Re: MEDM crash when "Plot Array" of sscan detector with large value

[Andrew Johnson <anj@aps.anl.gov>]

Tim,

On Monday 19 July 2010 16:54:46 Tim Mooney wrote:
> Lewis,  Just for completeness, I see this on my system as well.

Are you seeing this on a version of MEDM that is build with XTR/Graph, or with 
SciPlot?  Janet's /usr/local/epics/extensions/bin/<arch>/medm binaries are 
built with XRT/Graph on Solaris and Linux I believe, and if they show the 
problem I think it will increase the pressure for Janet to work on a fix.  We 
need to make sure this gets on her list when she gets back (I think she's on 
vacation at the moment).

> I could get the sscan record to avoid posting PV's that contain
> NaN's, but I have a lot of other stuff that also gets plotted,
> and that also can post NaN's, so my plan is to sit tight until
> we know whether/when MEDM will be fixed.

It certainly makes sense to fix MEDM rather than try to avoid generating bad 
data everywhere.

- Andrew
-- 
The best FOSS code is written to be read by other humans -- Harald Welte