This conversation is happening at a veryy opportune time, and I would
like to get your collective advice on a related issue:
We are in the process of deciding on a print output solution for the
Women Writers Project texts. These texts use a full TEI DTD, with some
modifications, and contain complex markup. It is likely that any output
would have to be further modified, by hand, to make it look nice, since
not all details of formatting are encoded.
The kinds of features we want to be able to get are: relatively easily
modifiable output for things like title pages, and margin notes. and
remember, this is the end result of a process that doesn't have to go
backwards, it's like making a pdf. Our archival form is the SGML, and we
can pre-process that as well.
1. We were originally leaning towards DSSSL -> tex. However, recent
discussions on the DSSSL list make it seem as if that is a dead end.
Won't be supported long, and output engines probably won't be refined.
Also, the tex output won't be easily tweakable.
2. XSL ain't there yet, and what is there is geared to DTD
transformations, not formatted output.
3. Frame+SGML may well be able to do what we want. It can produce great
output, and the results are easy to modify, has a fairly powerful SGML
import, and has a c/C++ API for adding your own modifications for things
like the margin notes, which have be placed in floating image frames.
Any thoughts or suggestions? Those of you with lots of DSSSL expertise;
do you recommend delving into DSSSL? I've done simple things with DSSSL,
but we want to process hundreds of texts reliably. Any experience with
Frame, especially negative? That's what we're leaning toward right now.
Scholarly Technology Group
Providence, RI 02912