Re: Tinier SVG

On Thursday, November 14, 2002, 3:46:36 PM, Simon wrote:


SSL> chris@w3.org (Chris Lilley) writes:
>>No, my point was simpler. Citing that one implementation was 97k does
>>not mean 'SVG Tiny is always 97k' or 'SVGT never fits into 64k' etc.

SSL> My original point was just that SVG Tiny is not particularly tiny,

Its a lot smaller than Full, and substantially smaller than Basic. But
the implementations speak for themselves.

SSL> and there's still a lot of room for something useful but smaller.

A bit smaller, yes. Lots smaller becomes rapidly not useful or,
alternatively, discards things like accessibility or
internationalisation.

Of course, it is always possible to make something smaller. For
example all transformations could be removed (but then, that gives an
increase in required significant digits for coordinates) or remove all
path comands except cubic beziers, (tradeoff being more complex
content generation, larger files and lower quality) or only have the
polygon command and no path. Or remove all text and just draw pictures
of the characters.

SSL> If the W3C's not interested in such things, that's fine, but they
SSL> shouldn't be surprised that people may laugh at claims that SVG
SSL> Tiny is actually 'tiny'.

I am never surprised when you laugh. Nor did I say W3C was not
interested in a smaller profile in future (although there is a
downside to more profiles as well, of course).

I disagree though with your veiled assertion that SVG Tiny is way too
big and unimplementable.

>>Yes; TinyLine runs on Zaurus (but in J2ME not SE).

SSL> I suspect we could waste days discussing various J2ME profiles,
SSL> PersonalJava, etc., but it's not that exciting.

True, but since you bring it up ....

SSL>  The Zaurus does appear
SSL> to run more than the MIDP J2ME devices are capable of, but whether it's
SSL> running the deprecated PersonalJava or the new J2ME Personal Profile
SSL> isn't clear from Sharp's site.

Neither. The TinyLine site mentions two JVMs that have bee used to run
it.

-- 
 Chris                            mailto:chris@w3.org

Received on Thursday, 14 November 2002 10:12:50 UTC