| Activity bindings vs model processors [message #892255] |
Wed, 27 June 2012 10:04  |
Eclipse User |
|
|
|
In an Eclipse 3.x app I worked on, we used activity bindings as part of a user authorization system to hide UI elements (menu items, view parts, perspectives) that an account did not have access to. The app was composed of bundles which each contributed "permissions" and activity bindings, and a global service tied permission enablement to the activity system.
In Eclipse 4.x, I read about the Model Processor which could crawl the application model and make changes.
Is it a recommended or even practical approach to use model processors in e4 as a replacement to activity bindings from 3.x, or do activity bindings still work as they did?
Is it a good practice to have a model processor react live to the application, or should its lifecycle simply be on application startup? How about having a model processor register listeners to the model that react?
Where can I look for more information about better use of model processors? I've checked the help.eclipse.org and a tutorial, but did not find info about it on wiki.eclipse.org
|
|
|
|
| Re: Activity bindings vs model processors [message #892529 is a reply to message #892271] |
Thu, 28 June 2012 10:35  |
| Eclipse User |
|
|
|
We often use model addons. We often use these to translate the model into the runtime, and as part of the startup they register listeners that can later react to model changes and update the runtime.
We've been using processors to update the model before it goes live ... but I'm less sure of the implications of adding a listener during them, as they have no dispose lifecycle.
It might be that the LifeCycleHandler is a better fit for this kind of use.
PW
|
|
|
Powered by
FUDForum. Page generated in 0.04009 seconds
Home