Hi folks, I've just added a build target for creating a DTD-only xercesImpl jar file. To all appearances, it seems to work, and I didn't have to make all that many changes to the rest of xerces (pretty much only taking out hard dependencies in various files on schema-related classes, and StandardParserConfiguration needed to be refactored to make it inherit from an explicitly dTD-only configuration.)
Aside from the fact that this seems (to me anyway :-)) kind of cool, do people think it would be sufficiently useful to add to xerces? Personally, I can't see how it would hurt, but perhaps others do. A related question: one of the motivators for doing this was to decrease the size of our huge implementation jar. The standard jar we distribute is around 1.7 Mb; if we compress it, it falls to about 860K. Should we start compressing our jarfiles? This question has come up before but has engendered little comment, so we've stuck with tradition (and probably faster loading times) and continued to distribute uncompressed archives. BTW, with compression on, the DTD-only implementation is only 530K. This is a savings of nearly 40% over the compressed schema-supporting archive... schema support does not come cheap! And I haven't even looked at removing the serializer--that task should be easier and I was hoping someone else might try that some time. :-) Cheers! Neil Neil Graham XML Parser Development IBM Toronto Lab Phone: 905-413-3519, T/L 969-3519 E-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
