Other GRDDL concerns which should not be lost

I wish to remind folks of the concerns I raised in this email:
	<http://www.w3.org/mid/FA273CF6-82A3-4A43-8B47-A6D6497112A0@cs.man.ac.uk 
 >

It seems rather negligent for the W3C to publish auto and invisibly  
downloaded code without:
	1) A security audit
	2) a QoS audit

I'll point to a potential issue with 2. Either GRDDL is widely used or  
it is not. If it is widely used, then that suggests that it will put a  
large load on an already overstressed W3C system. If it is not widely  
used, then that makes the provision of an XSLT in this matter much  
less compelling.

There are, ways out of that dilemma, of course! I do not mean to  
suggest otherwise. But I am very concerned that there was no response  
to my earlier email. I did the work of investigating a series of  
clients (including some of their source code).

<shrug/> It's pretty clear that technical issues and spec text matter  
very little on this issue. There's a vocal community that takes  
anything but an XSLT as cause for vocal objection. That's the reality,  
fair enough.

But I think the W3C should establish some sort of internal procedure  
for this. If then intend to host lots of GRDDL XSLT, these issues need  
to be dealt with. Indeed, the GRDDL recommendation *evidently* needs  
some sort of review. The spec has not done well outside people who  
were directly involved, which is a sign that there are issues with the  
spec.

Cheers,
Bijan.

Received on Thursday, 26 February 2009 09:01:57 UTC