info is saved. Tables
     and registers can be generated at every level.  

(5)  There is driver independant color support, there is an 
     advanced colored verbatim TeX mode and there are visual
     debugging tools. 

(6)  ConTeXt offers advanced interactive document support.
     Of course there are normal hyper links possible, but
     ConTeXt can provide navigational tools, like menus and
     scroll bars in a lot of flavors. It's all programmed
     in TeX in a driver independent way and nearly
     everything is done automatically. Special support is
     implemented as a kind of drivers. Tables and registers
     can be interactive in several ways. Register entries
     can be linked. One can design parallel documents and
     reader profiles are supported (soon in a more advanced
     way). 

(7)  A major feature is the multi-lingual interface. This
     means that one can give commands and specify
     parameters and keywords and receive messages in his or 
     her own language. We've implemented the dutch and english 
     interface, but are thinking of more. 

(8)  Itemizes lists (upto many levels, with optional
     specifications) are optimized in a second pass, just
     like margin words and illustrations. We made as few as 
     possible compromises when implementing things like this. 

(9)  Although in constant state of improvement, there are 
     multiple columns, including insert optimization. 

(10) There are lots of ways to process floats. Eps and Tif
     figures can be manipulated in many ways and placement
     is straightforward. For instance one command is used,
     which accepts an optional argument that specifies the
     location, a second one for the reference and two more.
     Side floats are supported too. One can also define his 
     own type of floats. 

(11) The layout can be very complex and includes areas for 
     navigational stuff and company logos. There is single 
     sided, double sided and combined are supported. 

(12) Project wise organisation of texts is supported, which
     enables structured management of texts (we use this
     for educational projects, which often consists of a
     lot of related and reused texts). This feature is also
     supported by the DOS oriented program TeXEdit, which
     supports fast browsing, on the fly spell checking,
     help-info and tex processing. 

(13) There is a lot to tell about the more common features,
     which go beyond those offered by most of the packages
     that are available today. One is invited to take a
     glance at the (still dutch) 450 page interactive manual 
     (which is cross linked to it's source) or the dutch and
     english quick reference cards. 

(14) Chemical typesetting is supported by a related package
     PPCHTEX and we are working on SCIENCE that handles units. 

(15) The whole package is structured in modules which are 
     all compiled to the format file ConTeXt.fmt. At the 
     moment we are documenting the source. We are half-way 
     now. The source will be optimized to the (which is of 
     course our) limit. The (growing) package itself is used 
     in a production environment for about 5 years now, which 
     means that it works!

We are still not sure about putting ConTeXt into the public
domain. It took us quite some years to write the package
and it will cost us some more to enhance it, provide more
hooks to the outside world, and fully document it. The
question however is: 

(?)  Is there a need for such a package? 

For novice users ConTeXt is far more easy to learn and
manage than existing packages and advanced users (or those
getting tired or impatient of waiting for announced 
packages) can take full advantage of all the features. We
will not go as far as stating that ConTeXt is the state of
the art macro package today, but some others did so for us.
But: 

(!)  We will only take the trouble if ConTeXt is needed and 
     wanted. 

We cannot support it in all those languages and our 
help-desk capacity is limited. Writing an english manual 
makes only sense when people use such a system and others 
have to carry out some support too. So again: 

(?)  IS THERE ANY NEED FOR SOMETHING LIKE CONTEXT? 

If so, we can download to ctan (where there are already a
CONTEXT entry housing PPCHTEX and some generic modules): 

(1)  the dutch quick reference manual
(2)  the english quick reference manual
(3)  the dutch interactive ConTeXT manual

Later, depending on the response, we can post, 

(4)  the dutch format file
(5)  the english format file
(6)  some auxiliary programs

For the moment the format files will be in Huge EmTeX format 
(ConTeXt needs a hash table >>10.000, due to its character).
 
And when people still want it, finally the whole bunch,
including source files and documentation will be posted. 

We will keep on improving and extensing ConTeXt, just
because it's our core tool. A next step will be providing
more hooks for those who want to do even more nasty things 
with ConTeXt. Our real challenge however is a seamless link 
between ConTeXt and SGML, probably by means of DSSSL. 

We would like to hear if there is any need for ConTeXt. If 
needed, the package can be provided with german, french etc 
interfaces too (implementing those is only a matter of building 
some translation tables), but for this we need the help of user 
groups.
  

Greetings, 

Hans Hagen & A.F. Otten 

pragma@pi.net