W3C home > Mailing lists > Public > public-forms@w3.org > November 2010

Status Forms WG

From: Steven Pemberton <Steven.Pemberton@cwi.nl>
Date: Fri, 19 Nov 2010 12:00:19 +0100
To: w3c-html-cg@w3.org
Cc: "Forms WG" <public-forms@w3.org>
Message-ID: <op.vmezitrxsmjzpq@steven-750g>
The Forms WG had a 4 day face-to-face (and ear-to-ear) at the TPAC.
End of year travel freezes prevented several members from attending in  
we think it would be valuable if TPAC were earlier in the year to prevent  
this happening.

Nevertheless we had a productive week.

In particular, we worked hard on producing an XForms 1.2 initial draft.  
Because we have lost our main editor, we have decided to produce our spec  
using a wiki. We have adapted a program written by Sandro Hawke for the  
RIF and OWL specs (which were produced in the same way), and have  
succeeded in reproducing the XForms 1.1 spec:

	Wiki: http://www.w3.org/MarkUp/Forms/wiki/XForms_1.1_in_Wiki
	Transformed Wiki: http://www.w3.org/MarkUp/Forms/Test/XForms1.2/
	Original Spec for comparison: http://www.w3.org/TR/xforms11

We are now working on the XForms 1.2 spec:
although this version doesn't yet reflect all the text that was written in  
separate parts of the wiki before it was ready  

Another useful result of the meeting was an agreement on how to process  
JSON data in XForms. Several implementations already do this (in slightly  
differing ways), and we were looking for a generalised and as far as  
possible opaque way of doing it.

Since XML and JSON are just two serialisations of data, and JSON is a  
subset of XML in what it can represent, it turned out to be rather easy to  
define (after several false-starts). The JSON gets transformed into  
equivalent XML, so that XPath expressions treat the data as if it were XML.

So in general it is not visible within a form that the data has come from  
a JSON source. The MIME type tells the processor how to treat the incoming  
data, and it gets transformed into an XPath data model. When the data is  
re-serialised out, by default it is serialised to JSON.

For example

	{"company":"example.com", "locations":[{"place": "Amsterdam"},{"place":  

is transformed to


which allows a bind like
	 <bind nodeset="locations/place[1]" ...

Since this is only to allow the processing of existing JSON data, there is  
no requirement to serialise general XML to JSON, so only the subset case  
needs to be dealt with.

More details: http://www.w3.org/MarkUp/Forms/wiki/Json


2010-11-01 to 2010-11-05 Lyons / TPAC F2F


Best wishes,
Steven Pemberton
Received on Friday, 19 November 2010 11:01:02 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:14:04 UTC