Yes, this is a topic I have been interested in for some years. Creating documentation as a specialty seems an ideal use for LyX--at least, as it may become in future.
One notion that becomes very important in document creation, updating, and management in general is the notion of "chunking" into topics. One firm that has built a business out of this idea is Information Mapping Inc., although their methods are still somewhat controversial in the field. Thinking of topical "chunking" is nearly an unnatural act for many--but it may prove extremely useful if we are to get very far with the implementation of XML-style technology in documentation. One large problem is the lack of uniformity in the way people think of and create documents. On the simplest level, that includes format differences from individual to individual--format differences that are aided and abetted by programs like Microsoft Word that allows--nay, *encourages*--the profligate use of style tags and many ways to get a particular visible result in a document.j My interest in LyX has been powered by this observation, and the thought that something must replace our existing tools if we are to get the real benefits of the separation of content and form. In short, the entire "what you get is what you mean" approach is long overdue! However, the question remains whether LyX itself is the answer, or a recast tool based around XML. In other words, all the "stuff" that accompanies TeX and its complexity designed originally for typesetting may make the learning curve sufficiently complex that many may find it not to be worthwhile, especially for technical documentation that lacks many of the needs of the scholarly dissertation, for example. That is one thing I am attempting to determine by going through that learning curve myself! I delayed turning to LyX for several years, since for a time it appeared to the casual observer to be somewhat of a dying application. An aborted "Klyx" (kde specific) version, for example, did not appear to be a sign of progress. I believe, though, that if a stylesheet is made for the creation of DITA topics, there is little of the TeX stuff that needs to be used at all. Resulting topics could then be used by some as-yet-unidentified DITA map/content management facility with very little learning curve for the writers involved in topic content creation. While I am personally fascinated with typography, we are a comparatively dying breed these days, it seems. One thing I am thinking of is the feasibility of a context-sensitive right-click menu that would enable a quicker selection of a particular style attribute rather than just the drop-down box in the present interface. Such a dropdown could also be easily triggered with a keyboard shortcut for those interested in true productivity--or perhaps a handful of keyboard shortcuts for the principal style designationsk, somewhat similar to the ones in FrameMaker. LyX seems to be extremely useful for creating indexes, which seems to me to be an often-neglected part of the written information trade. Indexing, though, could become extremely significant in content management--especially since indexing itself is in a major sense the extraction and processing of metadata. Thus, the information chunks represented by topic creation for content management may be best served by creating index references at the time the topics are written and placed in the repository originally. That would seem to permit the locating and employment of the right chunks to meet any arbitrary need. Then, creation of new documents as collections of these chunks would be greatly eased. Anyway, pardon my musings tonight...there is much that I am still at sea about in this whole subject. David On 6/5/06, Stephen Harris <[EMAIL PROTECTED]> wrote:
David Neeley wrote: > I am trying to get far enough up the learning curve with LyX that I > can feel good about creating style files. Then, I intend to do a style > file for DITA--because I believe that LyX is, potentially, a far > better tech writing solution than anything else out there at the > moment. I also believe that DITA has advantages over most other > XML-based solutions for the purpose...and it should be far easier for > folks to learn and use consistently than DocBook, for instance. >
