Open GoogleCodeExporter opened 9 years ago
There appears to be an issue related to the fact that we use blocking to wait
for the AWT thread to finish touching our data structures.
Specifically, as it is written, with a Glass Pane, we would ask for one pane to
be displayed, two panes get calls for repaint, and we would need to wait for
both of them to be updated before we could continue.
Will both of them always get calls for repainting? Maybe. Will the order they
get called be consistent? No idea.
Now, this whole process gets a whole lot simpler if the AWT thread had its own
copy of the AWT grid that was totally separate from the Terminal grid. Instead
of tweaking data structures, we'd be adding commands in to a queue that would
be processed at paint-time.
More than that, the command-based structure could be seen as a first-step
before we get our own display update protocol (a la xterm, ansi, etc).
So this is slipping.
Original comment by yam...@gmail.com
on 6 Jul 2012 at 8:51
Original comment by yam...@gmail.com
on 6 Jul 2012 at 9:03
Original issue reported on code.google.com by
yam...@gmail.com
on 6 Jul 2012 at 1:32