summaryrefslogtreecommitdiffstats
path: root/src/extension/execution-env.h (follow)
Commit message (Collapse)AuthorAgeFilesLines
* r18220@shi: ted | 2008-02-29 13:18:55 -0800Ted Gould2008-02-291-26/+53
| | | | | | | | | | | | | | | | | | | | | | | | Okay, sadly I'm not keeping the version history because I'm not convenced that SVK will do it right. One mega-patch, but that's life. Reshuffle the exection-env and prefdialog code so that the state machines aren't intertwines, which fixes a whole host of bugs with them. I think the behavior is correct now. Make it so that the effects can count how many preferences they have to determine if the dialog should be shown (fix above). Once this code was written it was easy to make it show an ellipsis on the verb if there is a dialog or not. This involved removing ellipsis from those effects that had it hard coded. Make it so that the parameters know that their command line options are going into a list. They don't have to acknowledge it, but they can, and specifically notebook does and handles it differently. This should fix the notebooks on Win32, but doesn't apparently completely. Change the script extension on windows to use pythonw instead of python so that the command line doesn't appear all the time. (bzr r4908)
* Add virtual to a lot of the destructors. Note: perhaps it will have to be ↵Johan B. C. Engelen2007-11-041-1/+1
| | | | | reverted for some files to keep them C-compatible. (bzr r4024)
* r16892@shi: ted | 2007-10-29 21:33:09 -0700Ted Gould2007-10-301-1/+10
| | | | | | Okay, so now the caches get into the execution environment and can be passed around to the effects. Now it's a matter of implementing the caches in the drivers. (bzr r3978)
* r16474@tres: ted | 2007-08-31 21:37:33 -0700Ted Gould2007-09-011-0/+1
| | | | | Changes that SVK borked. (bzr r3656)
* r16430@tres: ted | 2007-08-30 20:32:23 -0700Ted Gould2007-09-011-1/+3
| | | | | | Adding in a timer so that the preferences aren't acted on immediately. This should give the illusion of more interactivity. (bzr r3650)
* r16406@tres: ted | 2007-08-29 20:14:55 -0700Ted Gould2007-09-011-1/+3
| | | | | | Reroute the way that the preferences signal changes so that pinning works as expected. Fixes a segmentation fault. (bzr r3645)
* r16403@tres: ted | 2007-08-29 09:10:45 -0700Ted Gould2007-09-011-1/+2
| | | | | | Fixed a tricky segmentation fault. Multithreading is hard when you don't have threads :) (bzr r3642)
* r16395@tres: ted | 2007-08-27 19:41:32 -0700Ted Gould2007-09-011-0/+1
| | | | | | This is a good stopping point. I think things are working... more testing. (bzr r3640)
* r16340@tres: ted | 2007-08-20 19:02:50 -0700Ted Gould2007-09-011-1/+2
| | | | | Pinned is still not working, somewhat usable state. Worth checking in. (bzr r3639)
* r16339@tres: ted | 2007-08-19 19:47:57 -0700Ted Gould2007-09-011-0/+2
| | | | | | Live preview toggling is working. Pinning, not as much. But getting there. (bzr r3638)
* r16220@tres: ted | 2007-08-15 19:43:11 -0700Ted Gould2007-09-011-1/+2
| | | | | | | Okay, so now things are a little more connected. The widgets are parameters and they have signals. Work left is to actually USE those signals ;) (bzr r3634)
* r15704@tres: ted | 2007-07-02 16:03:34 -0700Ted Gould2007-07-031-0/+70
Split out the execution environment object. (bzr r3163)