Opera stealing focus (attn Eirik Byrkjeflot Anonsen)
TimC
About the opera-stealing-focus-under-Linux problem mentioned
previously (eg
http://groups.google.com.au/group/opera.linux/browse_frm/thread/31525d201df6e784/35d69f9b87541610?lnk=st&q=opera+timc+focus#35d69f9b87541610
). I'm a little surprised it's not bugging more people - I don't
think it requires a focus-follows-mouse window manager to trigger the
bug. For example, if you open the preferences window, change
something and save, you've got quite a few seconds to focus another
window, while opera sits there busily spinning away. As soon as it
gets around to redrawing the browser, it'll finish redrawing, then
steal the focus from whatever you had focussed in the meantime.
I just ran xtrace over opera, and it is certainly calling focus
related calls - but without seeing the source, I have no way of
knowing whether these are legimate calls (eg, to check who has the
current focus), or whether opera is "claiming" to be a pager
illegitimately (it doesn't help that I'm not an X programmer).
003:>:0xb8db:32: Reply to TranslateCoordinates: same-screen=true(0x01) child=0x02001428 dst-x=312 dst-y=500
003:<:b8dc: 12: Request(42): SetInputFocus revert-to=None(0x00) focus=0x04800a72 time=0x4f00b623
003:<:b8dd: 4: Request(43): GetInputFocus
003:>:0xb8dd:32: Reply to GetInputFocus: revert-to=Parent(0x02) focus=0x04800212
003:>:b8dd: Event PropertyNotify(28) window=0x0000004e atom=0x1ea("_NET_ACTIVE_WINDOW") time=0x4f00b629 state=NewValue(0x00)
003:<:b8de: 12: Request(42): SetInputFocus revert-to=Parent(0x02) focus=0x04800229 time=CurrentTime(0x00000000)
003:<:b8df: 4: Request(43): GetInputFocus
003:>:b8df: Event FocusOut(10) detail=Pointer(0x05) event=0x04800256 mode=Normal(0x00)
Do you mind doing a recursive grep over the opera source tree and
making sure calls involving XSetInputFocus, _NET_ACTIVE_WINDOW,
gdk_window_focus, WM_TAKE_FOCUS etc are all sane? Most of those
strings appear in the binary, but they could just be compiled in and
inactive. I don't believe the calls are being made from any plugins
-- they're not in the plugin wrapper, and I can 'killall
operapluginwrapper-ia32-linux', and the focus still keeps on getting
stolen (hence it's not flashplayer etc).
There may or may not be relevant information here
http://mail.gnome.org/archives/wm-spec-list/2007-March/msg00000.html
http://mail.gnome.org/archives/wm-spec-list/2007-March/msg00001.html
http://www.mail-archive.com/fvwm-w...@lists.math.uh.edu/msg14996.html
--
TimC
When the revolution comes, we'll need a longer wall. -- Tom De Mulder
Eirik Byrkjeflot Anonsen
> Hi Eirik or others,
>
> About the opera-stealing-focus-under-Linux problem mentioned
> previously (eg
> http://groups.google.com.au/group/opera.linux/browse_frm/thread/31525d201df6e784/35d69f9b87541610?lnk=st&q=opera+timc+focus#35d69f9b87541610
> ). I'm a little surprised it's not bugging more people - I don't
> think it requires a focus-follows-mouse window manager to trigger the
> bug. For example, if you open the preferences window, change
> something and save, you've got quite a few seconds to focus another
> window, while opera sits there busily spinning away. As soon as it
> gets around to redrawing the browser, it'll finish redrawing, then
> steal the focus from whatever you had focussed in the meantime.
I haven't noticed this, but I haven't really tried to trigger it
either.
>
> I just ran xtrace over opera, and it is certainly calling focus
> related calls - but without seeing the source, I have no way of
> knowing whether these are legimate calls (eg, to check who has the
> current focus), or whether opera is "claiming" to be a pager
> illegitimately (it doesn't help that I'm not an X programmer).
>
> 003:>:0xb8db:32: Reply to TranslateCoordinates: same-screen=true(0x01) child=0x02001428 dst-x=312 dst-y=500
> 003:<:b8dc: 12: Request(42): SetInputFocus revert-to=None(0x00) focus=0x04800a72 time=0x4f00b623
> 003:<:b8dd: 4: Request(43): GetInputFocus
> 003:>:0xb8dd:32: Reply to GetInputFocus: revert-to=Parent(0x02) focus=0x04800212
> 003:>:b8dd: Event PropertyNotify(28) window=0x0000004e atom=0x1ea("_NET_ACTIVE_WINDOW") time=0x4f00b629 state=NewValue(0x00)
> 003:<:b8de: 12: Request(42): SetInputFocus revert-to=Parent(0x02) focus=0x04800229 time=CurrentTime(0x00000000)
> 003:<:b8df: 4: Request(43): GetInputFocus
> 003:>:b8df: Event FocusOut(10) detail=Pointer(0x05) event=0x04800256 mode=Normal(0x00)
One thing that could be helpful is if you could identify which windows
the focus is being set to (0x04800a72 and 0x04800229 in the above
example). Maybe also which windows the GetInput messages returns
could be interesting in some cases.
Of course, I'm not actually working on the desktop browser these days,
so I probably won't have time to do much about it. I'll add this
information to your bug report, though.
>
>
> Do you mind doing a recursive grep over the opera source tree and
> making sure calls involving XSetInputFocus, _NET_ACTIVE_WINDOW,
> gdk_window_focus, WM_TAKE_FOCUS etc are all sane? Most of those
> strings appear in the binary, but they could just be compiled in and
> inactive. I don't believe the calls are being made from any plugins
> -- they're not in the plugin wrapper, and I can 'killall
> operapluginwrapper-ia32-linux', and the focus still keeps on getting
> stolen (hence it's not flashplayer etc).
There seems to be far more calls to XSetInputFocus in opera than there
used to be. Although I'm not sure all that code is actually in use...
eirik
Eirik Byrkjeflot Anonsen
> TimC <tcon...@no.spam.accepted.here-astro.swin.edu.au> writes:
>
>> Hi Eirik or others,
>>
>> About the opera-stealing-focus-under-Linux problem mentioned
>> previously (eg
>> http://groups.google.com.au/group/opera.linux/browse_frm/thread/31525d201df6e784/35d69f9b87541610?lnk=st&q=opera+timc+focus#35d69f9b87541610
>> ). I'm a little surprised it's not bugging more people - I don't
>> think it requires a focus-follows-mouse window manager to trigger the
>> bug. For example, if you open the preferences window, change
>> something and save, you've got quite a few seconds to focus another
>> window, while opera sits there busily spinning away. As soon as it
>> gets around to redrawing the browser, it'll finish redrawing, then
>> steal the focus from whatever you had focussed in the meantime.
>
> I haven't noticed this, but I haven't really tried to trigger it
> either.
Ok, I see. We are doing bad stuff with XSetInputFocus. We are
calling it in some places with time set to CurrentTime, and sometimes
that call will be delayed significantly. I've added that information
to the bug report. Hopefully Someone will Do Something About It in a
not too distant future :)
eirik
TimC
was almost, but not quite, entirely unlike tea:
> Ok, I see. We are doing bad stuff with XSetInputFocus. We are
> calling it in some places with time set to CurrentTime, and sometimes
> that call will be delayed significantly. I've added that information
> to the bug report. Hopefully Someone will Do Something About It in a
> not too distant future :)
Ah, thanks for that.
As someone not even working on the project anymore, you've done more
in a single post than I ever got from reporting bugs to the Other Folk
:)
Cheers.
--
TimC
Cult: (n) a small, unpopular religion.
Religion: (n) a large, popular cult.
Eirik Byrkjeflot Anonsen
> On 2008-10-24, Eirik Byrkjeflot Anonsen (aka Bruce)
> was almost, but not quite, entirely unlike tea:
>> Ok, I see. We are doing bad stuff with XSetInputFocus. We are
>> calling it in some places with time set to CurrentTime, and sometimes
>> that call will be delayed significantly. I've added that information
>> to the bug report. Hopefully Someone will Do Something About It in a
>> not too distant future :)
>
> Ah, thanks for that.
>
> As someone not even working on the project anymore, you've done more
> in a single post than I ever got from reporting bugs to the Other Folk
> :)
>
Yes, it is a common complaint (even internally) that our bug system is
largely a black hole for external users. Unfortunately, it isn't easy
to fix, as we want to be able to discuss "secrets" in the bug reports.
The biggest "secret" being our source code, of course :)
You can be sure that bug reports from our users are extremely
valuable. We have people who look through every bug report to figure
out what the reported problem really is and how important it is to do
something about it.
eirik
TimC
was almost, but not quite, entirely unlike tea:
>> Ah, thanks for that.
>>
>> As someone not even working on the project anymore, you've done more
>> in a single post than I ever got from reporting bugs to the Other Folk
>> :)
>>
>
> Yes, it is a common complaint (even internally) that our bug system is
> largely a black hole for external users. Unfortunately, it isn't easy
> to fix, as we want to be able to discuss "secrets" in the bug reports.
> The biggest "secret" being our source code, of course :)
>
> You can be sure that bug reports from our users are extremely
> valuable. We have people who look through every bug report to figure
> out what the reported problem really is and how important it is to do
> something about it.
Wups. I was ambiguous to which Other Folk I was referring to. I was
actually talking about the big open source fire breathing Lizard :)
(in particular, 8 year old open bugs in the bugzilla that present
serious usability issues that are still open. At least their URL
history database isn't made of bogosorting treacle anymore).
If it will still be useful, I do actually intend to get around to
doing a xtrace with the details you asked for, but for some reason my
weekends keep on getting stolen by recalictrant servers that are
protesting about being threatened with decomissioning.
--
TimC
There's only one way to better your memory. Everytime you forget
something, poke your brain with a stick. That'll teach it. -- Ash
Eirik Byrkjeflot Anonsen
> On 2008-10-27, Eirik Byrkjeflot Anonsen (aka Bruce)
> was almost, but not quite, entirely unlike tea:
>> TimC <tcon...@no.spam.accepted.here-astro.swin.edu.au> writes:
>>> Ah, thanks for that.
>>>
>>> As someone not even working on the project anymore, you've done more
>>> in a single post than I ever got from reporting bugs to the Other Folk
>>> :)
>>>
>>
>> Yes, it is a common complaint (even internally) that our bug system is
>> largely a black hole for external users. Unfortunately, it isn't easy
>> to fix, as we want to be able to discuss "secrets" in the bug reports.
>> The biggest "secret" being our source code, of course :)
>>
>> You can be sure that bug reports from our users are extremely
>> valuable. We have people who look through every bug report to figure
>> out what the reported problem really is and how important it is to do
>> something about it.
>
> Wups. I was ambiguous to which Other Folk I was referring to. I was
> actually talking about the big open source fire breathing Lizard :)
> (in particular, 8 year old open bugs in the bugzilla that present
> serious usability issues that are still open. At least their URL
> history database isn't made of bogosorting treacle anymore).
Ah. I guess every software project has its biggest challenge in
"issue management".
Oh well, it gives me the opportunity to say to anyone who happens to
listen: Please make bug reports for any problems you are seeing. And
thanks to everyone who has done so. Yes, you'll get minimal feedback,
but it really does increase the probability that the bug will be fixed
(and sooner rather than later).
> If it will still be useful, I do actually intend to get around to
> doing a xtrace with the details you asked for, but for some reason my
> weekends keep on getting stolen by recalictrant servers that are
> protesting about being threatened with decomissioning.
I suspect it isn't all that important. It is clear that at least some
of those XSetInputFocus calls can easily trigger the problem.
eirik
Jorgen Grahn
...
> Yes, it is a common complaint (even internally) that our bug system is
> largely a black hole for external users. Unfortunately, it isn't easy
> to fix, as we want to be able to discuss "secrets" in the bug reports.
> The biggest "secret" being our source code, of course :)
>
> You can be sure that bug reports from our users are extremely
> valuable. We have people who look through every bug report to figure
> out what the reported problem really is and how important it is to do
> something about it.
This is comforting to know (especially about the internal complaints;
it shows that people at Opera have a clue).
Yes, it's damaging to Opera and to its users. I know that *I* don't
file as many bug reports as I should/could, because of ... well,
"black hole" is a good metaphor.
/Jorgen
--
// Jorgen Grahn <grahn@ Ph'nglui mglw'nafh Cthulhu
\X/ snipabacken.se> R'lyeh wgah'nagl fhtagn!
Eirik Byrkjeflot Anonsen
> On Mon, 27 Oct 2008 09:28:03 +0100, Eirik Byrkjeflot Anonsen <ei...@opera.com> wrote:
> ...
>> Yes, it is a common complaint (even internally) that our bug system is
>> largely a black hole for external users. Unfortunately, it isn't easy
>> to fix, as we want to be able to discuss "secrets" in the bug reports.
>> The biggest "secret" being our source code, of course :)
>>
>> You can be sure that bug reports from our users are extremely
>> valuable. We have people who look through every bug report to figure
>> out what the reported problem really is and how important it is to do
>> something about it.
>
> This is comforting to know (especially about the internal complaints;
> it shows that people at Opera have a clue).
It's always surprising to find that other people have a minimal amount
of clue :)
> Yes, it's damaging to Opera and to its users. I know that *I* don't
> file as many bug reports as I should/could, because of ... well,
> "black hole" is a good metaphor.
Yes. And our users don't look through our bug reports to see if that
little annoyance is already reported and maybe add a little extra
information. Or keep their own bug reports up to date with further
discoveries. etc. etc.
Some of that is helped by our mailing lists, newsgroups and forums,
which I believe are monitored by some of our people. And they will
create bug reports for the problems that are discussed. Also, with
bug reports with a valid reporter email address we can ask the
original reporter for further information. But it would be very nice
if we could at least report back to the reporter "we think we fixed
it", "that's not a bug", "that's not important right now" or whatever
else we choose to do with the bug.
eirik
TimC
was almost, but not quite, entirely unlike tea:
>> This is comforting to know (especially about the internal complaints;
>> it shows that people at Opera have a clue).
>
> It's always surprising to find that other people have a minimal amount
> of clue :)
Heh. But if you start off assuming they don't, you're never
disappointed; only pleasantly surprised :)
>> Yes, it's damaging to Opera and to its users. I know that *I* don't
>> file as many bug reports as I should/could, because of ... well,
>> "black hole" is a good metaphor.
>
> Yes. And our users don't look through our bug reports to see if that
> little annoyance is already reported and maybe add a little extra
> information.
Is there a publicly accessible database of bugs somewhere? I hadn't
found one last time I looked.
Oh, I take it I can get an account here
https://bugs.opera.com/secure/Dashboard.jspa then login, giving me
search capability? For some reason, I had only seen the "for
employees" part before.
> Or keep their own bug reports up to date with further
> discoveries. etc. etc.
I'm a bit guilty of that. Not having found where to search for bugs
online means I've lost a couple I submitted from various email
addresses over the years.
--
TimC
Usage: fortune -P [-f] -a [xsz] Q: file [rKe9] -v6[+] file1 ...
Eirik Byrkjeflot Anonsen
> On 2008-10-29, Eirik Byrkjeflot Anonsen (aka Bruce)
> was almost, but not quite, entirely unlike tea:
>
>>> Yes, it's damaging to Opera and to its users. I know that *I* don't
>>> file as many bug reports as I should/could, because of ... well,
>>> "black hole" is a good metaphor.
>>
>> Yes. And our users don't look through our bug reports to see if that
>> little annoyance is already reported and maybe add a little extra
>> information.
>
> Is there a publicly accessible database of bugs somewhere? I hadn't
> found one last time I looked.
Not that I know of. Which is exactly what I meant :)
(People don't look through our bugs because they can't)
>
> Oh, I take it I can get an account here
> https://bugs.opera.com/secure/Dashboard.jspa then login, giving me
> search capability? For some reason, I had only seen the "for
> employees" part before.
Well, we've just replaced our bug tracking system. Maybe it is now
possible to get some information, at least on one's own bugs. If you
try it out, please let me know how badly it works :)
>
>> Or keep their own bug reports up to date with further
>> discoveries. etc. etc.
>
> I'm a bit guilty of that. Not having found where to search for bugs
> online means I've lost a couple I submitted from various email
> addresses over the years.
Of course. And since you never get any feedback on what we're doing
on a bug, you can't provide useful replies to that. A more open bug
system would have helped a lot on all of that.
> Usage: fortune -P [-f] -a [xsz] Q: file [rKe9] -v6[+] file1 ...
Hey, my favorite signature :)
eirik
Janek Kozicki
The focus stealing in opera is annoying thing for me too. When I
switch to desktop which has an opera window on it, I must be overly
catious about to where I type, but nevertheless I quite often end up
typing my email (or typing some xterm commands) into opera window
instead of into thunderbird window (or xterm). Which is ... you know -
annoying. Because I need to retype all that again, after I notice that
it went to wrong window. And as a side bonus opera sometimes does
weird stuff due to all those keystrokes.
And I'm using focus-follows mouse if that matters. I got used to
throwing mouse at any window that I currently need. And opera got used
to steal my effort back. Nevertheless opera is still a good browser
and I'm using it more often than firefox, but much less often than
galeon :)
Eirik Byrkjeflot Anonsen
If my analysis is correct, the focus stealing happens only if you
focus opera while opera is busy doing something else (like painting
the screen). As soon as opera finishes what it's doing, it will steal
the focus.
eirik
Alan Hoyle
> If my analysis is correct, the focus stealing happens only if you
> focus opera while opera is busy doing something else (like painting
> the screen). As soon as opera finishes what it's doing, it will steal
> the focus.
I am also experiencing this issue on Opera 9.62 on Ubuntu 8.04
desktop. Your analysis's conclusions and the previous poster's
experience mesh with mine pretty well.
-alan
--
Alan Hoyle - al...@unc.edu - http://www.alanhoyle.com/
Chris Wage
> If my analysis is correct, the focus stealing happens only if you
> focus opera while opera is busy doing something else (like painting
> the screen). As soon as opera finishes what it's doing, it will steal
> the focus.
I am seeing the exact same thing, and it's becoming more and more
annoying.. After using Opera for nearly 10 years, I'm considering
switching to Firefox purely because of this.. I can barely use my
desktop because opera snags focus at every opportunity..
Any chance this will be fixed any time soon?
Eirik Byrkjeflot Anonsen
I don't know. I haven't been working on the desktop browser for a
long time, so I don't know much about what they're doing these days.
eirik
Janek Kozicki
Please ping them if you can. Several of them, if possible. And point
to this thread. In fact I'm using opera much less than I'd like to,
just because of this bug.
GP lisper
> Chris Wage <cmw...@gmail.com> writes:
>
>> On Nov 17, 4:03 am, Eirik Byrkjeflot Anonsen <ei...@opera.com> wrote:
>>> If my analysis is correct, the focus stealing happens only if you
>>> focus opera while opera is busy doing something else (like painting
>>> the screen). As soon as opera finishes what it's doing, it will steal
>>> the focus.
>>
>> I am seeing the exact same thing, and it's becoming more and more
>> annoying.. After using Opera for nearly 10 years, I'm considering
>> switching to Firefox purely because of this.. I can barely use my
>> desktop because opera snags focus at every opportunity..
That's interesting. I used to have this problem, complained about it
once and in the next opera upgrade it was gone and I haven't seen it
again. Using opera 9.6x now, I test an upgrade on a laptop before
putting it on the main machine.
All I did was overhaul everything once. Upgraded java, cleaned up the
plugin problem, maybe even wiped opera for a fresh install. Perhaps
you are using 'mouse hints' ?
On plugins, I just created an /opt/plugins directory and made sure
that only /opt/opera and it were in opera search path. Java (which
probably has nothing to do with it) was cleared of multiple versions
and left with 1.6.
I do see a mouse stealing problem with a Tk program, but only at
startup.
--
"Most programmers use this on-line documentation nearly all of the
time, and thereby avoid the need to handle bulky manuals and perform
the translation from barbarous tongues." CMU CL User Manual
** Posted from http://www.teranews.com **
Alan Hoyle
> I do see a mouse stealing problem with a Tk program, but only at
> startup.
On my Ubuntu 8.04 box with O 9.62, I use "focus follows mouse" and get
focus following problems even without screen redraw. E.g. I have two
windows about 50 px apart with Opera in the background behind them and
move the mouse quickly between them, Opera doesn't relinquish the
focus even though the mouse is hovering over one or the other of the
app windows. If I go more slowly, the focus follows correctly, but
with a fast move, Opera retains focus maybe 1/5th of the time.
Here is an ASCII art picture of the basic arrangement:
+-------+ +----------+
+-| app1 |---| app2 |
| | | | |
| | | | |
| +-------+ +----------+
| |
| OPERA |
| |
+----------------------+
So far, in practice and in limited experimentation, I haven't noticed
any other app that does this. Only Opera seems to hold on to focus in
this.
Eirik Byrkjeflot Anonsen
Great example. Yes, this would maximize your problems. The problem
isn't that opera "holds on to" the focus. It is that opera moves
focus to one of its own windows when it gets a "focus in" event. So
what happens is that when moving from app1 to app2, first you hit the
opera window, the window manager sets focus to opera, and X sends a
"focus in" event to opera. Then you get to app2, and the window
manager sets focus to app2. Now opera finally gets around to acting
on the "focus in" event, and moves focus back to one of opera's
windows.
eirik
Alan Hoyle
Well, I'm glad that my regular window setup is a "great example" that
would "maximize [my] problems."
Count me as someone who would vote STRONGLY in favor of fixing this
bug ASAP. If writing in more CAPITAL LETTERS would help it get
resolved, let me know. ;-)
Seriously though: this issue is incredibly annoying to me. I hope it
gets fixed soon.
-a
Alan Hoyle
>> Alan Hoyle <al...@unc.edu> writes:
>>> +-------+ +----------+
>>> +-| app1 |---| app2 |
>>> | | | | |
>>> | | | | |
>>> | +-------+ +----------+
>>> | |
>>> | OPERA |
>>> | |
>>> +----------------------+
> Seriously though: this issue is incredibly annoying to me. I hope it
> gets fixed soon.
I suppose this isn't a surprise, but the update to 9.63 on Ubuntu 8.04
doesn't seem to fix the problem. Does the Opera 10.0 alpha fix it?
-alan
Chris Wage
> I suppose this isn't a surprise, but the update to 9.63 on Ubuntu 8.04
> doesn't seem to fix the problem. Does the Opera 10.0 alpha fix it?
Nope.. Tried opera 10 and it had the same problem.. back to using
firefox here
Janek Kozicki
> On Mon, 01 Dec 2008 15:15:53, Alan Hoyle wrote:
> >> Alan Hoyle <al...@unc.edu> writes:
> >>> +-------+ +----------+
> >>> +-| app1 |---| app2 |
> >>> | | | | |
> >>> | | | | |
> >>> | +-------+ +----------+
> >>> | |
> >>> | OPERA |
> >>> | |
> >>> +----------------------+
> > Seriously though: this issue is incredibly annoying to me. I hope it
> > gets fixed soon.
>
> I suppose this isn't a surprise, but the update to 9.63 on Ubuntu 8.04
> doesn't seem to fix the problem. Does the Opera 10.0 alpha fix it?
Exactly! That's the best way to reproduce this ERROR in opera focus
handling. How long do we need to wait, until opera 11.0? I'll stop
using opera by that time :( It's a great browser but c'mon...
Can opera developers add this to official bug reports database? I
couldn't find the opera's bugzilla although I tried really hard.
Eirik Byrkjeflot Anonsen
The bug is in our bug tracking system. We even think we know what
causes it. A proper fix may well be impossible, but we can do much
better than we do today. Unfortunately, it seems that the people who
could fix it are busy with "more important" issues. I can't tell you
what that is, simply because I don't know.
Our bug tracking system has one major flaw: The public is not allowed
to see our bugs. The main reason for this is that it allows us to say
anything we want to in the bug reports, even things that we consider
"secret" (in particular, descriptions of our source code). There is
no doubt that having (most of) our bugs generally accessible would
have been a great benefit, both to us and our users.
Rest assured that all reported bugs are examined, even if the system
feels like a black hole. Our bug system is at http://bugs.opera.com/
eirik
Janek Kozicki
thank you. Please try to convince them... again.
For everyone - it seems like I found a temporary solution to this
problem. In my window manager configuration I've set opera window to
never get focus. No matter what happens opera never gets focus, even
if I click on it madly. Not all window managers support this kind of
configuration, though... I'm using sawfish if you wanted to ask.
Alan Hoyle
>> >> >>> +-| app1 ??|---| app2 ?? ?? |
>> >> >>> | | ?? ?? ?? | ?? | ?? ?? ?? ?? ??|
>> >> >>> | | ?? ?? ?? | ?? | ?? ?? ?? ?? ??|
>> >> >>> | +-------+ ?? +----------+
>> >> >>> | ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ??|
>> >> >>> | ?? ?? ??OPERA ?? ?? ?? ?? ?? |
>> >> >>> | ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ??|
>> >> >>> +----------------------+
> thank you. Please try to convince them... again.
> For everyone - it seems like I found a temporary solution to this
> problem. In my window manager configuration I've set opera window to
> never get focus. No matter what happens opera never gets focus, even
> if I click on it madly. Not all window managers support this kind of
> configuration, though... I'm using sawfish if you wanted to ask.
Hear hear! Please get this fixed.
Janek Kozicki
forgot to turn off "focusable" afterwards. Now - this has never
happened to me before. Opera managed to steal focus, after I switched
to another viewport. Now not a single window on current desktop was
focused, but opera had focus which was on different desktop. I could
check that, because the text that I was typing on irc to my friends,
was in opera window (when I switched back, to check if maybe that's
where my typing goes).
Alan Hoyle
> Janek Kozicki <cos...@gmail.com> writes:
>> On Dec 16 2008, 2:40??pm, Alan Hoyle <al...@unc.edu> wrote:
>>> On Mon, 01 Dec 2008 15:15:53, Alan Hoyle wrote:
>>> >> Alan Hoyle <al...@unc.edu> writes:
>>> >>> ?? +-------+ ?? +----------+
>>> >>> +-| app1 ??|---| app2 ?? ?? |
>>> >>> | | ?? ?? ?? | ?? | ?? ?? ?? ?? ??|
>>> >>> | | ?? ?? ?? | ?? | ?? ?? ?? ?? ??|
>>> >>> | +-------+ ?? +----------+
>>> >>> | ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ??|
>>> >>> | ?? ?? ??OPERA ?? ?? ?? ?? ?? |
>>> >>> | ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ??|
>>> >>> +----------------------+
>>> > Seriously though: this issue is incredibly annoying to me. ??I hope it
>>> > gets fixed soon.
>>>
>>> I suppose this isn't a surprise, but the update to 9.63 on Ubuntu 8.04
>>> doesn't seem to fix the problem. ??Does the Opera 10.0 alpha fix it?
>>
>> Exactly! That's the best way to reproduce this ERROR in opera focus
>> handling. How long do we need to wait, until opera 11.0? I'll stop
>> using opera by that time :( It's a great browser but c'mon...
>>
>> Can opera developers add this to official bug reports database? I
>> couldn't find the opera's bugzilla although I tried really hard.
> The bug is in our bug tracking system. We even think we know what
> causes it. A proper fix may well be impossible, but we can do much
> better than we do today. Unfortunately, it seems that the people who
> could fix it are busy with "more important" issues. I can't tell you
> what that is, simply because I don't know.
I don't know if it's the case or not, but I recently updated to 9.64
and it seems like this problem has gotten worse.
please, please, please get them to fix this.