GtkStatusIcon: "Activate" on single click vs. double click

[milosz derezynski <internalerror gmail com>]

Hi everyone,

I'm one of the lead devels of BMPx (http://bmpx.berlios.de), and we've imported a copy of GtkStatusIcon from CVS HEAD into our own code tree, which i have modified to also
support single click and scroll event reporting (for controlling the volume in the player on scrolling over the tray icon). The enhancement request+patch is in GNOME bugzilla at: 
http://bugzilla.gnome.org/show_bug.cgi?id=318933

There was a short discussion about that in #gtk+, with Federico and Matthias, and it sorta ended in that both single-click and double-click are to be considered as an 'activate' action, and also that it doesn't make sense to support both types of 'activate' (single/double-click) and to make it even worse, support them both at the same time so the app can do something different depending on single or double click.

Now i'm in favor of single click. I find it more intuitive to just single click on a tray icon to e.g. hide/unhide the UI of an app (that's what the modified GtkStatusIcon in BMPx does), and e.g. the ICCCM tray icon of GAIM is doing that too (hides/unhides the buddy list from the window list/pager on _single_ click on the icon), but there are also other apps who do use a custom ICCCM-compliant tray icon and activate on double click (
e.g. Azureus, just cause i have it runnign right now :)

My argumentation for single click is:

Since the tray icons can't be "selected", like e.g. items in Nautilus, which works trough a single click, while double clicking (in Nautilus) activates the item, i don't really see a point for double click, i find it contra-intuitive. A tray icon is more something like a small gadget that you "prod" so something happens, and a single click in that case makes more sense, having to double click it is something i don't think people do expect with tray icons. It's the same as with applets or launchers in gnome-panel: they can't be really selected either, so there's no "need" for a distinction between single and double click; hence they are activated on a single click, because activation on double click would seem illogical to the user: he/she would ask him/herself "If i have to double click it to activate, what does single click do then?" because they are used to this "stepped" behaviour from filemanager: Single click to select, double click to activate. They just know that double-click does something "more" than single click, that's why i think if activation would happen on double click, they would expect single click to do something as well, but would be wondering what that might be or if the tray icon is not even broken at all.

Now what do you all think about this?

Regards,
Milosz

------------------------
In theory, theory and practice are the same. In practice, they aren't.
http://bmpx.berlios.de
 | http://beepmp.sf.net | http://www.relevantive.de