Issue 11847 - combo-box from ParagraphStyles frame fires events wrong
Summary: combo-box from ParagraphStyles frame fires events wrong
Status: CLOSED FIXED
Alias: None
Product: ui
Classification: Code
Component: ui (show other issues)
Version: current
Hardware: PC Linux, all
: P3 Trivial (vote)
Target Milestone: OOo 1.1 Beta2
Assignee: nospam4obr
QA Contact: issues@ui
URL:
Keywords: accessibility
Depends on:
Blocks:
 
Reported: 2003-02-26 10:42 UTC by Unknown
Modified: 2003-05-12 08:46 UTC (History)
6 users (show)

See Also:
Issue Type: DEFECT
Latest Confirmation in: ---
Developer Difficulty: ---


Attachments
simple-at inspecting StarOffice screen-shot (149.94 KB, image/jpeg)
2003-02-26 10:48 UTC, Unknown
no flags Details

Note You need to log in before you can comment on or make changes to this issue.
Description Unknown 2003-02-26 10:42:08 UTC
OVERVIEW DESCRIPTION :
Using gnome-2.0 stack from 3 Feb, java-access-bridge, at-spi, atk, gail, from
today's CVS (HEAD) , java 2 sdk 1.4.1_01, and StarOffice 6.1 Beta iterim release
06 Feb 2003, from
ftp://oem2:SmwkIdh55a@ftpmaster.stardiv.de
I found inconsistency in generating events for items within a list that belongs
to a combo-box.

STEPS TO REPRODUCE :
- start StarOffice 
- put focus on Paragraph Styles frame, on the only combo-box existing there
- start festival_server in the background (this is not mandatory, but it would
help you in having acoustic feedback from simple-at test program)
- start simple-at test progam (from at-spi/test source directory)
- switch back to ParagraphStyles and navigate through the items of combo's list

ACTUAL RESULTS:
You will notice that events are reported NOT for the focused object, BUT for the
next object in list (I think this is an index problem). The only time when focus
is reported correctly is when the focus is on the last item in list (in this
case Conditional Styles).

EXPECTED RESULTS:
I would expect to receive focus event for the real object that is focused: for
example,
for Special Styles object if this is focused, not for HTML styles.

ADDITIONAL INFORMATION:
I will attach a screen-shot with simple-at inspecting events from StarOffice to
prove my point.
Comment 1 Unknown 2003-02-26 10:48:07 UTC
Created attachment 4910 [details]
simple-at inspecting StarOffice screen-shot
Comment 2 Frank Schönheit 2003-02-26 13:12:58 UTC
AdiDascal, thanks for testing this sensible area of OOo/SO.

I cc'ed some people which may know more about this - if anybody feels
he is wrong on the cc'list: sorry, please remove yourself :)
Comment 3 nospam4obr 2003-02-26 13:43:33 UTC
Darren, does the JABG map ACTIVE_DESCENDANT events to FOCUS events for
ComboBoxes ? I think in the UAA, the list object has the focus all the
time.

The index problem could still be a problem on the OOo side though ..
Comment 4 nospam4obr 2003-02-27 14:10:34 UTC
Confirmed in m5s1.
Comment 5 nospam4obr 2003-02-27 14:14:30 UTC
Peter,

I can see in my logfiles that the wrong object is reported as new
value in the ACTIVE_DESCENDANT event fired by this ComboBox. I assume
there is only one implementation for ComboBoxes, so could you please
have a look at this ?

BTW: this happens also on Windows.
Comment 6 nospam4obr 2003-02-27 14:39:55 UTC
Actually this combo box instance seems to be the only that has this
problem - at least the only one I have found so far. Hmm, strange.
Comment 7 pb 2003-02-28 12:40:32 UTC
Accepted
Comment 8 pb 2003-03-17 10:25:54 UTC
Fixed in cws os7: sfx2/source/dialog/templdlg.cxx 1.31.2.1.28.1
Comment 9 pb 2003-04-04 07:21:46 UTC
.
Comment 10 pb 2003-04-04 07:22:39 UTC
PB -> OBR: please verify with your logfiles.
Comment 11 pb 2003-04-04 07:23:00 UTC
Fixed in cws os7
Comment 12 nospam4obr 2003-04-04 08:09:03 UTC
We now have to problem that the document frame claims to be active
after one selects the combo box, which causes the combo box list to
loose the focus. This bug is fixed in the CWS vcl07.

The active descendant events now look o.k., so I set this task to
verified.
Comment 13 nospam4obr 2003-05-12 08:46:38 UTC
I have verified in the build m13s2 that the fix has been successfully
integrated.

Our current bug handling policy requires this issue to be closed
before beta2 can officially be released. Feel free to re-open it if it
doesn't work for you as expected.