Bug 233054

Summary: Only show taks from current screen doesn't work correctly
Product: [Unmaintained] plasma4 Reporter: András Manţia <amantia>
Component: widget-taskbarAssignee: Plasma Bugs List <plasma-bugs>
Status: RESOLVED FIXED    
Severity: normal CC: aseigo, claim, sebas, thijs22nospam
Priority: NOR    
Version: 4.6.0   
Target Milestone: ---   
Platform: Compiled Sources   
OS: Linux   
Latest Commit: Version Fixed In:
Sentry Crash Report:

Description András Manţia 2010-04-02 11:29:21 UTC
Version:            (using Devel)
OS:                Linux
Installed from:    Compiled sources

I have a dual monitor setup, with taskbars on both screens. The taskbar on the second screen is configured to "Only show taks from current screen".  If I move an application from screen 1 to screen 2, it will not appear in the second taskbar. If I start an app on that screen, it appears, but moving to screen1 doesn't make it disappear from the task 2. 
Strangely, if now I modify the taskbar's configuration (e.g disable this setting and enabling it again), it will work as expected until plasma-desktop is restarted.
Comment 1 András Manţia 2010-04-02 11:30:13 UTC
Forgot to write, this worked fine a few weeks ago.
Comment 2 Sebastian Kügler 2010-05-19 18:17:51 UTC
Same problem here, one taskbar per screen with dualhead doesn't split the tasks correctly per taskbar.

Whoever picks this up, ping me if you need someone to test.
Comment 3 Sebastian Kügler 2010-06-10 15:04:10 UTC
Ok, found the reason for this:

Apparently, the "current screen" of the taskbar isn't changed when you move the taskbar (for example in a panel) from one screen to the other. It will still think it's on the previous screen, and hence mess up which tasks to show. Removing the taskbar and adding a new one in the same location "solves" this.

This is especially annoying when you switch a display on and off, and your taskbar moves screen.
Comment 4 Sebastian Kügler 2010-06-10 15:22:05 UTC
The problem is that the "show only tasks from current screen"
 gets messed up when you move the taskbar from one screen to another. This seems to be related to the Plasma::ScreenConstraint. This one is emitted when an applet is moved from one screen to another. It does *NOT* get emitted (or reach the tasks applet, anyway) when a panel is moved between screens, so the applet shows the wrong tasks after moving.
Comment 5 Aaron J. Seigo 2010-06-10 18:27:07 UTC
fixed in r1136747 :) thanks, sebas, for figuring the root cause of this one out.
Comment 6 Sebastian Kügler 2010-06-10 22:19:50 UTC
Fix confirmed. Yay. :-)
Comment 7 András Manţia 2010-06-19 20:29:40 UTC
Sorry to say, it is not working. :( I have a taskbar on the second screen, start an app (assistant in the particular case). It is full-size on the second screen, but it isn't in the taskbar. Strange enough, starting konqueror or dolphin on that screen make those to appear in the taskbar. And now assistant as well. It is unpredictable.
But if I drag them from that screen to the first one, they are still in the taskbar. :(
Comment 8 Waldemar Szostak 2011-01-02 18:20:22 UTC
I had this problem and now it's gone/it's fixed in Gentoo KDE 4.5.4.
Comment 9 Waldemar Szostak 2011-02-18 23:49:42 UTC
Heh it's back in 4.6.0 - it's not working again :P

[complaing mode]Generally, this 4.6.0 seems to be the worst KDE release ever - most crucial features like addressbook stop working there (yeah.. created a bug report for that :/)[/complaing mode]
Comment 10 Aaron J. Seigo 2011-02-19 00:35:49 UTC
"[complaing mode]"

this is not the right forum for this drek. i really don't need to add "random off-topic complaining from random people" to the burden of going through my bugs.kde.org emails. 

highly unimpressive, highly unhelpful and pretty demotivating. for everyone's sake, take it elsewhere.

the great irony here is that the bug was evidently _never_ fixed so your "it works, now it doesn't" is just more BS.  the trigger for the problem was never found. the one person who bothered to dig into it (sebas) found an issue, which i was then able to fix in fairly short order, though it seems to be different from the original reporter's and yours. until the triggering factor is found, there's little we're going to be able to do about it.
Comment 11 Waldemar Szostak 2011-02-19 01:17:58 UTC
Perhaps I indeed should have kept that for myself. My apologies. But - it's a fact that something went quite wrong with 4.6.0. And this was also highly unimpressive, unhelpful and demotivating for the users (like myself). Besides I wasn't speaking all that serious. If you can't take any humour, then it's your problem.

Secondly, please don't patronise me and act as if you were the only person around who cares, because I'm also concerned about KDE and its quality. I appreciate all the people that spend their time contributing to it as well.

Finally, if my previous comment was inappropriate, then what should your description of my report as bullshit be called, huh? I'm really not interested in some personal dispute here, but think for a second about what you write too before jumping to conclusions about others.


To the point - I don't know 'why' but the function in question didn't work, later at some point it did and it doesn't again. This is a user's observation.
Comment 12 Aaron J. Seigo 2011-02-19 02:12:25 UTC
"it's a fact that something went quite wrong with 4.6.0."

for you, perhaps. many others have been quite happy with it. are there bugs? yep. many fixed for 4.6.1. in the case of plasma, many of those are because i try to keep up with and read all of these bug reports as they come in. and guess what makes me not want to do that?

"And this was also highly unimpressive, unhelpful and demotivating for the users (like myself)."

so the answer is to complain about bugs in other software on an unrelated bug report so you can spread that around? no one here who needs that.

if you want to commiserate with others about how shitty things are, there's the web forums and other more appropriate places. bugs.kde.org, however, is part of our work area in which you are ultimately a guest. when useful feedback is provided, then it's a good thing. otherwise, it's less than useless, it's a negative. if you think that it shouldn't/doesn't matter, look at this example:

http://www.kdedevelopers.org/node/4131

it happens all too frequently, and that's why the kind of stunt you pulled gets to me so much. even as a user there are responsibilities when participating in an open creative environment such as this.

"Besides I wasn't speaking all that serious."

that is such a cop-out; nothing in what you wrote communicates anything other than it being said in all seriousness. there was nothing "funny" in it, right down to the emoticon.

"don't patronise me and act as if you were the only person around who cares,"

sharing how your behaviour impacts those of us doing the work you appreciate is not patronizing, it's stating facts. just because you don't like it doesn't mean it doesn't need to be said.

"because I'm also concerned about KDE and its quality."

great; but this isn't the right forum for sharing what you did in the way you did. the only impact that has is to help make the rest of us feel like bugs.kde.org is not a useful tool. that's not a good thing, as it is one of the few structure feedback systems we have.

so if you care about the quality of KDE, how about we start with the quality of participation on bugs.kde.org and try to prevent it becoming a PITA for those of us trying to improve KDE with our efforts.

"I appreciate all the people that spend their time contributing to it as well."

you can demonstrate this by helping keep our work environment productive and non-aggravating.

"then what should your description of my report as bullshit be called, huh?"

you're right, i shouldn't have called it BS. what you wrote was the usual "best guess" of a user that turns out to be wrong. there is certainly a qualitative emotional difference there. 

so why did i use the less tactful observation? in this case because sitting in my email inbox was a report that something had become re-broken/regressed in 4.6.0.  because i care about such things i came here to see what the history of the report was and see if it could be re-fixed. upon reading the full report it turns out wasn't re-broken at all, just never gotten fully fixed. and for that, i got a random bit of unnecessary, off-topic complaint tagged on. that hardly does anything good for one's motivation. what a horrible "reward" for caring enough to track these reports! it took me from a so-so place in my day to a generally less-than-happy one. great, so now we're both making each other annoyed. no one's winning there.

and it's easily avoided by not dropping complaints (let alone off-topic ones) that can have no useful result just because "we can".

"I don't know 'why' but the function in question didn't work,
later at some point it did and it doesn't again. This is a user's observation."

that's all that was needed. in future, please keep it to the on-topic observations and take your editorializing elsewhere.
Comment 13 Thijs 2012-01-19 14:36:28 UTC
Works perfectly smoothly for me in 4.7.4. I'd like to close; does anyone disagree?
Comment 14 Myriam Schweingruber 2012-05-16 18:44:50 UTC
Closing based on comment #13. Please feel free to reopen this report if you can still reproduce this with KDE 4.8.3 or later.