Software FX Posted January 16, 2002 Report Posted January 16, 2002 This is a good one. Even though it seems like magic it makes sense. The floating point overflow is generated by Chart FX not VB, CHART_HIDDEN can not be assigned to an XValue (shame on us for not documenting this), however Chart FX dos not show any message, the Floating Point library however, keeps a flag of the last exception therefore the first floating point operation is performed in VB after the Chart FX overflow causes VB to display the error message. The first time this happened to me I though I was going crazy but then after some research I found out what was happening. -- FP Software FX, Inc.
User (Legacy) Posted April 13, 2002 Author Report Posted April 13, 2002 Hi, Thought I'd share this little info titbit that would have been handy to me about 2 months ago: - Don't ever set the xvalue() to CHART_HIDDEN I did it "to be sure" and "cause i was a lazy" at the time, and have discovered that it causes overflows later in execution at fairly unrelated bits of code such as: dim test as double dim test2 as double test = test2 ' overflow! Unbelievable as it seems, killing the line that set the xvalue stops the overflow. Hope it helps someone Paul
User (Legacy) Posted April 15, 2002 Author Report Posted April 15, 2002 Ah so thats why it happens... It really bugged me and I had no idea wtf was going on, especially since the line the error would occur on would be in a completely seperate part of the app! Maybe Microsoft should put a little kb article about how these sort of "bugs" are handled by VB (ie confusingly - it should clear the flag before it performs the ops!). Thanks, Paul
Recommended Posts
Archived
This topic is now archived and is closed to further replies.