Subject: Re: Generating high-level formatting output From: "Sebastian Rahtz" <sebastian.rahtz@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx> Date: Sat, 12 Jun 1999 20:57:11 +0100 (BST) |
Adam Di Carlo writes: > > Personally, I think it would be wiser to go from SGML -> TeX, and skip > LaTeX altogther. That's because LaTeX has it's own ideas on how to > format stuff, whereas with the low-level TeX you'd have more control > and more robustness (fewer TeX infrastructure slings-and-arrows, such > as Babel breaking everything). > the corollary to that is that if you don't have LaTeX's Babel font selection scheme multicol, tabular packages backend-indepepdent graphics and color hypertext you have to invent it all for yourself. attracted as I am to Norman Gray's high-level LaTeX backend, I think its use would spell the death of DSSSL, whose only strength[1] is thea device-independent FOs. using it as a transformation language just makes it an `also-ran' alongside a myriad other languages. Most telling, to my mind, is James Clark's statement (I hope I read it right) that he thinks XSLT took all the lessons learnt from the transformation parts of Jade and did it even better. Given this, why would we not right write a "*ML to LaTeX" converter in XSL? I just hate the idea of DSSSL living on solely in the person of Jade running Docbook transformations for Linux documentors. Lets not give up on FOs, please! Sebastian [1] apart from the fact that Jade groks full SGML, unlike eg xt DSSSList info and archive: http://www.mulberrytech.com/dsssl/dssslist
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
Re: Generating high-level formattin, Adam Di Carlo | Thread | RE: Generating high-level formattin, Didier PH Martin |
RE: SGML entity mgmt stds (was Re: , Didier PH Martin | Date | RE: Generating high-level formattin, Didier PH Martin |
Month |