I complained recently about stacks that suddenly locked up to the point
where Rev had to be killed at the Windows task manager level. I may
have figured out why this happened. What follows are only my thoughts:
I've not actually tested this theory yet.
The first stack I tried to write was an image viewer. One issue was how
to resize the image as the stack was resized by the user. I started out
trying to use the Geometry facilities provided in Rev, but that failed
because the loaded image might have a different aspect ratio than the
designed image space holder. I then tried to write an on resize
handler, to manually resize the image as the stack size changed.
I think what happened was that I forgot to disable the Geometry on the
image, and was both trying to adjust the size manually and have Rev
adjust it automatically. The result may have been repeated or recursive
calls, as the two approaches battled it out endlessly.
My failing to disable the Geometry when I wrote the on Resize handler
is obviously a bug on my part. My only question/comment is to ask
whether the Rev run-time system might have detected an attempt to
manually alter a value that was being maintained automatically, and flag
a run-time error. This would have eliminated much of the mystery, and
all of the agony.
Just a suggestion.
:)
Jon
_______________________________________________
use-revolution mailing list
[email protected]
http://lists.runrev.com/mailman/listinfo/use-revolution
