Re[2]: [cgmo-webcgm] CGM_Metafile test

[Benoit Bezaire <benoit@itedo.com>]

Hi Lofton,

I'm not trying to reopen an old issue. But I did want to point out
that the test is wrong...

-- 
Regards,
 Benoit   mailto:benoit@itedo.com


Thursday, November 24, 2005, 11:54:13 AM, you wrote:

> At 10:20 AM 11/24/2005 -0500, Benoit Bezaire wrote:
>>Hi,
>>
>>   The current WebCGM 2.0 specification says the following about
>>   events:
>>   - If there are no graphics objects whose interactive region is under
>>   the mouse (i.e., there is no target object), the event is not
>>   dispatched.
>>
>>   This does imply that clicking on some graphical primitives that are
>>   not part of an Object (or clicking on the background) does _not_
>>   generate an event. This was raised by Lofton (if I remember
>>   correctly) and no body objected.

> Yes.  No target object => no event.  Discussed and resolved issue.

> Note that the CS text still has the unclear wording between the first 
> 3-bullet list and the second 3-bullet list.  The 3rd bullet of the first
> list seems contradictory to the execution of bullets 2 & 3 of the second
> list.  (Everyone seems to understand what it means, but no one ever 
> volunteered for requested rewrite and it slipped through the cracks.)


>>   The problem is that the CGM_Metafile test does not contain any
>>   objects... and we can't test the addEventListener function because
>>   of it.
>>
>>   If the specification is correct, then the test is incorrect and
>>   needs to be updated.

> I think the specification is correct.  Does anyone disagree?

> (I think this matches SVG 1.0, altho' 1.2 apparently has changed it so that
> the <svg> element (sorta like the picture) can capture the event.)

> -Lofton.