W3C home > Mailing lists > Public > www-svg@w3.org > November 2009

Re: Clarifications for <image> and Integration module

From: Doug Schepers <schepers@w3.org>
Date: Sat, 21 Nov 2009 19:24:59 -0500
Message-ID: <4B0884DB.3000805@w3.org>
To: Alex Danilo <alex@abbra.com>
CC: Patrick Dengler <patd@microsoft.com>, "www-svg@w3.org" <www-svg@w3.org>
Hi, Folks-

Alex Danilo wrote (on 11/21/09 5:38 PM):
>
> --Original Message-- (Patrick Dengler <patd@microsoft.com>):
>>
>> I'm hoping you can help with a few clarifications.
>>
>> <image>: The Document Structure module states:
>>
>> "For a single-channel raster image, the effect is as if the object
>> were converted into a 4-channel RGBA image, where the single
>> channel from the referenced object is used to compute the three
>> color channels and the alpha channel is uniformly set to 1."
>>
>> Is it expected that single channel raster images are actually
>> converted in memory to 4 channel, or just that the images can
>> "appear" (have 'the effect').
>
> As long as the result is the same, your in-memory representation
> doesn't matter.
>
> What that is implying is that if you have an RGBA backing store for
> example and you paint a single-channel PNG or similar that the alpha
> channel bits will get set.
>
> Remember the backing store, or new background when you go:
>
> <g enable-background="new">...</g>
>
> creates a new transparent black backing store. So if you paint a
> single channel image into it, you really do want to set the alpha on
> the pixels that get touched.

What Alex said.


>> Integration: I've been reading on the improvements to the methods
>> by which SVG can be surfaced on the web.  Is this module:
>>
>> http://dev.w3.org/SVG/modules/integration/SVGIntegration.html
>>
>> expected to be integrated back into 1.1 2nd edition, or is this to
>> be rolled into SVG 2.0?

This will be rolled into SVG 2.0, not SVG 1.1.  As much as possible, I 
would like to look forward at an improved SVG, rather than rehash SVG 
1.1.  If we can get general agreement from the major SVG implementers, I 
think we have a better chance at an interoperable set of functionality 
based around SVG 2.0 than if we focused only on SVG 1.1, because of 
known flaws in that specification.


> That looks like a really early work in progress.
>
> If you want to address integration issues, then I would highly
> recommend you read this document:
>
> http://www.w3.org/TR/WICD/
>
> You should take that document as the definitive guide to integration
> at the present time.

The SVG Integration spec is a bit early on, but its scope is very 
different than that of WICD... it is intended more for the purpose of 
other specs that wish to integrate SVG into that language, with detailed 
hooks and considerations for doing so, while WICD is intended as a truly 
well-rounded spec that details the interactions between compound 
documents by reference.

I think both are important, as they each cover different aspects of 
integration that aren't covered anywhere else.

WICD does have the minor drawback that it references SVG Tiny 1.2 
exclusively, with no mention of SVG 1.1, and also deals only with XHTML, 
not text/html.  This may have seemed like a logical choice at the time, 
but given the current climate, it would be more suitable to reference 
SVG 1.1 and HTML5 as well. This is a relatively minor change that could 
probably be incorporated into a revised draft, if we can get some 
resources to do it.

WICD is not quite at Recommendation status... because the group did not 
complete the process before rechartering, and attention of the 
participants was drawn away to HTML5, there were insufficient resources 
to continue the work of the group; however, because the deliverables 
were well-defined and mature, with complete test suites [1] that were 
missing only a few implementations of some tests [2], we decided to 
hibernate the group (rather than close it completely) while waiting for 
implementations to catch up.  I strongly encourage implementers of 
HTML+SVG+CSS engines to give it a serious look... many aspects of 
interactions between different document types are spelled out very 
comprehensively.  By comparison, HTML5 does not seem to go into quite 
the same level of detail and clarity for compound documents.

With relatively few resources, we could easily resurrect the CDF WG, 
revise the documents to include SVG 1.1 and HTML5, update the test 
suite, and bring the WICD specs to Recommendation.  I think that would 
really help get good interoperability for implementers and authors 
regarding compound documents.

[1] http://www.w3.org/2004/CDF/TestSuite/WICD_CDR_WP1/
[2] http://www.w3.org/2004/CDF/TestSuite/WICD_CDR_WP1/wicdmatrix.xhtml

Regards-
-Doug Schepers
W3C Team Contact, SVG and WebApps WGs
Received on Sunday, 22 November 2009 00:25:05 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:43 GMT