Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.
Bug 324956 - [Model] Integrate Dialogs/Wizards into the application model
Summary: [Model] Integrate Dialogs/Wizards into the application model
Status: CLOSED DUPLICATE of bug 421462
Alias: None
Product: Platform
Classification: Eclipse Project
Component: UI (show other bugs)
Version: 4.2   Edit
Hardware: PC Mac OS X - Carbon (unsup.)
: P3 normal (vote)
Target Milestone: ---   Edit
Assignee: Platform-UI-Inbox CLA
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-09-10 08:30 EDT by Thomas Schindl CLA
Modified: 2014-08-12 11:05 EDT (History)
8 users (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Thomas Schindl CLA 2010-09-10 08:30:30 EDT
We should think if we should discuss the integration of Wizards/Dialogs into our Application Model which could e.g. help with KeyBindings in there, ... .
Comment 1 Eric Moffatt CLA 2010-09-13 14:41:24 EDT
Tom, how do you see this working? 

How deep would you expect this model to go (all the way through SWT widgets or something at a higher level)?
Comment 2 Thomas Schindl CLA 2010-09-13 18:13:30 EDT
I'd see this at a very very highlevel:

Dialog
 1 Part

Wizard
 1 .. n Part

but I really haven't thought about it in detail but when it is part of the model we could assign keybindings, ... to it.
Comment 3 Brian de Alwis CLA 2010-09-14 08:50:43 EDT
Would we model *all* dialogs (e.g., MessageBox.openConfirm()), or just higher-level dialogs?  Would we try to capture trays?

We might want to allow dialogs to have trim, toolbars, and perspectives (e.g., like many MacOS X preference dialogs).
Comment 4 Thomas Schindl CLA 2010-09-14 08:57:22 EDT
(In reply to comment #3)
> Would we model *all* dialogs (e.g., MessageBox.openConfirm()), or just
> higher-level dialogs?  Would we try to capture trays?

Only highlevel dialogs. I think about the various WizardDialogs we currently have.

> 
> We might want to allow dialogs to have trim, toolbars, and perspectives (e.g.,
> like many MacOS X preference dialogs).


Well then they are Workbench Windows not?
Comment 5 Brian de Alwis CLA 2010-09-15 09:06:34 EDT
(In reply to comment #4)
> (In reply to comment #3)
> > We might want to allow dialogs to have trim, toolbars, and perspectives (e.g.,
> > like many MacOS X preference dialogs).
> 
> Well then they are Workbench Windows not?

Good point.  "Dialog" typically implies some kind of modal behaviour, so you're right, those preference dialogs are really windows.
Comment 6 Eric Moffatt CLA 2010-09-20 14:21:31 EDT
We can already demo that you can put parts into wizards by creating a local context and calling createGui directly.

I'm hoping that the various Commands infrastructure discussions will produce a better way to manage the key bindings problems but I can see where having the ability to define the bindings for such a part might be quite useful.

I'm less clear as to whether modeling the high-level dialogs will help very much versus ensuring that the key bindings story is open enough to allow KB's in a dialog...

Miscellaneous observations:

- any part used in a dialog/wizard *must* be multi-instance and cannot be shared (if not then the only choice for the dialog is to 'kidnap' the part from the main ui layout, leaving a hole where it was)

- Parenting the dialog's context is important...this is a KB issue in the sense that do we always want to use the current context as the basis (i.e. Debug Perspective) and use its ActionSet info...or could we make a particular wizard's context be based on a particular perspective (i.e. perhaps the Create Class wizard should always use the Java Perspective's KB set regardless of the currently active perspective)?
Comment 7 Thomas Schindl CLA 2010-12-18 04:07:25 EST
I'll remove this from the 4.1 model evolution - I still think having highlevel dialogs/wizards in the model is benefitial but it needs more time to make up the design
Comment 8 Lars Vogel CLA 2014-08-12 11:05:28 EDT

*** This bug has been marked as a duplicate of bug 421462 ***