Issue 30258 - API: forms.OComboBoxModel::com::sun::star::form::XUpdateBroadcaster
Summary: API: forms.OComboBoxModel::com::sun::star::form::XUpdateBroadcaster
Status: CLOSED FIXED
Alias: None
Product: App Dev
Classification: Unclassified
Component: api (show other issues)
Version: 3.3.0 or older (OOo)
Hardware: All Linux, all
: P2 Trivial
Target Milestone: ---
Assignee: chne
QA Contact: issues@api
URL:
Keywords:
: 30546 (view as issue list)
Depends on:
Blocks:
 
Reported: 2004-06-15 14:53 UTC by chne
Modified: 2013-02-24 21:09 UTC (History)
1 user (show)

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


Attachments
patches to fix this bug, not extensively tested (10.23 KB, application/x-compressed)
2004-06-29 11:52 UTC, Frank Schönheit
no flags Details

Note You need to log in before you can comment on or make changes to this issue.
Description chne 2004-06-15 14:53:50 UTC
while executing the test for XUpdateBroadcaster for object OComboBoxModel the
office hangs. This happens on Linux and Solaris. The concerning calls of test are:

oForm.load()
oObj.getPropertyValue("Text")

This code snipped was called after a listener is added. While I debug this
snipped sometimes it runs without failure. If I try then again the test all
works fine. A fresh started office is a garant to reproduce the freeze.
Just execute
%SOLARENV%\bin\checkapi -o
fforms.OComboBoxModel::com::sun::star::form::XUpdateBroadcaster
to reproduce the behaviour
Comment 1 Frank Schönheit 2004-06-15 16:15:39 UTC
accepting
Comment 2 Frank Schönheit 2004-06-22 09:35:41 UTC
*** Issue 30546 has been marked as a duplicate of this issue. ***
Comment 3 Frank Schönheit 2004-06-22 09:37:01 UTC
*** Issue 30546 has been marked as a duplicate of this issue. ***
Comment 4 Frank Schönheit 2004-06-29 11:51:06 UTC
note 1: The bug was unrevealed by a change in Writer which made it into the
master recently. Since another side effeft of this change was pretty annoying,
issue 30295 was submitted and fixed. Now when the fix for issue 30295 makes it
into the master, then this bug here will vanish magically :(

note 2: The deeper reason is the NumberFormatter service implementation relying
on the SolarMutex for thread safety, which is pretty bad. Fixing this issue here
*also* involves fixing the NumberFormatter to rely on an own mutex. However,
this is somewhat risky - it's easy to implement (I did a test implementation
which fixed the bug here), but it's difficult to overlook the consequences.

I'm going to attach patches which completely fix this bug here, including a
patch to the NumberFormatter service object(s).
Comment 5 Frank Schönheit 2004-06-29 11:52:14 UTC
Created attachment 16181 [details]
patches to fix this bug, not extensively tested
Comment 6 Frank Schönheit 2004-10-18 15:56:24 UTC
fixed in CWS dba18
Comment 7 Frank Schönheit 2004-11-08 12:59:30 UTC
re-opening for reassignment to QA
Comment 8 Frank Schönheit 2004-11-08 13:17:30 UTC
fs->cn: please verify in CWS dba18
Comment 9 Frank Schönheit 2004-11-08 13:19:40 UTC
fs->cn: please verify in CWS dba18
Comment 10 Frank Schönheit 2004-11-08 13:20:24 UTC
fixed in CWS dba18
Comment 11 chne 2004-11-08 15:12:15 UTC
verified in dba18
Comment 12 chne 2004-12-08 20:37:55 UTC
tested in src680_m65 -> close