Re: Preaching DSSSL (Re: using LOUT)

[Paul Prescod]

Chris Herborth wrote:
> ... LaTeX's formatting is driving us up
> the wall...  Strange side-effects abound, even simple things like
> spacing at the top of a page aren't 100% reliable.  Usually it's "good
> enough" (we've published a LOT of documents in the past two years), but
> it's next to impossible to tweak the format to look more like our
> traditional printed documents, and it's next to impossible to debug and
> fix the remaining formatting problems.
> 
> Moving to DSSSL (an ISO standard) using Jade (free and portable) and
> Lout (free and portable) would really make me happy.  I have long-range
> plans to move to Lout, but being able to just write up the document
> transformations in DSSSL and have it happen automagically would really
> be great.

Could you please describe to a clueless newbie how Lout will fix the
problems of LaTeX? Is that just a faith that LaTeX is bad and Lout is
good or is it the experience of people who use Lout that it is less
prone to mysterious behaviours.

Also, do you know whether DSSSL+TeX is prone to the problems you
describe? I would like to get a good handle (again, as a newbie) of the
advantage of a Lout back-end.

I think I planted (or re-planted) the DSSSL+Lout idea a few weeks back
asking various people about the connections. But before anyone spends a
bunch of effort making a Lout back-end it is worth deciding if it will
really offer significant value over the TeX back-end. My idea was that
DSSSL itself could be extended to allow the kind of precise layout
control that Lout allows. Lout's layout engine would do the actual
layout "work". 

But I believe that extending DSSSL in this way would be a research
project, not a lark. 

So what we are actually talking about is the more managable project of
making a Lout back-end of approximately the same sophistication as the
TeX back-end. Are there specific problems with that back-end that we
hope to fix with Lout? Are the problems more easily fixed with a Lout
back-end or corrections to the TeX back-end?

 Paul Prescod