For those who might miss <pattern> in FF take a look at
http://srufaculty.sru.edu/david.dailey/svg/pattern.svg . I had it
working before in IE/ASV but had to clean up a few things to make it
proper. I'd be interested in knowing if it displays anywhere else
(than FF and IE/ASV). It takes a lot of horsepower to run the
scripted animation. I suspect older equipment will choke.
Basically, it just populates the screen with replications of a single
group (offset a bit for a hex fill). Is there any way to do an
animated non-rectangular pattern using SVG <pattern>. I figure that
any periodic tiling of the plane can be rotated so as to find a
rectangular clip which can then be replicated (consider, for example,
the rectangle that completely encloses any three mutually adjacent
hexagons -- the prototiles -- in the regular hex-tiling... then that
rectangle tiles seamlessly -- I think that generalizes to all and
only the periodic tilings.) The problem occurs when we try to animate
activity within the nonrectangular tile so as to simulate activity
within the prototiles. I don't think the attributes associated with
<pattern> (including "overflow") will allow for activities in the
accidentally contiguous chunks which make up the prototiles. Does
this result in a suggestion for SVG k.n? Probably not... while the
periodic tilings of the plane by individual regular polygons include
only squares, triangles and hexagons, each of which might have their
own enthusiasts, it hardly makes sense to generalize to
non-rectangles unless one implements a more comprehensive theory of
tiles which might include all the periodic tilings as well as those
nonperiodic and aperiodic tilings which are recursively enumerable.
That would be a bit of an undertaking. For those who are still
reading this rather awkward paragraph, I am reminded of another
question: Consider a set of tiles which tile the plane, each of which
is filled with a gradient (either linear or radial -- and the radial
gradients need not be centered at the center of the tile)... are
there seamless tilings (gradient quilts) which do not expose harsh
edges of gradients other than the trivial ones?
By the trivial ones, I mean the linear gradient with infinitely many
stop points, the radial gradient with infinitely many stop points,
and the "raindrop" composite gradient where radial gradients of the form
<radialGradient cx cy r>
<stop offset<r stop-color=A/>
[<stop/>]*
<stop offset=r stop-color=B/>
</radialGradient>
are centered in tiles that are then dropped onto a monochromatic
background of color B at infinitely many points (cx,cy), and far
enough apart to not collide .
The fill pattern associated with <feTurbulence> is one such pattern
which is seamless and nontrivial, but it is not obvious to me that it
may be tiled in some sort of piecewise fashion with only these two
types of gradients. It seems like some topologists must have
discussed this when filling orientable surfaces with vectors ...
maybe Stokes theorem settles the issue ?.. it has been too many years
since I did this kind of analysis.
The fact that the edge of a tile filled with a linear gradient may be
laid seamlessly next to the edge of a tile filled with a radial
gradient, (see for example
http://srufaculty.sru.edu/david.dailey/svg/tritile.svg ) gives me a
sense that indeed nondeterministic seamless quilts may be built from
these two gradient types, but I have also been suspicious that there
are seamless quilts which would demand a richer set of primitive
gradients. If so then how to characterize the set of such primitives?
DD
------------------------ Yahoo! Groups Sponsor --------------------~-->
Most low income homes are not online. Make a difference this holiday season!
http://us.click.yahoo.com/5UeCyC/BWHMAA/TtwFAA/1U_rlB/TM
--------------------------------------------------------------------~->
-----
To unsubscribe send a message to: [EMAIL PROTECTED]
-or-
visit http://groups.yahoo.com/group/svg-developers and click "edit my
membership"
----
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/svg-developers/
<*> To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
