Closed stewa closed 11 years ago
Milestone: Someday. Label: #new. What's next? A reviewer should examine this issue.
+1
Milestone: Someday. Vote: 1. Label: #new. What's next? A reviewer should examine this issue.
+bug +AppKit
Milestone: Someday. Vote: 1. Labels: #needs-review, #new, AppKit, bug. What's next? This issue is pending an architectural or implementation design decision and should be discussed or voted on.
We can't really accept this without a test app that shows before/after the change.
Milestone: Someday. Vote: 1. Labels: #needs-reduction, #needs-review, AppKit, bug. What's next?
Tests/Manual/ArrayController1: 1) select a row start editing a text field 2) select another row => changes are lost without the fix and commited with the fix applied.
I cannot reproduce. If I follow your steps, the field is updated without any problem
i can reproduce it:
Oh we were talking about the textfield, not the row itself. Ok I can reproduce then.
Milestone: Someday. Vote: 1. Labels: #accepted, #needs-reduction, AppKit, bug. What's next? A minimal test app should be created which demonstrates the concern of this issue in isolation.
Merged. Thanks!
Milestone: Someday. Vote: 1. Labels: #fixed, AppKit, bug. What's next? This issue is considered successfully resolved.
We finally have some concern about this one. It makes the CPTableViewTest.j failing.
Let me copy paste conversation about it here :
primalmotion: this one is the problem
https://github.com/cappuccino/cappuccino/commit/19804720af58d9a4617729b626563a21b8e0eaf5
it makes sense why the test is failing
I think this patch is correct, we just need to update the test
14:07 aljungberg
Hmm.
Well that depends.
Isn't it possible to double click on items when a table is not the first responder?
14:08 primalmotion
it would then become the first responder at the first click?
14:09 aljungberg
Not if whatever was the first responder didn't resign first responder, right?
Also I wonder what happens if the window is a cppanel which can't become the key window. Do controls become the first responder there?
14:10 primalmotion
hum, and in that case are we able to double click ?
14:10 aljungberg
I don't know
So we don't know :)
Milestone: Someday. Vote: 1. Labels: #accepted, AppKit, bug. What's next? A reviewer should examine this issue.
Until we have an answer on this I am going to fix the test.
I have confirmed that in Cocoa on OS X 10.8 that on a table which refuses to become the first responder it's none the less possible to:
I tested this by creating a subclass of NSTableView
which responds NO
to acceptsFirstResponder
and which emits a logging message in becomeFirstResponder
.
So this change is not entirely incorrect. It's true that the table does try to become the first responder when clicked, but if this fails it still handles mouse actions as normal.
(Note: oddly, triggering editing of a table row seems to make the table view the first responder even if it returns NO for acceptsFirstResponder
.)
-#fixed
milestone=0.9.6 assignee=aljungberg
Assignee: aljungberg. Milestone: 0.9.6. Vote: 1. Labels: #accepted, AppKit, bug. What's next? A reviewer should examine this issue.
This pull request as described in the title has been reverted. One part of the change still remains though: now the table tries to make itself the first responder earlier than before. Maybe this will resolve the original problem in #1435, although there doesn't seem to be any test for this issue.
+#fixed
Assignee: aljungberg. Milestone: 0.9.6. Vote: 1. Labels: #fixed, AppKit, bug. What's next? This issue is considered successfully resolved.
Fix for issue #1435