[IPython-user] ipython -p numeric problem

Fernando Perez fperez at colorado.edu
Sat Mar 29 23:01:46 CST 2003

Quoting Jack Moffitt <jack at xiph.org>:

> > It is caused by the fact that I'm trying to enable mouse support, but 
> > unfortunately I think only the CVS gnuplot has this.  In a sense it's 
> > forcing a bit the users's hand, but I've left it in because Gnuplot with
> > mouse is so incredibly better than without, that it's worth for users 
> > the pain of moving over to CVS code.  
> It really depends on how soon the Gnuplot CVS will make it to a real
> release?  I have no idea what their timelines are.

Well, slow as far as I can see.

> > Unfortunately I don't know of a way to trap the error to inform the 
> > users of all this more gracefully.  Ideally I'd check for this in some 
> > manner and give a nice message, but because of the way ipython 
> > communicates with gnuplot, I've failed so far.
> Can't you just run gnuplot --version and check the version string?

The problem is that I'm not really sure in which patch level the mouse 
support appeared.  But to be honest, I haven't put much effort into this at 
all.  It may also be possible to start gnuplot with the mouse option on from 
the command line, and check for an error code.  So few people seem to use 
gnuplot that I've been a bit lazy with this.  The ones that use it seem to 
be happy to get the mouse-enhanced version.   But I should really cook up a 
cleaner solution.



More information about the IPython-user mailing list