On 4/6/06, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> Hi Seth,
>
> "Seth Tager" <[EMAIL PROTECTED]> wrote on 04/06/2006 06:13:13 PM:
>
> > ...I'm not sure I understand your
> > statement about 'creating classes to hold the data between DOM and
> > GVT'. I think you're suggesting that one way to decouple them would be
> > to create an intermediate set of classes to hold the data and then
> > pass the data objects off to some object in gvt that would build the
> > tree. Another way would be to require bridge clients (like gvt) to
> > implement an interface, and then have the clients build the tree
> > themselves as the document is parsed.
>
>    In the end I don't think there is a large difference between the
> two.  Perhaps you save some intermediate memory but if I give you
> an java.awt.Path with the data what is the difference from calling
> 20 different methods (like PathParser does)?  You can pull the data
> out of the path object at least as easily as implement the PathHandler
> interface...

I think you lose the potential to handle dynamic content. I'm not
concerned about dynamic content at the moment,  but if I'm thinking
about a general solution I figure I should work toward one that would
support it in theory.


> > Here's what I've done so far and where I'm headed. Since I have
> > limited batik experience, please let me know if you think I'm
> > looking for trouble.
>
>    Well, I'm not sure what you code looks like so it's a little hard
> to say.  I'll say that in general the most challenging parts of batik
> are the 'use' element, and 'text' (text is _really_ hard).  Also of
> course efficient dynamic updates are tricky (dependency tracking).

I did notice a lot of code related to dynamic updates, and I see a lot
of classes and complexity around text, but I appreciate the
confirmation of where the rough spots might be. I didn't notice anything
related to the 'use' element, so I'll look for that.


> > ... It seems to me that a framework
> > for building arbitrary class trees from svg would do much to promote
> > svg as it would lower the barrier to entry for 3rd party projects that
> > wanted to read svg and transform it to their internal data format.
>
>    I don't know for certain but it seems to me that this interface would
> be huge (filters, clips, patterns, gradients, fill, stroke, use) I
> would expect that many people would prefer to implement the Graphics2D
> interface to import SVG data - it may be useful for importing other data
> as well.  However it can be a bit lossy.

I hadn't thought of that, and it sounds like the easier way to go for
some situations, but the lossy aspect is important. I assume things
like filters get transformed into drawImage calls, which is what you
mean by lossy, no?


>    I really don't think I follow you here.  Do you care to share your
> work so far?

I'd be happy to send it to you. I tried posting it to the list about a
week ago but it was rejected as being too large (300k). I cc'd you
directly, but I didn't get a reply so I'm not sure if you got it. Can
I send you a copy offline?

Seth

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to