HTML WG comments on DI Select

(Apologies for late comments: we discussed DI select extensively last week  
at our FtF [1].)

It pains me to have to write this message, but I am sorry to say that the  
HTML WG has instructed me to report to you that it does not support the  
underlying approach of DI Select. This should not surprise you however,  
since we have reported this position before at HCG teleconferences.

The basic problems are:

1) It takes us back to the bad days of mixing content with presentation  
details.
2) The selection mechanism is reminiscent of the C #ifdef style of content  
selection, which has proven to be difficult to manage, and results in a  
spaghetti code that has to be tested on all combinations of platform  
before you can be reasonably sure that it is correctly organised.

CSS and XSLT have successfully shown that it is possible to repurpose  
content for different devices without having to impinge unnecessarily on  
the content. One reason that this is such a good approach is that when a  
new device comes along you do not have to touch the document(s) but just  
add to the stylesheet. It seems to us that the main missing part for DI  
selection in the CSS/XSLT approach is preventing unwanted content from  
being sent to the device in situations where bandwidth is a limiting  
factor, and allowing it to be sent later on request.

Our recommendation is therefore to take the lessons from CSS and XSLT and  
if necessary add to them to allow your extra use cases, and allow  
transformation and filtering of documents where bandwidth is scarce. We  
regret to say that we cannot support the current approach; we therefore  
also refrain from any further detailed comment.

For the HTML WG,

Steven Pemberton

[1] http://www.w3.org/2005/06/13-html-minutes.html#item06

Received on Wednesday, 22 June 2005 14:55:51 UTC