W3C home > Mailing lists > Public > public-forms@w3.org > March 2011

Impressions concerning XForms at XML Prague 2011

From: COUTHURES Alain <alain.couthures@agencexml.com>
Date: Tue, 29 Mar 2011 23:23:05 +0200
Message-ID: <4D924DB9.7070105@agencexml.com>
To: Forms WG <public-forms@w3.org>
It was a pleasure to see at XML Prague 2011 that XForms has been 
recurrently mentioned in presentations and demos.

My own paper about JSON for XForms was received by some W3C Group 
members as in conflict with other works about the future XML Data Model. 
My argument about this objection was that this is already successfully 
implemented and supported by any browser due to good support of XML 1.0 
and XPath 1.0 without having to wait for browsers vendors to release 
versions supporting future recommendations.

 From my point of view, JSON support in MarkLogic Server is less 
interesting for XForms because it doesn't profit from xsi:type use and 
because it is using an extra element for arrays which is not good for 
XPath querying. Non XML names are converted using the _  character for 
escaping the decimal value (Micah's idea, apparently) of the not 
supported character; depending on implementations this might be slower 
or faster than using an extra attribut as we suggested. MarkLogic has 
developed a new query language for JSON in MarkLogic Server: this 
language is written in JSON and can be transformed into XQuery for 
processing. I wasn't personally convinced by this notation and some 
attendees pointed that it wasn't quite readable for them (JSON was 
heavily mentioned at the conference but wasn't it because it was a 
proposed theme for this session of XML Prague?).

At the MarkLogic DemoJam, I demonstrated the "file://" support in 
XSLTForms. I showed the StratML form and, also, a basic XML editor form 
(just for documents having text contents on leaves elements). Betterform 
people demonstrated their own extension for embedding forms claiming 
that forms are becoming far too big if such a functionality is not 
implemented. They argued that loading times are now acceptable, even for 
mobile phones. Another guy proposed "nanoforms" because of "XForms being 
too complex" and I was pleased to see the "...Loading..." dialog box of 
XSLTForms appearing in yet another demo also based on exist-db.

The main preoccupation at XML Prague this year was about XML 
technologies within browsers: Schema validation (open source alternative 
to Xopus), XSLT 2.0, XQuery and XMLHttpRequest enhancements. Interactive 
considerations were indeed studied for XSLT 2.0 (Saxon-ce) and XQuery: 
HTML event listeners, HTML DOM/properties access from functions or 
shadow elements. In fact, it was said that "XForms has gone much further 
than Saxon-ce in declarations of state/input interactions". XSLT 2.0 
templates being fired from event listeners is another approach for 
actions and, comparing to XForms 1.1, XSLT 2.0 has variables and sort 
capabilities in xsl:for-each. It was said that this is quite similar 
with XQuery in the browser and a need of coordination has been expressed.

People contacted me for training and support for XForms. Some 
implementors of other technologies told me that they worked sometimes 
ago with XForms developers but they didn't have the occasion to use it 
themselves. It sounds like a sort of reborn of XForms for them while new 
people surely assumed that XForms is a mature technology to be used as 
others.

Michael Sperberg-McQueen has already scheduled a new session for his 
XForms course (with XSLTForms!) for June in California and flyers about 
this training were proposed to attendees.

-Alain
Received on Tuesday, 29 March 2011 21:22:17 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 October 2013 22:06:55 UTC