a todo list for gui

[lyhlbyl]

High Priority

• [x] connect status, also set the default status as listening
• [x] the text filed in the atomic edit panel is too small and too short
• [x] double click a nested node to "open" it rather than to edit the name
• [x] export graph as a dot file
• [ ] overlapping buttons, need to be in a drop down style when
• [ ] to be able to delete a tactic definition (similarly for gt)
• [x] a locker for the hierarchy inspector which will lock the current window and prevent from displaying other subgraph
• [x] key binding support, e.g., pressing enter to confirm and esc to cancle
• [x] not to mark as changed when changing the current location of the psgraph
• [x] the hierarchy in tree view becomes unreadable, i.e. having problem in fitting in one screen, especially when graph becomes more complicated (see the conv example) need a better representation.
• [ ] a json debug interface
• [x] change the loop shape
• [x] zoom in and zoom out (real)
• [ ] movable edge label
• [ ] curve edge

Low Priority

• [ ] number of branches ? maybe not to show the full number, just show 0 ..n and n+, because the way how sequence is unfolded

A TODO pool

• [ ] command : RSP_EXCEPTION when if _interrupt is false, try backtrack automatically
• [ ] interactive inspection are re-active if user click on some item on the graph when waiting for prover
• [ ] moving the gui to Scala 2.11.8 (latest) current is 2.10.0
• [ ] improve dialog layout

[plebras]

gui now has three connection status, with the following behavior when icon is clicked :

• disconnected --> open connection
• connecting (default) --> interrupt connection request
• connected --> close connection

[ggrov]

Great. Is it easy to understand which state you are in from the icon? How about connected and disconnected as now (one with full red/green colours, the other without), and then connecting we make red and green flash?

[plebras]

at the moment it's simply the same icon with faded fills. I'll have to see if I can easily have something changing the icon regularly to make it flashing.

[ggrov]

at a minimum we at least need 3 different icons if there are 3 different states (maybe just find a standard "connecting" icon for the connecting state if animation is too much work) Having the same icon for 2 different states will not be very user friendly as it will just looks like the tool behaves randomly.

[plebras]

Here are the three icons : They are somewhat different, but I reckon that it may be unusable for users with visual impairments. But usually a connecting state is represented with animated icons, so it would be worth investigating that.

[plebras]

Adding a blinking icon was less difficult than expected. So now when connecting the icon changes every 500 milliseconds between the three icons above.

[plebras]

@lyhlbyl the changes the discussed are commited, turns out I had a method for that already defined.

[plebras]

Regarding tactic definition suppression, I suggest the following :

• introduce a loop to remove definitions of all tactic that users have removed the line, tactic will still be in the model
• print out all tactic and their definitions currently in the model when opening the editor, even the empty definitions (already implemented)
• disable the deletion of tactic when the last node has been deleted
• introduce a "clean" function, where either users can select which unused tactic (i.e. with no nodes) they want to remove, or the gui automatically deletes every unused tactics

[lyhlbyl]

@plebras for the second item, i.e. the text filed in the atomic edit panel is too small and too short

• [ ] allow users to resize the box
• [ ] softwarp the line if it is too long