On Jul 22, 2008, at 3:13 PM, Martin Wierschin wrote:
Hi Gerd,
[..]
In your case I would worry about the call sequence when line
fragments are calculated. There's no guarantee that each fragment
will only be swept out once and in an order sorted by vertical
position. Especially if you have noncontiguous layout enabled. I
think it would be much safer to have your implementation of
lineFragmentRectForProposedRect:etc: do some simple geometric tests,
eg:
I realize that my example made certain unsafe assumptions, and your
way is a better one. I had a similar method in place at one time, but
I ran into some confusion with flipped coordinate systems that has
since been solved. I made a note to revisit that.
However in the few cases where I was able to reproduce the problem,
the call order was as expected and identical to the cases when the
problem does not occur (and even a hardcoded fixed offset for all
lines was not honored). I simply meant to offer anecdotal evidence
that under Leopard there appears to have been (or still is) a rarely
triggered circumstance where text rendering behaves oddly. And in a
hard to figure out problem every little shred of information might
lead to a solution...
Thanks for your suggestion!
Gerd
_______________________________________________
Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)
Please do not post admin requests or moderator comments to the list.
Contact the moderators at cocoa-dev-admins(at)lists.apple.com
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com
This email sent to [EMAIL PROTECTED]
