Re: Proposal: Change end attribute behaviour

Dear Olaf,

Thanks once again for your response!

2011/3/25 Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de>:
> Wouldn't it be better to discuss this with the SYMM group?

The references at the foot of my previous mail list a number of
messages on this and related topics to www-smil over a few years.

Amongst these is the comment from Sjoerd Mullender regarding the reset
issue I raised in my previous mail:

  It might be a good idea if a new SYMM working group is to start and
  consider a new version of SMIL, to provide this issue as input. ...
  This is, however, a change in behavior and can therefore not be
  considered as an erratum to the current spec.[1]

That was in 2009. I think I've gone through the proper channels for
now and that we can't wait for another version of SMIL in the distant
future.

> I think, it does not help to deviate in SVG from SMIL just for fun and
> after 10 years with a behaviour already defined otherwise. The purpose of
> recommendations is, that one can rely on them, if the behaviour is
> often changed, for authors it simply means, that the changed feature
> or the complete recommenation cannot be used at all - this impression
> should be avoided to help authors to believe in SVG and that it can
> be really used for published documents not just for the moment, but
> for many years and in the future as well. Nobody really wants to check
> all published documents every year to new recommendations to be
> sure, that nothing was changed that modified the meaning of already
> published documents. It is annoying to have to fixed issues due to
> modified recommendations (what should be limited to contradictory
> or undefined issues).

I agree we don't want to change behaviour arbitrarily. If I understand
correctly, your concern is about existing content and existing
experience.

I'm not aware of existing content that relies on this behaviour but
would be very interested to hear from anyone who does.

Regarding existing experience, I am not concerned. I anticipate that
the number of authors with previous experience with SMIL and who wish
to apply that experience to SVG could be counted on one hand. By
comparison, the number of authors without such experience who wish to
create SVG content must not only be orders of magnitude greater
already, but is likely to grow significantly in the near future now
that Opera, Chrome, Safari, and Firefox all support SVG Animation.

Given that most people will be starting with SVG (not SMIL) and that
there's not a lot of content using this feature compared to what is
likely to appear in the near future I think now is the time to fix SVG
Animation, to remove the traps, to iron out inconsistencies between
browsers caused by unnecessary idiosyncracies in the specification,
and to reward authors' intuition by providing consistent behaviour.

We're definitely on the same page about trying to get consistency
between user agents and I sincerely appreciate all your work in this
area. However, if the spec is "author-hostile" on some point I'd
rather fix the spec first rather than the implementations.

All of that aside. I'm genuinely interested to hear feedback about the
specific proposal. Has anyone else been tripped up by this end
instance condition? (I have been, too many times to count, and I
implemented it!) Does anyone rely on this behaviour?

Thanks very much once again Olaf for your help.

Best regards,

Brian Birtles

[1] http://lists.w3.org/Archives/Public/www-smil/2009OctDec/0004.html

Received on Friday, 25 March 2011 12:29:45 UTC