[rekonq] [RFC] Better Integration of Web Applications into Window Management

[Martin Gräßlin]

Hi rekonq developers,

I am not subscribed to this list, please keep me in CC!

As you might have seen, I added an entry to your agenda for your IRC meeting 
about better integrating web applications into window management. I want to 
outline the ideas I have in my mind in this mail and want to get some feedback 
what you think about it.

First let me describe the current situation:
More and more web applications are replacing standalone desktop applications. 
A good example is gmail replacing kmail, but there are also applications like 
Facebook or Google StreetView which never existed as a desktop client. As 
"cloud" seems to be the buzzword of the decade I do not expect the trend to 
stop, even if I do not like web applications (as my day job I am writing web 
applications).

The entry point to web applications is a browser. And here is the problem. We 
have one application which embends multiple applications. We have a great 
desktop experience to manage applications, we have a tasks applet, present 
windows, desktop grid, Alt+Tab, Window Tabbing and so on. But this is all not 
available for web applications and that makes working with multiple 
webapplication a terrible thing to do IMHO.

Solutions:
There are diffierent solutions to tackle the problem. Google tries to free the 
devices from everything what is not a web application with Chrome OS. This is 
IMHO a terrible step as desktop apps are in general better.

The mobile devices try to free the web from the browser by providing apps. 
This is more of a solution to the problem that most web sites are unuseable on 
small and touch screens.

A third solution is the Silk project. An in my opinion awesome idea, but it 
seems to lack manpower and by that I gave up the hope on it :-(

All of theses solutions seem to not fix our (KDE's) need. So here comes my 
idea;

Proposal:
Let's make web applications integrate into the desktop through the browser.

The idea is to make web apps behave like a normal application. They get an 
entry in alt+tab they have a tasks item, they have a preview in the toolbar, 
they show a window in Present Windows. Nevertheless they would still live in 
the browser.

So let's say that a user has one browser tab with Gmail and one with Facebok, 
then he will not only have the entry of rekonq (which might even be dropped), 
but also an entry for those web apps. If the user selects Gmail through e.g. 
alt+tab the browser window get's activated and the browser switches to the tab 
containing gmail.

Technical Solutions:
One possible solution would be window tabs. Each web application would be an 
own rekonq window grouped together by tabs. It would be an improvement but 
after lots of thinking I came to the conclusion that it is not the best from 
user experience point of view.

My idea would be more revolutionary and is nevertheless technically feasible. 
Rekonq would have to tell us  (kwin/plasma) which web applications are 
available (either through dbus or X), there names, their icon and their 
thumbnail. Now KWin could include these web applications as "children" to the 
actual window. Both our effect framework as our alt+tab solution are 
abstracted enough from windows to allow this. E.g. alt+tab is also able to 
manage desktops. Also in Plasma I think that this is no problem to integrate 
as there is the libtaskmanager and the tasks dataengine.

Btw. our Alt+Tab code is so strongly abstracted from KWin that you could 
easily integrate it as a ctrl+tab to walk through browser tabs ;-)

What would it offer to KDE?:
Such a solution would be pretty unique and could be quite a selling factor for 
KDE. This is something no other plattform can provide (except Microsoft 
Windows + IE resulting in probably another fight against EU commission :-). We 
are the only other project having control over the desktop shell, window 
manager and browser.

Further discussions:
If you think that this idea might work, I would like to discuss it with you in 
more detail. I even think that this is a possible topic for the usability 
sprint mentioned by Celeste some time ago. If it would be accepted to the 
sprint I would go there although usability is not my most liked working area 
:-)

Implementation:
That is the bad part. I won't have time for such a project for quite some 
time. If you would want to experiment in that area, you would have to work on 
the kwin part. Of course I would provide help. This might be something nice 
for a GSoC project which I would co-mentor. If as GSoC I would only co-mentor 
if it is carried out by an existing member of the community (we had in KWin 
too many code dumbs which are now unmaintained :-( ).

So please let me now what you think about it.

Cheers
Martin Gräßlin
(KWin Maintainer)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 316 bytes
Desc: This is a digitally signed message part.
Url : http://mail.kde.org/pipermail/rekonq/attachments/20110110/de9f582b/attachment.sig