Hello all.I noticed this past weekend that using LogScale with a model that contains a 0.0 will cause the WAxis object to asymptotically extend the axis toward zero, which in operation makes it exhaust system memory, and probably crash eventually. I saw this under Linux and Mac OS X.
Attached is a patch that modifies ExtremesIterator to consider the AxisScale, and for WAxis' render range to pad the extrema relative to the log value for LogScale mode for better consistency in the size of the padding. The change still means the automatically determined axis minimum is the model's next largest value. This shifts the issue away from the presence of a zero to just the available precision of the model, which I think would be acceptable. I'm not sure if this is the intended direction, but does cover the bug.
Any comments or advice on this? Thanks. -- Joe Garcia Instrumentation Technologist UW Lidar Group Univeristy of Wisconsin - Madison
WAxis_Log0asymptotic.patch
Description: Binary data
------------------------------------------------------------------------------ Register Now & Save for Velocity, the Web Performance & Operations Conference from O'Reilly Media. Velocity features a full day of expert-led, hands-on workshops and two days of sessions from industry leaders in dedicated Performance & Operations tracks. Use code vel09scf and Save an extra 15% before 5/3. http://p.sf.net/sfu/velocityconf
_______________________________________________ witty-interest mailing list witty-interest@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/witty-interest
