Jean-Christophe Rioux wrote:
>
> If so, how could the Adobe plugin do a better job?
I would challenge whether it's really a 'better job' given the SVG
you are sending it. You are of course seeing a 'better looking'
result,
I won't challenge that.
> The image is even smaller on Adobe plugin and still I get a better
> definition!
> I think that's a bug and I will certainly not edit every SVG so that it
> display correctly.
What we are discussing then is viewer-specific behavior, right ? - not
the SVG spec. If a line width is 2.5% of a pixel, I would expect it
to
nearly go away when using aliasing, and to be jaggy if not.
> There should be a way to do it by some coding (and no I will not modify the
> XML internally in my code)
Of course code changes can change this behavior to be more nearly that
of the Adobe viewer. But...
There is a downside to hacking Batik, to make it mimic the behavior of
another viewer, rather than asking what the 'correct' behavior
should be (according to the standard, or from what one might logically
expect independent of the viewer).
While this sort of thing crops up frequently when one
implementor of a standard 'gets there first', I think in this case it
is pretty questionable. Many other SVG files could be quite
negatively impacted by this change, for instance the apparent darkness
of a fine sub-pixel grid (for instance, used as a crosshatch pattern)
would get much darker as you zoomed out, which would ruin an image
which
relied on crosshatched shading.
Do you see my point?
Changing the XML need not be too painful, it could be done via XSLT
for
instance. If the images needed 'compatibility' with Adobe (which
exhibits
what I'd call broken behavior in this case) then the SVG could use
CSS
styles I believe.
I hope that we can find a workable solution for you that is either
based
in the SVG spec or does not negatively impact other SVG files.
Best regards,
-Bill
> Jean-Christophe Rioux
--
--------------
Bill Haneman
Gnome Accessibility / Batik SVG Toolkit
Sun Microsystems Ireland
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
