W3C home > Mailing lists > Public > www-svg@w3.org > July 2008

Re: Reference image of animate-elem-40-t

From: Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de>
Date: Mon, 7 Jul 2008 15:01:45 +0200
To: "Julien Reichel" <Julien.Reichel@spinetix.com>, www-svg@w3.org
Message-Id: <200807071501.45924.Dr.O.Hoffmann@gmx.de>

Julien Reichel:
> Hi,
> I was referring to #markergroup and indeed, and if you consider that the
> PNG refers to the images after the last animation, then you are right:
> the PNG is correct. Sorry for the noise of my previous email.
> In a general manner, I must admit that many of the test of the test in
> the test suite using animation would benefit of having PNG references
> for different animation times. As it is often not easy to exactly
> understand what should be displayed in the player :-) and one can easily
> overlook problems when testing document.

I personally think, that the reference image is anyway often of very limited
use for animation tests. Animation tests I create typically do not rely on
reference images and several of them try to be selfconsistent, indicating
at least typical errors of the tested feature in an easy way even for
smaller deviations. But this is not completely trivial to have for any 
feature to be tested and for many things tested it is not really necessary
to have a precision test. As here the timing of the marker is not centered,
therefore it is only a help to get a rough impression of the timing - it is
already much better than to have no indication at all ;o)

And for example if you have animations with repeatDur="indefinite"
obviously there is not reference image for the 'final' display typically
and the PNG shows something different. And if the animation is not
frozen with an interesting value, obviously the PNG for the final
state is not very informative at all ;o)

> The animation of the #markergroup in animate-elem-40-t is actually a
> very good example of small details that can be easily overlooked. And
> the fact that its an animation of a use that will itself be referenced
> by another use, makes it a complicate case to test.

Well for this test it is the intention of some subtests to test the 
behaviour of use, therefore it should be no complication to reuse
the markers. 

Received on Monday, 7 July 2008 13:07:03 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:54:19 UTC