DR-09-0029: WML, Custom XML and Smart Tags: Specification for Validation Lacking

Shawn Villaron shawnv at microsoft.com
Mon Jun 8 04:45:08 CEST 2009


Sorry for letting this sit for so long.  Looking at my email regarding this defect report, there were a few open issues:

1. Caroline asked about why we have doNotValidateAgainstSchema. 

The short answer is that we felt that there could be the case where content is included in the document but that it's not ready to schema validate.  For example, suppose you had a schema that required three pieces of data.  The customer is using the application and only enters two pieces.  But now the customer wants to save their data and go home for the evening.  If this case, you could take the incomplete data and set doNotValidateAgainstSchema to true.  Then, when the document is opened, the application wouldn't report an error on open.  When the customer completes filling in the data, doNotValidateAgainstSchema is set to false and the data can be validated.  

I'm not sure how useful this is to put into the standard; naturally, if WG4 believes that there is value, we can include it as part of the guidance/note/rationale for this item ...

2. Murata-san asked about the differences between WML and SML.  

Again, I'm not sure this warrants modifying the text in the standard, but our thoughts where that the schemas was critical to the SML scenarios but what helpful in the WML scenarios.  So we felt it was okay to allow the schema to exist externally for WML, but weren't comfortable with that for SML.  You can imagine the scenarios where an externally stored schema gets detached from the document ( you email it to someone, move it to a memory stick, etc. ). 

3.  Murata-san was going to log a new DR regarding the requirements of the schemaLocation attribute.  We'll handle related questions with that defect report.

Modulo the first two issues, I believe we're okay with respect to the required changes to start the closing process for DR-09-0029.  Please advise as to what you believe our next steps are here.

Thanks,

shawn

-----Original Message-----
From: MURATA Makoto (FAMILY Given) [mailto:eb2m-mrt at asahi-net.or.jp] 
Sent: Wednesday, April 22, 2009 7:48 PM
To: 'e-SC34-WG4 at ecma-international.org'
Subject: Re: DR-09-0029: WML, Custom XML and Smart Tags: Specification for Validation Lacking

Shawn,

Thank you for providing the text.

I am thinking about the differences between custom schema lookup for WML and that for SML.  In my understanding, custom schema lookup for SML always uses  Schema elements within the MapInfo element, while that for WML references to a schema by specifying an external location.  A note about this difference might help.

BTW, I find that 29500 specifies almost nothing about the values of the schemaLocation attribute.  Can it a be an absolute IRI or relative IRI reference?  Are fragment identifiers allowed?  I do believe that non-ASCII characters (i.e., IRIs) are allowed.  I will submit another DR for this.

Cheers,

--
Makoto <EB2M-MRT at asahi-net.or.jp>





More information about the sc34wg4 mailing list