Lists Home |
Date Index |
Dare Obasanjo wrote:
> I read it. It reminds me of Joe Gregorio's posts on Regex-able XML[0,1] which makes similar
> complaints about working with XML that point to problems with APIs
> as opposed to problems with XML.
Dare, I agree that many of the problems are with poor APIs,
which we've discussed in the past , and as I pointed out
there, some of the APIs are better than others because they
go beyond the standards, in this case something Microsoft deserves
> Most of the suckage in XML is abstracted away by the XML Infoset and the technologies for processing infosets.
The crux of my complaints was that 'abstracted away by the XML Infoset'
is a pre-requisite, and that is a problem. That is, the bits-on-a-wire format of XML is
such that is *necessitates* an API to work with it.
> If anything my top 15 list would mostly be complaints about APIs and maybe a gripe or two about QNames in content.
> As for Bradford's clean namespaces proposal, I didn't like it.
> It seemed like an overly verbose way of getting around the scoping issues that can exist in
> documents that conform to the Namespaces in XML recommendation.
I am kind of suprised at a verbosity argument, this is XML after all.
Despite it's verbosity, I liked it because it is a backward
compatible solution and it would get me that much closer to my pipe-dream
of regex-able XML. Throw in xml:href, xml:src and xml:id and I'd
be in heaven :)
>  http://bitworking.org/news/40
>  http://bitworking.org/news/42