W3C home > Mailing lists > Public > www-tag@w3.org > March 2013

Re: Why polyglot is needed

From: Robin Berjon <robin@w3.org>
Date: Wed, 20 Mar 2013 13:25:40 +0100
Message-ID: <5149AAC4.4080602@w3.org>
To: "Henry S. Thompson" <ht@inf.ed.ac.uk>
CC: www-tag@w3.org
On 19/03/2013 20:40 , Henry S. Thompson wrote:
> The point is this is a common situation as technologies transition.

I fully agree that transition strategies is an architectural topic, a 
difficult one at that, and that it could be an interesting subject for a 
finding (likely one more generally on versioning).

But what you're describing isn't a transition problem. You're trying to 
publish XML content to a system that has presumably, in one form or 
another, been serving HTML since before XML existed. It's not seeing 
pains from transitioning from XHTML to HTML — it never supported XHTML 
to start with.

I'd like to remind the TAG that we've worked on this before:


It is customary in groups not to reopen discussions unless there is 
substantive new input. So can anyone clearly tell me what problem:

     a) Isn't solved by placing HTML serialisation at the end of an XML 
tool chain (as supported by XSLT); or
     b) Isn't solved by placing an HTML parser, making use of Infoset 
coercion, at the front of an XML tool chain; and
     c) Is of an architectural nature.

Otherwise, as Chaals says, this is a deployment issue. It is the HTML 
WG's, and the broader HTML community's, job to help XHTML users 
transition to HTML.

I don't have a strong opinion on whether polyglot is an important 
component in helping that happen. I used it for a couple years at first, 
but now that the (non-browser) HTML tool set has sufficiently improved I 
no longer see the need for it. I do however not see where this is a 
current architectural issue. At best it is a problem of improving the 
existing libraries — something I do believe we should encourage and support.

Additionally, I remain convinced that there is a goldmine of value for 
all were the XML community to cast its nighted colour off and not 
forever with its vailèd lids look seek for well-formedness in the dust. 
Something like JSONiq is absolutely wonderful, and XQuery on HTML would 
be very valuable (I know I'd use it).

But that's not an architectural issue either.

Robin Berjon - http://berjon.com/ - @robinberjon
Received on Wednesday, 20 March 2013 12:25:49 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:33:19 UTC