Re: Validity of test DOM 2 HTML HTMLAppletElement 05

On Feb 2, 2007, at 8:56 AM, Joćo Eiras wrote (on www-dom-ts):

>
>
> The test HTMLAppletElement 05 of DOM 2 HTML
> http://www.w3.org/2004/04/ecmascript/level2/html/ 
> HTMLAppletElement05.html
>
> is not much valid, I think.
> The test checks if the value held by applet.codeBase is the same as  
> declared in the page markup.
> Both Opera and Mozilla resolver the src, href, data and codebase  
> attributes for elements. Accessing these properties directly in DOM  
> elements returns a resolved url, while accessing using getAttribute 
> (..) returns the original value, also for the sake of using selectors.
> Is there as part of the specification where these properties are  
> required to be the original values ?

Probably more effective to have interpretation questions on www-dom,  
not www-dom-ts, so moving the conversation there.

This case seems similar to errata item html-1 (http://www.w3.org/ 
2000/11/DOM-Level-2-errata#html-1), but there isn't a "ditto for  
HTMLAppletElement.codeBase" in the errata.  The discussion that  
resulted in the errata is http://lists.w3.org/Archives/Public/www-dom/ 
2003JanMar/0064.html

Also, "base uri" is used in the description for codeBase and  
according to http://www.ietf.org/rfc/rfc2396.txt a base URI must  
include a scheme which would support a case for resolving a relative  
URI.

I'm not sure of the process now that the DOM WG has expired.  There  
paths forward seem to be:

1. Endorse the test and say that resolving the URI is prohibited
2. Modify the test to accept either a absolute or relative URI as the  
return value from HTMLAppletElement.codebase.
3. Issue an errata mandating resolution and modify the test  
appropriately.

Probably would be good to survey the implementations to see their  
behaviors and review the recommendation to see if there are other  
attributes that might have the same ambiguity.

Received on Friday, 2 February 2007 20:12:16 UTC