Summary: | [EditorMgmt] Need a new API IEditorPart.isSaveAllowed() to control Save menu | ||
---|---|---|---|
Product: | [Eclipse Project] Platform | Reporter: | Tetsuro Asahata <asahata> |
Component: | UI | Assignee: | Platform UI Triaged <platform-ui-triaged> |
Status: | NEW --- | QA Contact: | |
Severity: | enhancement | ||
Priority: | P3 | CC: | mistria, tomasz.zarna |
Version: | 2.0 | Keywords: | helpwanted, investigate |
Target Milestone: | --- | ||
Hardware: | All | ||
OS: | All | ||
Whiteboard: |
Description
Tetsuro Asahata
2002-04-29 11:10:24 EDT
In this case, since the encoding is part of the editor's state, changing the output encoding should dirty the editor. Will reconsider this for v3 though. Changing dirty state will confuse user, since user only changed global preference and not each editor's content. In this case editor should not change dirty state. Also output encoding is just an example. Editor may update their own meta data such as GENERATOR in HTML, editor's private information which usually user don't care or don't change, and so on... However, I agree this is a V3 item. Reopen for investigation Moving Dougs bugs What if the editor is opened on a read-only input (e.g. CVS revision), should changing the encoding dirty the editor as well? Observed in N20090101-2000. Remy is now responsible for watching the [EditorMgmt] component area. Usually, on all editors I can figure out, Save is always enabled independently of the state of the file being edited, even if it contains "wrong" data. It's actually important for user to be able to save even invalid content, as a milestone. So despite the idea is interesting, I question how users will really like it or not, compared to established workflows (of Eclipse IDE and all editors in general). It seems to me that in most case, error in files have to be reported in the editor and the decision of whether to save or not should entirely belong to the user. |