W3C home > Mailing lists > Public > www-svg@w3.org > February 2010

Re: Rendering and Pointer-Events (ACTION-2372, ACTION-2358)

From: Domenico Strazzullo <nst@dotuscomus.com>
Date: Fri, 26 Feb 2010 17:44:33 +0100
Message-ID: <F8A52599CB4645F0A51E50125C6E54BD@RAPAX>
To: <www-svg@w3.org>
Hi,

I am totally pleased that this issue is being resolved and I approve the 
resolutions. If ever the different interpretations could be considered 
reasonable, that may be true in theory. In practice, when dealing with 
complex situations, the webkit behavior (Chrome only) in relation to 
clip-paths makes an interface unusable, very simply speaking. While 
finalizing GEMÔ 2, frustrated of not being capable of finding a valid 
workaround to palliate for this, let's call it, peculiarity I have had to 
shamefully hack by replacing some relevant (offending) <g>s + clip-paths by 
<svg>s. I'm sorry, this is inadmissible. In the case of <g>, which, like 
Thomas De Weese rightfully says, is most often the element referencing a 
clipPath, this is possible, but when using one on a primitive you end up 
with this kind of comic situation, we have:

- One red object in the spatial region of one clipped green object in an 
upper layer.
- The two objects appear then one next to the other.
- They both have events; the red says "I'm red" while the green says "I'm 
green".

Question: what do you expect the end user to expect when she clicks on the 
red object? Unless we all want to reconvert to prestidigitation games 
developers, we would like our user to see the "I'm red" message.

In relation to the spec I again agree with Thomas De Weese when he says:

> The standard makes no reference to clipping with regards to hit testing,
> why someone would decide that they should added such a thing when it is
> totally absent from the standard makes no sense to me.

This is not an errata, it's a clarification at best. The point is, a spec, 
however exhaustive it may be, cannot possibly cover all of its subjects in a 
binary like manner, as it is not addressing AI enabled machines after all, 
or is it? And my question to the implementors is this: in doubt, why not 
simply ask, communicate, so that possible doubts on interpretation are 
dissipated? Who pays cash when dealing with oddities that are so reminiscent 
of the old html-css demons? The developers. Please, think also of the time 
and energy that go into these discussions to solve incertitudes that should 
be solved beforehand.

Regards,
Domenico Strazzullo 
Received on Saturday, 27 February 2010 13:13:44 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:44 GMT