← Back to team overview

unity-design team mailing list archive

Re: Papercut or not? Bug #495403 in One Hundred Paper Cuts: “Do not raise windows or dialogs without user input”

 

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Frederik Nnaji wrote on 30/05/10 15:36:
>...
> On Fri, May 28, 2010 at 12:31, Matthew Paul Thomas <mpt@xxxxxxxxxxxxx
>...
>> So when should the window manager switch from assuming you want a
>> new window focused, to assuming you don't? After five seconds? Ten?
>> Twenty? And how can the window manager, by itself, tell which was the
>> action that resulted in the window eventually opening?
>  
> Let me interject a few questions..
> Focus, Mouse Focus (e.g. for scrolling), Keyboard Focus, raising to the
> foreground.. aren't these all different things?

No. In Metacity and Compiz, the first three are the same. The fourth is
the same as the other three, except with "Always On Top" or utility windows.

> What exactly do you mean with "assuming you want a new window focused"?
> do you mean i want it raised above all other windows, or rather do i
> want it focused?

Focused.

>...
>> I don't know the exact details of the algorithm Metacity and Compiz
>> use to make this guess, and most of the people in this discussion
>> seem not to know that there even is one. So I think the next step is
>> for someone to dig into the code and find out exactly what the
>> algorithm is in Metacity, in Compiz, and in kwin (and ideally
>> research what it is in Windows and Mac OS X, too). Only then can we
>> report informed bugs where the algorithm isn't working properly, and
>> only then can we make sensible suggestions for how it could be
>> improved.
> 
> yes, such an algorithm exists indeed, i found this [1].
> I suppose that is the behaviour you are referring to.
> For Compiz, there are some GConf entries that had me a little
> confused.. just CTRL+F Compiz in gconf-editor to see what i mean.
> 
> [1] http://blogs.gnome.org/metacity/2007/12/24/stacking/

That's it, well found. Next, find a set of exact reproducible steps
where the algorithm doesn't produce the desired result.

- -- 
Matthew Paul Thomas
http://mpt.net.nz/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkwfHcAACgkQ6PUxNfU6ecqpjQCbBFNC2wM9pymNXs/wON2jPpVT
EMIAoJbI7piTerWKUdeAPKkAcyjdM0iE
=7Crw
-----END PGP SIGNATURE-----



References