Re: ISSUE-56 (urls-webarch) - Chairs Solicit Alternate Proposals or Counter-Proposals

On 04.03.2010 02:02, Ian Hickson wrote:
> On Wed, 3 Mar 2010, Julian Reschke wrote:
>>>
>>> Note that this is an escalation of this bug:
>>>
>>>      http://www.w3.org/Bugs/Public/show_bug.cgi?id=8207
>>>
>>> ...which was not rejected. I don't think anyone disagrees with the
>>> proposal, the problem is just that the text to be referenced -- the
>>> new IRI spec -- isn't done yet, so it's not possible to know _how_ to
>>> update the HTML5 spec.
>>
>> Actually, that change proposal includes a rewrite of the affected
>> section.
>
> With all due respect, the proposed text is so confusing I don't even know

It's not nearly as confusing as WEBADDRESSES which has no owner, no 
issue tracker, and nobody answering questions about in a timely manner.

> how to begin to review it. The biggest problem is that since the
> underlying text (the new IRI spec) isn't yet done, there's no way to know
> if the new text is accurate or not. For example, right now the IRI spec

That's the same situation as for a ton of other specs HTML5 is citing.

At least it has a working group where issues can be reported and discussed.

> 'Web Address processing' section is destructive to invalid absolute URLs,
> but I can't tell if that's a bug in the proposed IRI text, the proposed
> replacement text for HTML5, or something else.

1) Example?

2) Assuming you are right, do you have any evidence this is a problem in 
practice?

3) If it's a bug in HTML5, report an issue to *this* WG. If it's a bug 
in IRIbis, report a bug to the IRI WG. If you're not sure, discuss the 
issue on one of the mailing lists. Or both. That's what they are for.

> There's no point putting the roof on this house. We're still building it.

There's also no point in building a house on top of quicksand.

> Trying to update the "URL" definitions in the HTML5 spec is premature.

I'd appreciate if the chairs asked for a Counter Proposal then -- this 
would need to address the problem of a normative reference to an 
abandoned spec (and yes, that might mean moving it back into HTML5, but 
IMHO that would still be preferable to what we have now).

Best regards, Julian

Received on Thursday, 4 March 2010 09:04:24 UTC