I want to figure out what's the problem.. By blocking you have in mind that after you call invoke, following canvas commands must be saved in a buffer, before they will actually applyed on a canvas? When you get image data, canvas buffer must be flushed, right? Is it problem local for IE or for all browsers?
Some code snippets may be helpful..
> Deli
--
Evgeny Burzak <buz...@gmail.com>
Hi Evgeny,
What I have in mind is much simpler (Its possible I dont understand the
current behaviour properly).
1. I draw something on the canvas (Text usually)
2. I call getImageData to record the pixels that are now not blank
3. I save the not blank pixels in a cache when getImageData returns
4. I search for a blank area to draw the next item using the data
returned in 2
The sequence above is what I get using ff, chrome and opera with an
actual canvas, but when I use a debugger and console logs using fxcanvas
I see.
1.
2.
4.
1.
2.
4.
3.
3.
Resulting in the code drawing more text on areas it thinks are blank as
the cache has not been updated.
You can see the actual code here
http://blog.odonnell.nu/static/wurdle/wurdle.html
The important file is http://blog.odonnell.nu/static/wurdle/js/wurdle.js
The function update_word_cache at line 417 has both the fxcanvas and
normal canvas functionality in it.
It generates a sort of tag cloud given text,
easiest way to run it is to put a twitter username in the textfield
Thanks for you help
Sean
I really appreciate the help, this is just a hobby project, so no panic.
All the best, and thanks for fxcanvas
Sean