76

One of the main advantages of using KDE is the amazing amount of customizability one gets from the plasma desktop. I believe this freedom should be added to other parts of the desktop environment. My idea is to make it possible to add plugins to the title bar of applications.

This is, in a way, already there, as it's currently possible to add and remove different buttons. What I propose is to extend on this functionality: make it easy for developers to write simple plugins which can be added to the titlebar, and offer a GHNS interface to add these plugins.

Many of the ideas that are currently floating around here could be implemented in this way:

If all plugins are forced to be simple buttons:

brainstorm.php#idea64847 (a button which resizes the screen tot maximum size without overlapping other windows)
http://forum.kde.org/brainstorm.php#idea82868 (Title bar per-application volume buttons)

If plugins can be anything (not just buttons):

http://forum.kde.org/brainstorm.php#idea61999 (Combining title and main menu bars)
Can't seem to find the url (group multiple windows of the same application (google chrome style))


The advantage of this system would be that these plugins would be very easy to add for users, but they wouldn't get in the way of users who don't want them. It would also allow extreme amounts of innovations as everyone who has an idea would be able to implement it by themselves, and it doesn't have any disadvantages as nothing changes for users who don't care about these features.

Being able to add custom buttons to title bar simply rocks
+1

The title bar really is a lot of wasted space. Although this is an interesting idea I would much prefer that the menu bar and the title bars are merged.

tubasoldier wrote:The title bar really is a lot of wasted space. Although this is an interesting idea I would much prefer that the menu bar and the title bars are merged.

If plugins can be more than just a single button such functionality could be added using a similar function.

Basically, if you want menu bar on the title bar -> put it on there.

If you prefer to group applications google chrome style -> add the corresponding plugin.

The more functionality you can add to the title bar the less a waste of space it'll be, and the advantage is that most of these things can then be optional extras (so each user gets the KDE experience he wants)

If I was a developer being asked to create this I would object on the basis of cost vs benefit.

I think... and i am not sure, but it will add bloat to kwin which is unnecesary to at least 95% of users. Above are 3 uses, but I think to justify this there needs to be alot more. Plugins may also cause stability issues depending on how deeply they integrate with Kwin. Firefox plugins can bring the whole application down. It's very bad if your window manager suddenly fails.

I think KWin should remain simple and focus on reliability.

anoneemouse wrote:If I was a developer being asked to create this I would object on the basis of cost vs benefit.

I think... and i am not sure, but it will add bloat to kwin which is unnecesary to at least 95% of users. Above are 3 uses, but I think to justify this there needs to be alot more. Plugins may also cause stability issues depending on how deeply they integrate with Kwin. Firefox plugins can bring the whole application down. It's very bad if your window manager suddenly fails.

I think KWin should remain simple and focus on reliability.

Yes and no... It depends upon where one wants to go with KDE and the title bar: If everyone wants to keep it as is it would indeed not be very useful. But there are already quite a few (popular) feature requests concerning it.

Either one tries to add all of those features, adding bloat to kwin for users not interested in these extras such as yourself, or a simple plugin interface is offered. The basic kwin has the features you need and others can add any features they want using plugins.

As for stability: plasma allows plugins and doesn't crash often (since kde 4.3 plasma hasn't crashed once for me). Furthermore if you don't use these plugins, they can't introduce new bugs. To anyone not needing these extra features a plugin system is probably a good thing...

It does not matter if the plugin interface is simple, any form of running of non KWin code inside KWin can easily cause a crash with simple programming error.

I virtually daily get bug reports for any control panel module in System Settings, and have to reassign these to the respective module components.

Why do not use XEmbed or XReparentNotify?

Solution will be very simple: We only change parent of widget with special flag.

I must also tell, that WMs doing the same think - reparenting window to own window.

Kwin developers have said on other ideas they do not want to use xembed because of all the problems with it.

anoneemouse wrote:If I was a developer being asked to create this I would object on the basis of cost vs benefit.

I think... and i am not sure, but it will add bloat to kwin which is unnecesary to at least 95% of users. Above are 3 uses, but I think to justify this there needs to be alot more. Plugins may also cause stability issues depending on how deeply they integrate with Kwin. Firefox plugins can bring the whole application down. It's very bad if your window manager suddenly fails.

I think KWin should remain simple and focus on reliability.

ok it make a perfect reason for not do that, but why not the inverse way, how hard can be put the three buttons of title bar in que menubar area?, can invoke the max/min funtion just like file->quit? and we could get a ride ther title bar


|doc name 'and move spot'-(app) ||(min)(max)(Exit)|menu... ...........|


//please forgive me usability teem, just an idea

I like this idea. But as it was said by bcooksley it can have some problems.
And also by doing this Kwin could become more "bloated/heavy" and not everyone wants that.

But it could be (somehow) done with aurora plug-in for plasma win-deco.
So that this plug-ins would be a "plasma" based.

(Hope I'm making any sense at all, I just got this idea and had no time to see how much of it can be implemented etc...)

From a usability perspective I think this is also a bad idea. How will volume control for instance know whether an application has sound? Will I now have volume control on Kate? Window managers should manage windows, and KWin already does an excellent job at this.

"plasma allows plugins and doesn't crash often (since kde 4.3 plasma hasn't crashed once for me)"

My desktop refuses to start without an internet connection. All I get is a black screen. If KWin was also dead that would be very very bad.

There has been a lot of discussion about this and kwin developers have chimed on on the issue so it is clearly known to them. Therefore I marking this as submitted.