DCF and Waterloo SCRIPT had a few differences in the interpretation of the dot commands, so often macros written for one didn't work on the other. Waterloo SCRIPT did support GML, but again, a slightly different set of tags than DCF, so that documents written for one often looked different on the other (or didn't format correctly).
The key to what I want is the Bookmaster tag set, which doesn't work with Waterloo SCRIPT. Ordinary GML is pretty boring and utilitarian; Bookmaster is the set of specialized tags that IBM created to write their own manuals, and the combination of a really good understanding of what's necessary to easily create significant amounts of technical documentation and the practicality of how to get that documentation written is what makes the DCF+Bookie combination interesting. I can knock out a really stunning set of docs for a product in a tiny amount of time, and you'd never know they didn't come straight out of IBM Information Design in POK -- it looks, smells, and feels like IBM documentation, and better yet, it works like IBM documentation. Good example: the OpenSolaris for z documents are created with DCF and Bookie: the same source generates plain text, PDF (via Adobe Distiller and the LISTPS file that comes out of DCF), HTML, and (via Bookmaster/BUILD VM), a Library Manager compatible file. You install them like IBM docs, they're structured like IBM docs, and they work with the same tools that IBM docs. (I know that IBM produces PDF versions of manuals; I have the CMS version of DCF and the rest of the Bookmaster tools and I use them to create my own. I just wish I could run them somewhere else, since IBM seems hell-bent on neutering CMS into just a virtualization layer management tool) > I know that DocBook has been mentioned in this thread and compared > unfavorably with DCF. I've used both, although I haven't used DCF for > many years. > We recently started using DocBook on a couple of projects, and overall we > were pretty pleased with it. I've used both as well; the comparison is closer if you compare GML to DocBook; raw DCF is rather like raw troff macros; not for the faint of heart. I think the problem I have with DocBook is twofold: 1) documentation on how to USE DocBook is nigh unto nonexistent. There's plenty of discussion about how it should work, and how various DTDs are applied and distributed, but there's almost nothing about how you actually *author* useful documents. Compare to Bookie: 3 page intro to what's happening, and you're producing useful output. This has been changing lately, but in comparison to the Bookie user guide, it's still very difficult to determine how to do simple things without inordinate amounts of research. (I know, I know, write your own damn book, but still...) If anyone knows of a good tutorial for DocBook, I'd sure like to know about it too. The Oreilly book on DocBook is pretty much useless, and it's the best I've seen (it's also visually ugly as sin, which is unusual for a Oreilly book). 2) XML is much more difficult to read and parse for humans than the simpler GML tag structure. I have editors that can do both, but if you're hunting for some weird formatting problem, it's a LOT harder to hunt that down in the XML files than in a flat text file with the simpler GML tagging. It's hard to create XML with XEDIT or ISPF, but as you say, Eclipse does a fine job. I can't run that on CMS, though (and AFAIK, no formatters exist for DocBook on CMS or TSO). A side note: another nice thing about Bookie is that it easily enables the native source file control stuff (ie CMS update) so versioning and maintenance of the docs is a LOT simpler. XML makes that very difficult -- you should see what a update file against a XML doc looks like after running EXECUPDT. 8-) I guess I should just buckle down and take the Bookie syntax definition and write a set of macros for TeX or troff that emulate them. It'd still be nice to have something on Linux that understands 1403 listings, though. Oh, well. C'est la vie. -- db ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
