|
|
|||||||||||||||
|
XML-SIG Developers Day SessionThe XML-SIG has been mostly drifting for the last year, with little attempt made to push out a 1.0 version. Part of this stemmed from external forces (little development on a Python Unicode type, no namespace support in SAX or DOM), and part stemmed from internal reasons (my getting distracted, few people having time to work on it).We need to finish a 1.0 version as soon as possible. Paul Prescod listed some standards that should be supported: XML SAX Unicode XPath XPointer DOM XSLT. XSLT seems too far afield, but the rest are probably worthwhile candidates. (Can we prioritize them?) Issues remaining: DOM Level 2 supportShould we switch to 4Thought's implementations of DOM, XPath, and XPointer? Participants should try to at least read through some of 4Thought's code, so they can form an opinion of its quality.Pros:
Cons:
SAX Namespace supportThis requires that XML-DEV converges to some consensus on SAX2. Has it done so? Are there issues facing a SAX2 binding that we should discuss?UnicodeWe've been waiting on the official Python solution. M.-A. Lemburg is implementing a proposal (http://starship.skyport.net/~lemburg/unicode-proposal.txt) but it hasn't reached the Python CVS tree yet. Therefore, we still haven't dropped wstrop. My inclination is to do this as soon as Unicode is reasonably stable in the CVS tree.Adding qp_xml.pyTo the xml.parser package, presumably?xmllib.pyA touchy catfight ensued on the xml-sig mailing about xmllib.py's standards compliance (with or without sgmlop). Should xmllib.py be dropped? Can we do an xmllib.py compatible class on top of Expat?Glue for Java and COM parsersWe don't have any particular support for Java-based parsers, or Microsoft's XML-parsing COM component, but we probably should. (Still, this is probably a low priority.) |