264325 - Home/End/PgUp/PgDn doesn't work in Input/TextArea/Composer/MailNews on Solaris Xserver

Status: RESOLVED FIXED

(SeaMonkey :: General, defect)

Description

[Ginn Chen]

Use Soalris Xserver,

Input something in Input/TextArea/Composer/MailNews,
Press Home/End/PgUp/PgDn, nothing happens.

It can be reproduced on Mozilla 1.7.3, Mozilla 1.8a5, Firefox 1.0PR

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.3) Gecko/20041014 Firefox/0.10

Comment 1

[Ginn Chen]

a regression of bug 257405

Comment 2

[Ginn Chen]

Attachment: patch

bryner,
I'm not sure if these lines can help bug 257405.

Comment 3

[Roy Lamberton]

In Composer, when you right click on your mouse, the command window [the one
that opens up with all the chices] opens across the page.

It should open directly next to the pointer.

Thanks

Comment 4

[Brian Ryner (not reading)]

(In reply to comment #2)
> Created an attachment (id=162059)
> patch
> 
> bryner,
> I'm not sure if these lines can help bug 257405.
> 

Ginn, can you explain why we don't want to do this mapping?

Comment 5

[Ginn Chen]

(In reply to comment #4)
> Ginn, can you explain why we don't want to do this mapping?

bryner, I'm not quite understand bug 257405.
But these code prevent "inputfields", "textareas" & "editor" binding 
inherit "browser" binding.
See mozilla/content/xbl/builtin/gtk2/platformHTMLBindings.xml
If I copy PGUP/PGDN... bindings from "browser" to "editor" ...
It will also work. (Actually DOMKeyCodeToGdkKeyCode isn't called in this case)

Comment 6

[Brendan Eich [:brendan]]

Ginn, please don't set blocking* flags, those are for project managers to
approve or reject.  I'll return the flags to ? status for now, but I'll help
make sure that this bug gets attention for 1.0.

We need a patch that bryner can review, though, so you may have to give a more
detailed rationale about it. ;-)  Even better if you put the rationale details,
or a URL to persistent docs that explain the patch or the feature it uses, in
the patch itself as a code comment.

/be

Comment 7

[Ginn Chen]

Hi Brendan,

I was thinking bryner changed "the blocking flag" by accident.
Since you had set them at 15:29 PDT, and bryner commented at 15:30 PDT, and set
them back to "?"

Anyway, I leave flags for you.

Comment 8

[Brian Ryner (not reading)]

Ok, I think I understand the problem.  The mappings in nsSunKeycodes are not
two-way... that is, GDK_F29 becomes NS_VK_PAGE_UP when it enters our event
system, but that doesn't mean all NS_VK_PAGE_UP events are GDK_F29.  Since it
looks like this is the case for all of the mappings in that array, your original
patch looks fine.  You might want to add a comment to this effect.

Comment 9

[Ginn Chen]

I've tested it with Xsun and Xorg on Solaris,
both work fine with the patch.

Comment 10

[Brendan Eich [:brendan]]

Ginn, sorry about that -- right you are.  I wish bugzilla would flag those
request mid-air collisions.  Restoring +'d nominations; sounds like this will
make it with no problem.

/be

Comment 11

[Johnny Stenback (:jst)]

Comment on attachment 162059 [details] [diff] [review]
patch

sr=jst

Comment 12

[Brendan Eich [:brendan]]

Comment on attachment 162059 [details] [diff] [review]
patch

a=[e-mail] for 1.7 and aviary branches -- please note checkins into
both here, and set the fixed1.7.x and fixed-aviary1.0 keywords.  Thanks,

/be

Comment 13

[Johnny Stenback (:jst)]

Fix landed on trunk and branches.

Comment 14

[Ginn Chen]

Thank you, bryner, jst and brendan.