I don't believe that animation can cause "intermittent rounding errors". I can believe that there is a floating point quantization error due to the math being done in the wrong order. Unless we _know_ that unscaling/rescaling is as cheap as two fp operations, it would seem better to just get the math right.
On 2010-03-09, at 18:53, Max Carlson wrote: > Yes, that's true! I didn't check back to see. Unscaling, measuring, > then rescaling is the only reliable way I know of to get an accurate size. > > In LPP-8814 the scale was being animated (stretches/width changing) > which caused an intermittent rounding error when trying to measure using > the stored scale. > > Regards, > Max Carlson > OpenLaszlo.org > > On 3/9/10 7:15 AM, André Bargull wrote: >> So, you're effectively only reverting change r7933, aren't you? >> >> >> On 3/9/2010 1:23 AM, Max Carlson wrote: >>> Change 20100308-maxcarlson-F by maxcarl...@bank on 2010-03-08 16:20:37 >>> PST >>> in /Users/maxcarlson/openlaszlo/trunk-clean >>> for http://svn.openlaszlo.org/openlaszlo/trunk >>> >>> Summary: Make resource size measurement more reliable in Flash 8 >>> >>> Bugs Fixed: resource gets wrong width when stretches equals width >>> >>> Technical Reviewer: ptw >>> QA Reviewer: mdemmon >>> >>> Details: Unscale, measure, and reapply the scale to get the true size. >>> >>> Tests: IWFM. >>> >>> Files: >>> M WEB-INF/lps/lfc/kernel/swf/LzSprite.as >>> >>> Changeset: >>> http://svn.openlaszlo.org/openlaszlo/patches/20100308-maxcarlson-F.tar >>> > > _______________________________________________ > Laszlo-reviews mailing list > [email protected] > http://www.openlaszlo.org/mailman/listinfo/laszlo-reviews
