RFR: 8284521: Write an automated regression test for RFE 4371575
Phil Race
prr at openjdk.java.net
Thu Apr 7 21:04:34 UTC 2022
On Thu, 7 Apr 2022 11:08:38 GMT, Manukumar V S <mvs at openjdk.org> wrote:
> Write an automated regression test for [JDK-4371575](https://bugs.openjdk.java.net/browse/JDK-4371575)
>
> Issue:
> As part of the Merlin focus project, JComponent.setRequestFocusEnabled was
> deprecated and its implementation was changed to map exactly to the new
> method Component.setFocusable. Scott believes that the old behavior may be
> preferable. He would like to be able to specify that a Component is focusable,
> and should receive focus during keyboard traversal, but that it should not
> automatically take focus when the user clicks on it with the mouse.
>
> We are concerned that the accessibility team would be against this behavior,
> and this also seems like more of a PLAF issue. Nevertheless, we should look
> into it before beta ships.
>
> Fix:
> After some discussion we decided that the best balance of the old and new would be to make this an advisory property. This property will not be synonymous with focusable. Instead our mouse listeners will check this property before requesting focus. This provides as closely as possible the old behavior, while allowing people to use the new focusable property if they don't want a component focusable at all.
>
> Testing:
> Tested in mach5, 10 times with all the 3 available platform like macosx, windows and linux and got all Pass.
Before I look at the test, can you get rid of the pointless directory 4371575
A directory might be needed if you had a lot of files for the test - but you don't !
And I really don't like naming directories OR tests with the bugid. Someone can find
that out from the @bug tag.
I know you've pushed a bunch of bugs already that follow this pattern but update all the "in progress" ones to NOT create pointless extra folders.
-------------
Changes requested by prr (Reviewer).
PR: https://git.openjdk.java.net/jdk/pull/8143
More information about the client-libs-dev
mailing list