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

Re: Scope of W3C recommendations; core issue for polyglot & DRM

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Tue, 29 Jan 2013 08:03:20 -0500
Message-ID: <5107C898.5010702@openlinksw.com>
To: www-tag@w3.org
On 1/28/13 9:52 PM, David Sheets wrote:
> On Mon, Jan 28, 2013 at 6:36 PM, Kingsley Idehen <kidehen@openlinksw.com> wrote:
>> On 1/28/13 6:30 PM, Karl Dubost wrote:
>>> Kingsley,
>>>
>>> Le 28 janv. 2013 à 14:07, Kingsley Idehen a écrit :
>>>> Put differently, XML usage and relevance (at broad Web-scale) is on the
>>>> decline.
>>> I have seen that sentence a few times. I have no idea how we backup this
>>> by real data on a timeline. Do you have hints? Or how do you evaluate it
>>> yourself?
>>>
>> For the "Web Developer" profile, all you have to do it look at places like
>> Github and other Open Source collectives. The trend is away from XML to
>> JSON.
> XML and JSON serve different use cases. XML is subpar for data
> structures but superior for structured and marked-up documents. JSON
> is subpar for documents but far superior for data structures (except
> for missing lots of useful type constructors like tuple product...).
> That lots of people decided to use XML for protocols and data is
> unfortunate and has been a major contributor to its poor reputation
> among typical web devs.
>
> Do you have evidence that XML is in decline for document processing
> workflows in favor of JSON?

No, and I don't believe I am making such a claim. The issue here is 
special status of XML. A status so special that we are ready to force it 
into HTML in ways that are net negative to the target audience, as 
already demonstrated by this entire thread.

>
> As far as I know, there is no standard technology with XML's features
> and deployed base. Are there other generic, extensible document
> mark-up languages out there?
>
>> BTW -- I use XML extensively and appreciate its utility for data
>> transformation and exchange, the problem is that it needs to succeed or fail
>> on its own merits. Squeezing it into HTML will never help XML survive or
>> succeed.
> What squeezing is necessary? HTML has a syntactic subset which is
> well-formed XML.

And how many folks do you seriously believe buy that in the HTML 
community? I can assure you, that's a tiny *specialist* minority.

>
> You make the assertion that XML *ought* "to succeed or fail on its own
> merits". One of XML's merits is being close in syntax to HTML (enough
> so that polyglot actually exists).

Yes, and as I've already stated in prior posts, the polyglot simply lays 
the foundation for problems, as I outlined re. schema.org and the net 
effect on developers of Linked Data consumption and publication tools.

>
> What normative principle leads you to the conclusion that suppressing
> XML-HTML compatibility Recommendations is letting XML "succeed or fail
> on its own merits"?

The fact that they aren't the same. XML as a subClassOf of HTML is a 
broken construction prone to the very problems accentuated by this 
massive thread. The retrofit isn't one that's accepted by the target 
audience of Web document publishers or developers.
>
> It seems to me, if you want to see XML succeed or fail, you shouldn't
> assume it has failed and then condemn it to failure by restricting
> compatibility information.

FWIW -- We've already been through the hell I am trying to prevent 
others from embarking upon. We have more than a 100+ transformers for a 
variety of XML based data sources (including XHTML, XHTML, HTML5) and I 
know that what we've been through isn't something others will embark upon.

My fundamental concern: most (X)HTML5 is published without appropriate 
hints to processors. Thus, you have to sniff on the content.

If you haven't attempted to extract a Microdata, RDFa, Microformats data 
island from an (X)HTML5 document you won't be aware of these problems 
out in the field. In addition, beyond schema.org, most online retailers 
(as a consequence of Schema.org) are producing the kind of problematic 
(X)HTML5 that I describe.



>
> David
>
>
>


-- 

Regards,

Kingsley Idehen	
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen






Received on Tuesday, 29 January 2013 13:03:42 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:56:51 UTC