Push Button Paradise
Micah Dubinko
Fri, 10 Mar 2006
Is XHTML Modularization "pointless"?
Not too long ago, I said
XHTML Modularization is kind of, well, pointless.
After making a statement like that, it turns out that I'll be using the technology in my day-to-day work soon. More details on the work situation later, but it's actually the second time I've been face-to-face with M12N. The first was when editing XForms.
So, if you're in the business of documenting a markup language, M12N isn't pointless. From the viewpoint of an implementer (or by extension the viewpoint of the whole W3C Recommendation Track), however, it's not in the same ballpark.
Someone implementing a markup language might indirectly benefit from having a better laid-out specification, but otherwise won't care about M12N. -m
posted at: 23:54 | under: 2006-03 | 3 comment(s)
I guess you didn't follow my pointer in my comment when you originally said that there was no benefit in XHTML 1.1's modularization, so I'll summarize it here, because the benefits are not indirect at all. They're completely direct.
The paper accompanied an XML 2005 talk that I did about evaluating DTDs and schemas for possible adoption. I said that modularization is a key thing to look for when adopting any schema, for two reasons:
- It lets you adopt a specific subset of the modular DTD without worrying about your users trying to use features of the full schema that you don't want them to use. WAP is an obvious example.
- A modular architecture is much more amenable to your addition of new features specific to your needs, because it provides hooks for plugging your new parts in.
In addition to WAP, I described another schema that benefited plenty from building on XHTML 1.1 instead of XHTML 1.0: PRISM, for which I was on the working group.
I also wrote "while some people saw no new features in the transition from XHTML 1.0 to XHTML 1.1, the PRISM group saw an obvious improvement to take advantage of." There's nothing shadowy or indirect about it.
Again, check it out: http://www.idealliance.org/proceedings/xml05/ship/30/industryschemas.HTML#d0e281
More on PRISM:
http://www.prismstandard.org/
Bob
Posted by Bob DuCharme at Sat Mar 11 13:13:00 2006
I think we're mostly agreeing.
I didn't say that XHTML 1.1's modularization was pointless--in fact I agree that it was a great improvement.
Going back to my earlier posting, the general XHTML Modularization spec has a different relationship to the W3C Rec Track than "primary" specifications like XPath or XHTML, for which one can sit down and implement in code. As an example, if the entire Modularization spec was an unofficial page from the Working Group (along the lines of similar things* the group has produced), it would still serve basically the same function. Not so with most other things on the Rec track.
I agree with all the benefits you've listed. That's why I'll be using M12N for my upcoming project. :) -m
* e.g. http://www.w3.org/MarkUp/Forms/2003/xforms-for-html-authors.html
Posted by Micah Dubinko at Sat Mar 11 23:16:34 2006
Posted by Sebastian Schnitzenbaumer at Thu Mar 23 22:54:54 2006
