Hi Michael,

I agree that we need to make sure we have a certain level of quality for 
the major features introduced in XML Schema 1.1. The one area I am aware 
of that need some investigation is CTA (type alternatives). When we 
originally implemented CTA (thanks Hiranya) it was based on the earlier 
version of the 1.1 spec where a very specific XPath subset was defined. 
The latest spec has removed that restriction. So, any XPath expression can 
be used (of course it will only work if it evaluates to attribute node(s) 
on the given element). So we need to revisit that. We could use the 
current CTA XPath evaluator for the older subset and maybe fall back to 
Psychopath if a more general XPath expression is used.

It would be great if Mukul and/or Hiranya can take a look and see what can 
be done, if they have some time :-).

Regards,
Khaled





From:
Michael Glavassevich/Toronto/i...@ibmca
To:
[email protected]
Date:
11/19/2010 09:03 AM
Subject:
Re: [jira] Resolved: (XERCESJ-1433) [GSoC]: Implementing XML Schema 1.1 
overriding component definitions (<xs:override>)



Hi Mukul,

Mukul Gandhi <[email protected]> wrote on 11/18/2010 04:03:26 AM:

> Hello,
>     I've been testing a bit the xs:override implementation that has
> been recently committed to Xerces SVN repos. It seems to be working
> fine and I'm happy with it's implementation.
> 
> Congratulations to Udayanga and also thanks to Khaled :)
> 
> I believe we now have nearly all pieces of XML Schema 1.1
> implementation available in Xerces-J and I think we might now work to
> release the Xerces 2.11.0 version formally. I think Michael is perhaps
> the best person to start that process.

I was hoping either Khaled and/or Sandy would chime in on other thread I 
started as they've been monitoring the XML Schema 1.1 development more 
than I have. The implementation is still experimental but I think there's 
a certain level of quality we'd want (at least basic / critical things 
working for the features we say we've finished) before we release it. I 
recall you were doing some testing. I know Sandy has been and think we 
should clean up any major issues prior to building a release candidate and 
bringing it to a vote.

Thanks.

Michael Glavassevich
XML Parser Development
IBM Toronto Lab
E-mail: [email protected]
E-mail: [email protected]

Reply via email to