Re: SVG IG Torture tests

I've updated the MIME types.  This can be done before check-in like so:

$ svn propset svn:mime-type text/html foo.html

or image/png or image/svg+xml etc

http://svgtorture.googlecode.com/svn/trunk/unsorted/radial_gradient_test.html

Anyway, this looks like a pretty involved test!  Don't have a lot of
time to go over it atm - but it's pretty :)

Regards,
Jeff

On Sun, Feb 7, 2010 at 7:23 PM, G. Wade Johnson <gwadej@anomaly.org> wrote:
> On Sun, 7 Feb 2010 17:32:03 -0600
> Jeff Schiller <codedread@gmail.com> wrote:
>
>> Wade,
>>
>> The 'unsorted' is where I've been putting my ad hoc tests.  You can
>> take a look at the examples there to get an idea of what I considered
>> an acceptable format - but feel free to invent your own.
>
> I've put it under unsorted. I'm debating how much explanation should go
> on the page to explain expected behavior. (For that matter, I'm a bit
> unclear on whether this is actually the expected behavior.
>
> G. Wade
>
>> On Sun, Feb 7, 2010 at 5:30 PM, Doug Schepers <schepers@w3.org> wrote:
>> > Hi, Wade-
>> >
>> > G. Wade Johnson wrote (on 2/7/10 4:45 PM):
>> >>
>> >> Now taking the list in a completely different direction, I've got a
>> >> question on the torture tests.
>> >>
>> >> I stumbled across an issue with (at least) radial gradients that is
>> >> different between different browsers/viewers. It's not exactly a
>> >> "torture test", but it is a good compatibility test.
>> >
>> > That sounds perfect as a start on a set of practical tests.
>> >
>> >
>> >> Do we have a preferred form for this kind of test?
>> >>
>> >> I was thinking of an HTML page with variants of the SVG listed on
>> >> the left and PNGs with the expected output on the right, labeled
>> >> appropriately. Does that match other people's expectations? Should
>> >> each variant have it's own HTML page? Do we have any suggestions on
>> >> standardized naming, yet?
>> >>
>> >> If there are no defined approaches yet, I'll probably toss my
>> >> examples into svgtorture, assuming we'll move things around as
>> >> consensus gels.
>> >
>> > My suggestion would be to simply put whatever you have up there.
>> >  We can always change the format later, what we need is the meat of
>> > the tests.
>> >
>> > Regards-
>> > -Doug Schepers
>> > W3C Team Contact, SVG and WebApps WGs
>> >
>> >
>>
>
>
> --
> To vacillate or not to vacillate, that is the question ... or is it?
>
>

Received on Monday, 8 February 2010 01:39:44 UTC