<div dir="ltr">Michael Wilber kindly sent me the link: <div> <a href="http://lists.racket-lang.org/users/archive/2012-April/051485.html">http://lists.racket-lang.org/users/archive/2012-April/051485.html</a> and I think that this is not really a good argument for changing the default behavior. A keyword argument that says "create a new eventspace" seems okay, but the docs should explain that this is multi-threading, and if you were to invoke methods of the plot frame without first going to the new eventspace (via queue-callback or something like that), then you'll probably get bad behavior.</div> <div> Robby On Mon, Mar 11, 2013 at 1:55 PM, Eli Barzilay <<a href="mailto:eli@barzilay.org">eli@barzilay.org</a>> wrote: > > A few minutes ago, Robby Findler wrote: > > There is a tension here, but I actually think Racket's default > > strategy is the best one (Matthew and I have discussed this a fair > > amount way back when we were first getting GUI stuff going in > > Racket; Java's strategy is different and leads to race-conditions > > easily, for example.). > > In this case, things are kind of obvious if you know about eventspaces > and about the per-plot eventspace feature -- but since this is a quirk > that only plot is doing, most people don't expect it and get > surprised. > > > > So if plot's function were just something like: > > > > (define (plot-frmae the-args) > > (define f (new frame% ...)) > > ..put stuff in frame.. > > (send f show #t)) > > > > Then just hitting run in DrRacket would show you the multiple plots in > > separate frames fine. At the command-line if you script was > > > > #lang racket > > (require plot) > > (plot-frame ...) > > (plot-frame ...) > > (plot-frame ...) > > > > then that would also work fine (not exiting until you closed all the > > windows). > > +17 if it can work like this. Otherwise, if there's a reason it > can't, then I think that the new-namespace could be available as a > keyword or parameter thing. But the default should be the > no-surprises version. > > > > I can't recall anymore the use-case that let Neil to what he's done > > here, so we probably should revisit that. > > -- > ((lambda (x) (x x)) (lambda (x) (x x))) Eli Barzilay: > <a href="http://barzilay.org/">http://barzilay.org/</a> Maze is Life!</div></div>