Re: [Ayatana] unity and notifications

[Matthew Paul Thomas <mpt@xxxxxxxxxxxxx>]

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

Diego Moya wrote on 20/09/10 15:59:
>...
> Diego Moya wrote on 17/09/10 13:59:
>>
>>> Yes, those bubbles are actually quite similar to what NotifyOSD tries
>>> to do. But look how Jef required a way for the user to take control
>>> of how bubbles are presented.
>>
>> Which was unrealistic. By the time you had remembered, and then typed,
>> the command to reduce the opacity, the error would have faded away by
>> itself anyway.
>
> I think it was meant to configure the default opacity to a level at
> which the user feels comfortable. The Ayatana design dismissed the idea
> to let users configure notifications to their needs, and I never got
> the reason for that.

For the same reason operating systems don't have visible controls for
configuring the appearance of tooltips. If they're designed right, they
should be below the threshold of triviality. If a substantial proportion
of users want to configure them, there's something wrong that we need to
fix.

>>> The Ayatana notification design took the
>>> original idea and then changed it to its opposite by removing every
>>> safeguard feature to put user in control and keep it humane.
>>
>> That is not true. I had read the book years ago, and seen that idea,
>> and thought it was crack (mainly because the mockup was obviously
>> rigged to ensure the error text didn't overlap the background text).
>> But by 2008 I'd forgotten about it.
>
> And then you went designing a notification system based on translucent,
> non-dismissible bubbles? I didn't mean using a similar design was a
> deliberate decision, but surely it had some influence?

No, not at all. When I say "I'd forgotten about it", I mean, I'd
forgotten about it. :-)

>...
>> We could do better with the fading, but I find myself unable to
>> imagine a situation where someone simultaneously (a) needs to see
>> what's under a notification bubble and (b) "needs the cursor
>> elsewhere". Can you give an example?
>
> I think Michael Jonker already provided one at this very thread; that's
> why I elaborated on it. When working with the Gimp, he needed to see
> the layers panel while working on the scene, and it was obscured by the
> bubble.

As Mark Curtis said, that's a potential problem no matter where the
bubble appears. And any interaction to move or dismiss the bubble would
be just as much of an interruption as moving to fade it.

>...
>> When we designed Notify OSD, we did not have touch screens in mind.
>
> Does that mean they will not be supported, ever?

They're "supported" right now.

>                                                  Or are suggestions
> welcome to also support touch screens, even if it implies changing the
> original design?

That too.

>>> The starting point i feel is: You touch it and it goes away.
>...
>> So, how would you avoid touching it by accident just after it appears?
>
> Several viable options to avoid the problem:
> - Touching it don't dismiss it for the first 0.5 second, enough for the
> user to notice it.

That's a possibility, but we would need to define more carefully what
"don't dismiss it" means. If I clicked on the part of a bubble above a
window's close button, 0.45 seconds after the bubble appeared, would it
close the window, or do nothing at all?

> - Don't show notifications if there has been user interaction in the
> previous 5 seconds.

Interesting idea, though it might result in a long queue.

> - Even if the bubble is closed, allow the user to reopen it from the Me
> menu.

Many if not most notification bubbles have nothing to do with social
networks or instant messaging, so that would be a poor fit.

- -- 
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/

iEYEARECAAYFAkyYueAACgkQ6PUxNfU6ecrnmACgqwtxbKPfsmpNHIOr7Zx/T2y1
WSMAoMXnuisj7UuaWm1tZKFfSXRz8iFx
=7zXZ
-----END PGP SIGNATURE-----