Details about WLAN subsystem future

Carla Schroder gave some more details about the future of the WLAN stack, showing the future of the development once mac80211 is included in the mainline kernel.

The new subsystem is still not part of the latest kernel. But since that is just a question of time the future of the stack is already in development.

The problem with even the new driver subsystem is that it only supports drivers where part of the work of the WLAN card is done on the host CPU – and is therefore implemented in software (“soft-MAC”). Other cards, “full-MAC” called, implement everything in hardware. These are not supported by the new subsystem.

However, the current subsystem ieee80211 supports some of these. Therefore the aim is now to take some of the existing code and merge it into the new subsystem. But since this will add new base functionality it will also get a new name – “lib80211″.

This new library will then replace all drivers – with two exceptions: the ipw2100 and ipw2200 wlan cards will need there own library. But as long as these are the only ones the Linux kernel might survive that.

Thanks to Carla Schroder for an excellent article about the future development.

kwin composite supports blur effects

The kwin composite extension now supports blurring for better reading. However, the hardware needs for this feature are quite high.

The new feature by Rivo Laks was mentioned in the last commit-digest, issue 56:

Blur effects blurs out background of translucent windows, thus improving e.g. readability of text in such windows.

The reason to have this functionality comes clear if you compare the following two screenshots:

Both images show a transparent window in front of another window. In the first image the text of the active window is difficult to read because the text of the underlying window disturbs the reading. The second image now has a blur effect activated, making it much easier to read the text of the transparent window.

You can now ask how much sense it makes to have a transparent window when you want to read text anyway, but this is about eye candy, and several people like it that way. I even impressed my (totally windows only) room mate two days ago by just showing some funky Beryl things.

Anyway, whether you like it or not, the implementation of the blur effect is quite hardware intensive. As Rivo Laks points out in an additional e-mail:

It uses shaders and render targets, so ATM it needs GeForce 6000 or up (not so
sure about ATI cards, Radeon x500 or something and up may work).
Once I get our GLRenderTarget to work with ARB_texture_rectangle, it will also
work on cards not supporting NPOT textures, that is GeForce 5000 / Radeon
9500 and up. I’m not sure if those cards are fast enough for everyday usage
though…

That’s quite a lot – I wonder which Intel chips (which have the best free drivers available) will work with that.
I doubt that most people have such powerful graphic cards, so I think many people will not see this new feature anytime soon…

Shiny Dolphin Mockups

Nookie provided some Dolphin Mockups to show how he imagines Dolphin. While some of the enhancements do look a bit like a file manager of a certain wide spread operating system they do look good.

Nookie is a kbfx developer. I’m not sure why he produces Dolphin Mockups, but they look very slick and shiny anyway.
Of course, if you have a closer look it is all a bit clustered and might even be seen as cramped. On the other hand Dolphin is a file manager and therefore is intended to show lots of information. And, last but not least, these are mockups, so there is no reason to complain or whine. Just look and enjoy – or hit the close button.

So, here are his mockups:

The first mockup with the media bar is explained in this post, the second mockup with the video file preview is explained here.

I pretty much like both mockups because they add a necessary touch of glass like “shiny-ness” which I like. Sure, it still would be important to have the ability to turn it off, but that’s the smaller problem.
Also I very much like the idea of the media bar! Of course, together with the space stealing “Dolphin” image there is not much space left for the actual display of the files – but imagine a system where you have the shiny “Dolphin” title bar, and embedded in this bar the icons for the different view types and such stuff.

In any case, these mockups show that there are people around who have ideas of how to improve the visual appealing of KDE 4. Sure these mockups are not ready for productivity yet, but they show a direction and I would like to see more of that.

Discussion about current state of Plasma

Frerich Raabe raised the question on planet KDE what the current state of Plasma is. He asks if there is enough code in shape to have something ready for KDE 4 and calls for new developers to step up and fill this suspect ed gap.

The question doesn’t come from no where: compared to all other KDE 4 base technologies, Plasma is the one which got most of the attention while just few code bits and almost no previews were released. Also, it is one of the few things Troy Unrau haven’t introduced in his famous “Road To KDE 4″ series.

Still, it is not entirely like Frerich states: there is not nothing. krunner for example is developed together with Plasma, and it is already working. But I do agree that there was nothing shown yet which comes close to certain mockups which were shown when Plasma was announced. Also the raptor menu developed by the kbfx guys is supposed to be a Plasma widget.

The discussion below the mentioned thread shows that there are more people uncertain about the current state: almost all people agree about the state, that they haven’t heard anything. Some people suppose that the development might take place outside of the SVN repository to release it in one big bang. However, even Oxygen is included in SVN now and I don’t see reasons not to include cool Plasma things. Additionally Plasma should be included early enough so that other people can start writing widgets for it.
But at the moment there is not too much to see.

I’m not sure what to think about it – on the one hand I do trust Aaron Seigo, one of the main figures behind Plasma, to do the right job. I doubt that he will disappoint the community. On the other hand Plasma is supposed to replace very important – and very visible – parts of the desktop: kdesktop, kicker, k-menu, superkaramba. So I would like to have clarity about that topic.

Maybe the Plasma developers will see the need for some clear words about the topic, now after the question was raised on the planet.

Skype for Linux 1.4 [Update]

Skype for Linux 1.4 will be released after this weekend.

This was just released by Skype’s Linux blog:

Skype for Linux 1.4 has miserably failed to launch today.
Because of some technical problems we took more time than expected to get it out and then it was decided to abandon the release today.
Unfortunately, we can not release on friday, because during weekend we could not wipe your ass fast enough in case of emergency and that is a problem.
So, you’ll have to wait whole week more before you can take a look. Sorry for that.

A changelog is not yet available (or, better said: I don’t find one), but we can expect a Qt4 client. Also, this note by a developer at least gives hope:

Also never mind the version number – 1.4 just means “right after 1.3” and is not tied to any particular Skype for
Windows version or specific feature set.

Update
The original post quoted above has been removed from Skype’s Linux blog. You can still see it however here.