W3C home > Mailing lists > Public > www-svg@w3.org > March 2004

Re: implementation size for SVG Tiny

From: Robin Berjon <robin.berjon@expway.fr>
Date: Thu, 11 Mar 2004 18:50:12 +0100
Message-ID: <4050A6D4.8020204@expway.fr>
To: Jean-Claude Dufourd <Jean-Claude.Dufourd@enst.fr>
Cc: www-svg@w3.org

Jean-Claude Dufourd wrote:
> Robin Berjon wrote:
>>> http://www.bitflash.com/purchase/buy_now.asp
>>> 4th line
>>
>> That is quite clearly an implementation of SVG Basic, not SVG Tiny.
> 
> OK, you are probably right. I have heard that their SVG Tiny player is 
> still in that range, not much smaller than the SVG Basic player.

There are big difference between Tiny and Basic, so if their player is 
still in that range I would expect that they started with a Basic 
implementation and removed things instead of doing Tiny directly (but I 
could be wrong). That would likely result in larger sizes.

> Now, looking at the w3c site, I see "PocketSVG", which is purported to 
> be Tiny plus something else, is 390Ko on PocketPC 2002.

According to their website, they support Tiny, half of Basic, and some 
stuff from Full. So it would seem to be quite a bunch more than a Tiny 
implementation.

You also want to note that eSVG is also a Basic implementation so using 
its numbers as a proof of SVG's size doesn't work either ;)

> On second reading, you are right, even though the sentence "very close 
> to being completely conformant" feels like "my program is almost 
> working" :)

Yeah I know, I was simply quoting TinyLine as an example, and as I said 
I am unsure how representative it is both ways (it probably has more 
bugs than one would want, and the code doesn't seem to be optimised for 
mobile as aggressively as it could).

> However, how much of property inheritance (the ability to put a property 
> like color anywhere above your rectangle) does the test suite test ?
> Or is this an optional feature ?

I have no idea how much it is tested, but I doubt it would be optional 
(can't find a place that would make it so). It would be a bit strange 
not to support that as one could then not animate several elements with 
a single animation. It's a bit like requiring transform to be on all 
elements instead of inherited.

-- 
Robin Berjon
Received on Thursday, 11 March 2004 12:50:45 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:26 GMT