Migrating to XML: The Case of the GUST Bulletin Archive

摘要

Ten years of experience with TEX publishing of the GUST bulletin shows that Knuth's dream of highly portable TEX files is apparently an illusion in practice. Over the last decade, articles in the GUST bulletin have used at least six major formats (LATEX 2.09, transitional LATEX+NFSS, LATEX 2ε, plain-based TUGboat, Eplain, and ConTEXt), numerous macro packages, fonts, and graphic formats. Many old articles are typeset differently nowadays, and some even cause TEX errors. This situation motivates the following question: how do we avoid the same problem in the future? As the World Wide Web is quickly becoming the mainstream both of publishing and of information exchange we argue for migration to XML - a Web compatible data format. In the paper we examine a possible strategy for storing GUST articles in a custom XML format and publishing them with both TEX and XSLT/PO. Finally, the problems of converting the TEX files to XML and possibility of using TEX4ht - an authoring system for producing hypertext - are discussed.