[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [E-devel] Showing icons on other desks and screens in the window list (or not).....



Committed in cvs...Thanks :)

dh

Ravenlock wrote:
On 12/14/2006 09:53, Ravenlock wrote:
On 12/14/2006 09:29, Ravenlock wrote:
On 12/14/2006 10:43, Aleksej Struk wrote:
On Thu, Dec 14, 2006 at 08:36:35AM -0600, Ravenlock wrote:
On 12/14/2006 03:43, Christopher Michael wrote:
Ravenlock wrote:
On 12/13/2006 23:00, Ravenlock wrote:
Hello,

I wanted a slightly different behavior than what was available, so I've changed it up a bit on my copy, and wanted others opinions as to whether the change might be suitable for the general population.

I applaud the effort, but can you provide a patch that others can use/try ? Sometimes ideas have a better chance if they come with examples :)
K. Try the attached. Do you (the e-developers) have a preferred diff format? If you have trouble with the diff let me know.

usually cvs diff -u is preferred.

heh.  Sorry.  Here you go (I hope).


Here is a final one with the ipc updates as well. If this patch does not apply cleanly please let me know.

Please take a look at the constants defined in e_ipc_handlers_list.h
, as I took the "easy way out"... and slipped them in out of numeric order. Would you prefer I update ALL constants that follow my updates?


Ok.  One more try.

WRT the constants in e_ipc_handlers_list.h... Given our conversation, and after a little more thought, I decided you were right. They should be near the bottom, with the appropriate values. I think keeping them out of order with a note at the top would simply cause bookeeping headaches for others. So I opted for a note in the middle pointing you to where the new constants are, since they are not where you might think they would be.

Let me know if you have any troubles.


Please excuse me if I am getting my terminology (screens vs desks) mixed up.

Presently in e17.037...
when you iconify a window, you can choose whether you want to have them enumerated in the window list or not. If you choose not to, then they are not present at all. If you wish to see them, you see *all* iconified windows on the same screen. Therefore, if you have only one screen like myself, and you switch to one of your other virtual desks... you see them *all* in the list.

I would prefer them to not be there. I prefer to use each virtual desk as entirely separate from the others.

Others may also...
Presently in my patched version...
I have created options in the Configuration Panel -> Advanced -> Window List -> Window List Settings dialog which read:
 - Show iconified windows (<- This was already present)
 - Show iconified windows from other desks
 - Show iconified windows from other screens

and modified the winlist_border_add function with the following rule:
   if (bd->iconic)
     {
   if (!e_config->winlist_list_show_iconified) ok = 0;
   if ((bd->zone != zone) &&
(!e_config->winlist_list_show_other_screen_iconified)) ok = 0; if ((bd->desk != desk) &&
       (!e_config->winlist_list_show_other_desk_iconified)) ok = 0;
     }

This allows me more flexibility over what I see in the window list. If I do not want to see those windows which are iconified on desk0 *from* desk1, then I can uncheck the "Show iconified windows from other desks".
[Forgot to mention]
This matches (I think) the functionality of the ibox, in a sense. The icon boxes can show (!or not!) windows from other screens and desks.

I have given the following default values
   winlist_list_show_iconified = 1
   winlist_list_show_other_desk_iconified = 1
   winlist_list_show_other_screen_iconified = 0
as I think these represent the original behavior before the changes.

One thing I am aware of that I *did not* do (which may need addressing)... I did not modify the IPC_Handlers in any way. I was hoping for comments on that. Is it necessary? Adding something to e_int_config_winlist.h looks, well.... painful.

It's general practice that when new config properties get created, they should be put into enlightenment_remote ipc calls, yes. In the future, the ipc code is going to get a much-needed cleaning, but for now your option should go in.
ok. I started down this road. May have even completed it, though I haven't built it yet. So I backed it out for the diff attached. How would I go about testing them? using enlightenment_remote? I've never used that nor do I have any idea what its for (aside from what the name implies). I'll look into it.

Your thoughts?  Would anyone be interested in this functionality?

Regards,
-R
Cheers,
dh