Re: [Plplot-cvs] CVS update notice (plplot/examples/tcl)

[Maurice L. <mj...@ga...>]

Alan W. Irwin writes:
 > Update of /cvsroot/plplot/plplot/examples/tcl
 > In directory usw-pr-cvs1:/tmp/cvs-serv30944
 > 
 > Modified Files:
 > 	plot.tcl 
 > Log Message:
 > Removed configuration of dp driver that does not even exist.  Also, commented
 > out configuration of xwin and tk drivers which killed pausing since that
 > seemed to generate bad behaviour (any part of X window that was obscured
 > during drawing remained black when it gained focus, plots would not remain
 > on screen if run from scripts).  The net result now of these
 > changes is the plots run fine interactively or from a script using either
 > the xwin or tk drivers.

As an aside, for single plots using the tk driver I still think it's better to
turn off the pause so that control returns immediately to the interpreter,
without having to hit the redundant RETURN in the tk window.  

For the X driver, it's a different matter (I don't know what I was thinking).
Since the X driver is part of the same process, when control is elsewhere
there is no way to process its event loop.  Geoff and I have long discussed
the merits of rewriting the X driver to be in a separate process so that it
can always be processing events, but it would be a significant piece of work
and never really had time for it.

The tk driver communicates to a plserver that is however a separate process
and so it doesn't matter if control is returned to the pltcl -- the plserver
will process events just fine regardless.

The only real problem IMO with turning off pause on just the tk driver is that
it introduces an asymmetry in how plot.tcl behaves for X vs Tk output.  So
it's cool with me to keep your change as-is.

BTW, I notice in README.tcldemos that you are unable to get r.dat to work
with the plot command.  What happens?  'plot r.dat' works fine for me.

-- 
Maurice LeBrun    mj...@ga...