KDE Commit-Digest for 3rd August 2008

In this week's KDE Commit-Digest: The Plasma "extenders" project is merged into kdebase, with initial integration into the kuiserver applet. Continued work on the systray-refactor, and more work on the "Weather" Plasmoid. A whole load of bugfixes for Kicker 3.5.10. A new "Magic Lamp" minimize effect, and a rework of the "Grid" effect in kwin-composite. Support for extracting artwork from iPod's, tag editing and removing files from MTP devices, and scriptable services (including a "web control" script), and lots of other developments in Amarok 2.0. An automatic image fetching script/plugin added to Parley. Basic XLIFF support in Lokalize. Support for regular expressions in KSysGuard graphing. Improved support for password protected archives in Ark. Support for saving file fonts embedded into a PDF file in Okular. A new, enhanced Strigi service (using KDE technologies) for interfacing with NEPOMUK. KJots and KTimeTracker can now be deactivated (while KMail, KOrganizer and KAddressbook cannot) in Kontact. Beginnings of "master pages" support in KWord. Rocs, a graph algorithm tool, added to playground/edu. "Google Gadgets for Plasma" moved to kdereview, "Timer" Plasmoid moved to kdeplasma-addons. Read the rest of the Digest here.

Dot Categories: 

Comments

by T (not verified)

I'm excited about the improved Strigi, it seems that a lot of "corner cases" are being carefully considered.

Anyone know if the Kicker fixes will make it in to Lenny?

And, thanks Danny!

by Martin (not verified)

That would rather be in Sebastian's fork of Strigi, I believe? http://www.kdedevelopers.org/node/3573

by Stefan Majewsky (not verified)

> Anyone know if the Kicker fixes will make it in to Lenny?

KDE 3.5.10 will be tagged this month. It's just a bugfix release, but I do not know if it is sufficient for Lenny which comes next month (hopefully).

by Richard Van Den Boom (not verified)

Thanks for the report, Danny, always read and always appreciated! :-)

Is there a way to bind Grid usage to a middle button / wheel click, for instance?
And is there a way to get Alt-Tab use a composited way of moving through windows, once it seems set to the old way? Once I requested that circulation was done over all windows in all virtual desktops, I could not set Alt-Tab to have a different behaviour than the old one, without any nice composited effects.
Last question : where is located the configuration to use NTP now? Cannot find it anymore.

by Tom (not verified)

Will those be available soon? At the rate KDE is evolving they probably are only up to date for a few more days ;)

BTW I hope they will be as good as the debconf ones. I find it amazing that volunteer projects produce way better videos than "commercial" entities like the Linux Foundation etc.

by Vide (not verified)

Benoit's work on KDE· kicker is just so wonderful... upstream devs are devoliping fulltime KDE4 but he's got to use KDE3, has some issues with it, has the knowledge and... patches pop up to fix those issue, because the code is opensurce. How great is that?
:)

by Stefan Majewsky (not verified)

Interestingly, the Nvidia driver problem seems to have been fixed just today. I cannot comment on that as my Nvidia card does not make any problems but: http://pinaraf.blogspot.com/2008/08/nvidia-driver-at-last.html

by ac (not verified)

hmm. for some cards. And it's beta.

by Bobby (not verified)

NVidia is a catastrophe. I was thinking that the driver situation is being improved until I bought an IBM Thinkpad R50e (old laptop) with an Intel card and compared it to my much more powerful PC which has an nVidia Geforce 6600 LE video card. The Think Pad has a little more that half of the power that my PC has but it beats it (the PC) by leaps and bounds in performance! I just couldn't believe my eyes. Smooth scrolling, YouTube videos play full screen without the slightest problem and a very fast KDE4. A performance that I can only dream of on my PC. All I have to say is, people don't touch nVidia if you haven't as yet.

by Martin (not verified)

ATI: open documentation: yes - contributing open source drivers: no
Intel: open documentation: no - contributing open source drivers: yes
nVidia: open documentation: no - contributing open source drivers: no

Take your pick. nVidia recently restated their position of providing proprietary drivers only, making the Nouveau reverse-engineering project more than they deserve.

by sds (not verified)

nope, intel give out docs on 965

by Diego (not verified)

> ATI: open documentation: yes - contributing open source drivers: no

Also this is not true: they hired Alex Deucher, one of the developers of the open source driver.
http://www.phoronix.com/scan.php?page=news_item&px=NjIwNg

by Martin (not verified)

Yes, it actually looks like they will adopt the Phoronix driver or something like it at some point, which is great. But presently, the driver that they do support, and which supposedly is developed by their core team, is closed.

by Bobby (not verified)

Users don't care a red rat about hardware documentation, all they want is that it works! Take my old Thinkpad for example; I installed openSuse 11 with KDE4 recently, which was over in no time. The system booted and there it was! everything worked like a charm. 3D was activated and I had no need to bore around with drivers. That's what users want and what nVidia and ATI don't want to deliver to Linux so why should I give them my money?
I am planning to buy a Dell at year's end and I have already decided that it has to have an Intel card.

by superfan! (not verified)

Wow! I can't wait to try Benoit Minisini's improvements on kicker ;-)
There are a lot of good improvements for people like me who are stuck on KDE3 for various reasons.
Compliments guy!

by R. J. (not verified)

Does anyone have problems with Kmail? I write my e-mails, but when I look at what is sent all spaces and lines are removed from the e-mail and it is all just jumbled together. It's with the new kmail in kde 4.1. I've just switched to thunderbird just so people can actually read what I write, lol.

by Jonathan Thomas (not verified)

I haven't received any complaints yet. ;-)

by Mrugesh Karnik (not verified)

It actually sends the mail as HTML. I'd been pissed about it till it got sorted out with the newer factory packages in opensuse.

Another kmail bug that bugged me bigtime was that it put the cursor in when replying, for top posting. I ended up editing the template and it got sorted out too.

by Bobby (not verified)

Why don't you file a bug report?

by R. J. (not verified)

I'm not really sure how to, what information is required etc. Last time I filed a report on a bug I was told I was missing something, but when I asked they wouldn't tell me what I needed to install to do a proper bug report, so I'm not that sure on how to do one, what debugging stuff I need installed etc.

by Xanadu (not verified)

The "Magic Lamp" effect is now in KWin. Whoop-dee-do. "Weather" Plasmoid. Wow. (yawn)

How about some usefull Eye Candy like, oh, I don't know...

auto-hide kicker or a desktop background when "Old School View" is enabled???

(sorry, this is just getting on my nerves a bit now...)

by Anon (not verified)

"The "Magic Lamp" effect is now in KWin. Whoop-dee-do. "Weather" Plasmoid. Wow. (yawn)

How about some usefull Eye Candy like, oh, I don't know...

auto-hide kicker or a desktop background when "Old School View" is enabled???"

As has been stated over and over again, those are both in progress.

"(sorry, this is just getting on my nerves a bit now...)"

That's no excuse for shitting on someone else's work.

by Jan Ask (not verified)

Well said.

by Vide (not verified)

These features are waiting for someone like you to be implemented. Show us the code.

by Xanadu (not verified)

How many people have been begging for the above mentioned things vs. a "magic lamp" effect. I'm simply saying the priority list needs some serious adjusting...

by Martin (not verified)

If you want to know: there was a feature request for Magic Lamp: http://bugs.kde.org/show_bug.cgi?id=167387

I implemented this feature because *I* wanted to have it. Because it looks cool. I used it in Compiz, I know it from MacOS and *I* wanted to have it for KWin.

If you want to know even more: It took me about three to four hours to code this effect. I don't think that your long missed panel hiding can be implemented in three hours. If it would be that trivial it would have been in 4.0.

And just that you understand it. I'm working on KDE in my free time. I'm not paid for it (I do not count my Summer of Code project) and I am not a Plasma dev (yes not all KDE devs are working on Plasma). I think you are not in any position to tell me what I should do in my free time and what I should work on. So don't tell me that I have my priorities the wrong way.

by Bobby (not verified)

The Magic lamp implementation is nice but it looks quite different from that of Compiz-Fusion. Was that on purpose or will changes be made?

by Martin (not verified)

Does it have to look the same? If you have ideas to improve it please tell us.

I do not know exactly how the effect looks like on MacOS or on Compiz Fusion. I don't have an Apple and Compiz refuses to work properly on my machine.

by Bobby (not verified)

I only made the comparison, not meaning that one is better or worse than the other. The Compiz-Fusion Magic Lamp has a sort of winding effect when minimizing and it kind of sucks in but the KDE4 magic lamp is more like a minimize zoom effect with a magic lamp shape.

If you could get Compiz-Fusion properly installed then you could use the Compiz-Fusion tray icon to change to Compiz or KDE4 in order to make a comparison.

by Aaron Seigo (not verified)

> How about some usefull Eye Candy like, oh, I don't know...

someone's work on kwin and another person's work on weather information has nothing to do with the plasma workspace. there are people who like and enjoy what they are doing, and your attitude towards their efforts really stinks.

> auto-hide kicker or

not written yet; will be there for 4.2.

> a desktop background when "Old School View" is enabled???

already in svn for 4.2.

> (sorry, this is just getting on my nerves a bit now...)

as the person closest to the center of the target you are spitting at, i don't accept your apology. if you were truly sorry, you would take a different approach to it. and indeed: half the things you complained about are already fixed, and your comment was totally off topic in this story.

so you can take your apology back as it has no value here.

by Xanadu (not verified)

> half the things you complained about are
> already fixed, and your comment was totally
> off topic in this story.

I don't feel that it is off topic. The topic is about what has been written for the next "release" of KDE. I have been reading your blogs (from time to time - I don't get to check it every single day I'm sad to say) and you have indeed touched on the very things I (scathingly - it seems) brought up.

Are they in SVN? Well, I can say as of this just past weekend (Aug 17) they weren't. I compiled SVN on my Gentoo machine. There was an issue with the xine-plasma-(something) not compiling that was updated later that afternoon (east coast, USA). That's about the time frame I tried it. Then I see pretty effects for KWin beating out things us lowly (l)users have been bringing up endlessly for over a year.

(there are some serious issues with the i865G Intel vid chip in this machine and the transparency of the Folder View "window" (plasmoid), but this isn't a bug tracker, I understand that.)

> so you can take your apology back as it has no value here.

Perhaps. I see your point either way. I respectfully agree to disagree. I haven't been using KDE since 0.8 or 0.9 because I don't like it. I simply don't think eyecandy should beating out functionality on the priority list especially when that functionality has been being begged for. If that is an "attack", well, I can't do anything about that. This isn't back-pedaling. This is explaining, nothing more.

by xian (not verified)

I'm kinda at a loss of words for this besides "Dude! Seriously?"

What kind of reaction do you expect from useless whining? Should Aaron and other plasma devs suddenly go "Gee, this guy is asking for a hide-able kicker. Why didn't we think of that?!"

What you did was not "explaining" whatever you think that means. It was a silly attack with no value added to the discussion. This is a story about a single weeks developments in KDE. You show up and start yelling for random crap that has nothing to do with the topic. Someone in KDE (looks like Shawn Starr and Teemu Rytilahti thanks guys!) thought they would like to make a weather applet work. You show up and start whining and complaining.

If you feel so strongly about it, start contributing yourself. You may have gotten off on the wrong foot, but I'm sure your patches or contributions would be welcomed. Otherwise it might be best to stop bugging people and making them reply to flame posts on the internet, and get back to coding and wow'ing the rest of us.

by Aaron Seigo (not verified)

> There was an issue with the xine-plasma-(something)

hm. there is no xine-plasma- in kdebase or plasma-addons ... perhaps you mean phonon-xine?

> Then I see pretty effects for KWin beating out things us lowly (l)users have been bringing
> up endlessly for over a year.

but you see, that's not what's happening. firstly, users have not been requesting some of these things for over a year, such as backgrounds for folderview, as those things haven't existed that long. secondly, feature requests that have been made have been getting attention paid to them. people moaned endlessly about not being able to resize a panel, and rightfully so; we said we'd implement that and we did. that's one example of literally dozens such cases between 4.0 and 4.1. we've typically handled the oldest and most painful issues first, as well.

and we continue to improve things, such as making wallpapers plugins so that you can use wallpaper $FOO with containment $BAR. it has nothing to do with kwin effects, and the progress in plasma is actually even faster than what we're seeing in kwin (which is moving at a nice pace as well).

> there are some serious issues with the i865G Intel vid chip in this machine

those are issues with the driver, not plasma or kde4. these things are getting sorted out upstream as we actually exercise these features in the drivers and then reporting problems to the appropriate upstreams. other than doing that, there is nothing we can do but be patient. the alternative is to live with a desktop that forever looks and behaves like its 1995 all over again.

thankfully there are many chipsets which work wonderfully already, and many more are being fixed. the latest nvidia beta drivers seem to resolve the issues people were seeing on their hardware, for instance.

> I simply don't think eyecandy should beating out functionality on the priority list

it isn't.

> If that is an "attack"

i don't think that's an attack at all, because you shared your thoughts openly and with thoughtfulness. contrast to the tone and content of your original rant; the fact you felt the need to apologize says it all imho. but with your reply here you showed that you can say the exact same things without being a dick about it, and that goes a long ways to having a rational conversation.

by Xanadu (not verified)

> perhaps you mean phonon-xine?

Yes. Still, a working rev was checked in not long after I tried. I like Open Source. :)

> (in reference to my rant about the new kicker)..it has nothing to do with kwin

Point made, Aaron. I hadn't looked at it that way.

> those are issues with the driver, not plasma or kde4.

Indeed. I just (not being a codeist) have a hard time understanding why I can play RTCW or Q3A on the box, but it can't resize/move a transparent window. I know, I know; that's for me to figure out. Devs don't "hand-hold" the users.

> i don't think that's an attack at all, because you shared your thoughts
> openly and with thoughtfulness. contrast to the tone and content of your
> original rant; the fact you felt the need to apologize says it all imho. but
> with your reply here you showed that you can say the exact same things
> without being a dick about it, and that goes a long ways to having a rational
> conversation

You are a stand up guy, Aaron. Top-notch.

by Karl Günter Wünsch (not verified)

I'm there with you.
As it currently seems there is too much emphasis in the development for new eye candy which really is of limited use it it doesn't outright diminish the value of the whole environment (if regular non-plasmoid-malarchy applications get dropped in favour of this).
I still don't get the need for new plasmoids especially as they disappear completely behind the used applications and thus are worthless in a non "boah ey, eye candy" environment. Useful things like different backgrounds on different virtual desktop screens still are missing sorely so that productivity wise KDE 4.1 still is lightyears away from being a valid challenger to KDE 3.5 in terms of usefulness...
And the more plasmoids take over from formerly useful gadget windows which you could raise above the rest of the other windows the further it moves away from ever getting to this state!
If you look at the applications, sometimes I think the developers have lost the plot - period. Take gwenview. Formerly a nice configurable, fast image browser on KDE 3 it now is a mess of reduced options slower than ever before (even after the last efforts to regain a bit of speed). Worse still the image thumbnail browser is now fixed horizontally which takes up so much space vertically where screen estate comes at a premium, especially as wide screen displays are getting widespread.
I have been using KDE from the very beginning, KDE was the reason for our company to go with Qt for multiplatform development - on my recommendation. Almost 10 years have passed since then and given the state of Qt I am slowly but surely revising my opinion.

by Anon (not verified)

*sigh*

"As it currently seems there is too much emphasis in the development for new eye candy which really is of limited use it it doesn't outright diminish the value of the whole environment"

a) Just because *some* new contributors, who would probably not contribute to KDE at all if they didn't work on what they're working on now, happen to work on eye-candy related stuff, does not mean there's some shadowy conspiracy to abandon functionality in favour of eye-candy.

b) Eye-candy is generally configurable: in fact, the two examples the original poster took issue with are not only configurable but are disabled by default! So I don't see how this diminised the "value of the whole environment".

"(if regular non-plasmoid-malarchy applications get dropped in favour of this)."

Any examples of this happening?

"I still don't get the need for new plasmoids especially as they disappear completely behind the used applications and thus are worthless in a non "boah ey, eye candy" environment."

You don't see the need for menus, panels, system trays, task managers, the Jobs server (which I personally consider rather ugly, but very useful) the folderview, etc? And hint: At least two of the plasmoids I've just mentioned do indeed display over other windows.

"Useful things like different backgrounds on different virtual desktop screens"

I'd consider this an absolute prime example of "useless eyecandy". A lesson, here: one man's useless eye-candy is another man's essential feature.

I don't use Gwenview so I won't comment on that.

"I have been using KDE from the very beginning, KDE was the reason for our company to go with Qt for multiplatform development - on my recommendation. Almost 10 years have passed since then and given the state of Qt I am slowly but surely revising my opinion."

And what "state" is Qt in that's so terrible compared to 10 years ago?

by Aaron Seigo (not verified)

you took the words right out of my mouth.

the idea that wallpaper-per-desktop impacted productivity particularly made me smile.

by T. J. Brumfield (not verified)

Playing devil's advocate here as I don't really use multiple desktops, but I can see where having different wallpapers on your different desktops would help you differentiate the desktops at a quick glance. One could argue that it does affect productivity in that sense.

by Morty (not verified)

And one maximized window remove the usefulness of it, reducing it to nothing more than eye candy. As opposed to quick glance at the desktop pager, which always work. By highlighting the active one, helping you differentiate the desktops.

by Paul Eggleton (not verified)

> Useful things like different backgrounds on different virtual desktop screens

So, how is this "useful" and not just "eyecandy" by your reckoning?

> And the more plasmoids take over from formerly useful gadget windows which
> you could raise above the rest of the other windows

Are you aware you can use CTRL+F12 to bring the Plasma desktop to the foreground?

> given the state of Qt I am slowly but surely revising my opinion.

Wait, are you complaining about Qt or KDE now?

by Tim (not verified)

Not to side with the ranting people, but Ctrl-F12 is hardly an obvious or well-known shortcut! Nor are users told of this shortcut at any point.

There should be a "Show Desktop" button on the panel.

by lmarteau (not verified)

There is already one you can add in 4.1.

and another one for show Dashboard...

you have the choice :)

by Aaron Seigo (not verified)

> disappear completely behind the used applications

the dashboard view (Ctrl+F12), panels and plasmoidviewer (which shows applets in their own window; will eventually be integrated into plasma itself, though i'm not sure when yet as we have other things to do) negates this point.

> the more plasmoids take over from formerly useful gadget windows

there's something fundamentally missing in this viewpoint, which becomes obvious when you look at amarok2: it uses plasmoids extensively within its window. it's not about "taking over from formerly useful gadget windows" at all, but rather providing a better way to create such things that can be widely used. your "useful gadget windows" were pretty much destined to remain their own little windows forever; as a plasmoid they can be their own little window, part of the desktop, on a panel or in other apps that care about them.

in that sense, they are a bit like kparts, but with a completely different scope and purpose.

> different backgrounds on different virtual desktop screens

we'll be doing one better than that: different widget layouts for different virtual desktop screens.

besides the fact that "different backgrounds" as done in every other shell out there doesn't work at all for composited window managers (think of desktop cubes and grids), it would be fairly easy to add this to the Image wallpaper plugin if you so desired.

.. not to mention we're doing a lot of things now with wallpapers that will likely be in 4.2 like Edje driven papers that are generally pretty cool looking. ooops, more eyecandy! ;) this is being done by people who never contributed to KDE before, though, who are doing this as an experiment on using Edje in KDE. we as users get to benefit from that experimentation. nothing in this case is lost (developer time or effort, usability, etc) even though the eye candy factor goes up yet again. in fact, we get more interested developers out of the deal.

so yes, we don't have per-desktop wallpapers, but it's easy to implement and we've done (and are continuing to do) a hell of a lot more than that which makes that feature look really rather trivial.

> productivity wise

the only connection between desktop wallpapers and productivity i can think of is that you might use it as a visual hint as to what desktop you are on. but that seems like a real stretch, so i call BS on this point.

if you wish to talk productivity ... as opposed to "different wallpapers per desktop", apparently most people are interested on working on really useful and needed things like improving the system tray so that it doesn't suck anymore, something that has languished for *years* and which, only because of the approach we're taking in plasma, is finally getting attention and work done on it. that sort of thing can really improve productivity.

> I think the developers have lost the plot - period. Take gwenview.

i couldn't stand the rather arcane interface of gwenview previously; it also didn't fit at all on smaller screens (e.g. the EEE PC). the new gwenview has all the features i ever used in a much nicer package. it's something i'm not only more comfortable using, but also more comfortable giving to others to use as well as it looks far more polished and is easier to figure out how to use now.

so maybe KDE has moved on from your personal needs/viewpoints-on-the-world. that would be unfortunate, but KDE can't make everyone happy all the time. what we can do is make more people more happy, and it's pretty difficult to look at something like gwenview in kde3 and gwenview in kde4 and not come to the conclusion that that is what the developers are doing.

by Sebastian (not verified)

I agree in some of your points and I can add a whole list of others.

But the thing is:
o Trolltech/Nokia pushes KDE to mobile devices, where the priority of a *NIX-like desktop behavior is irrelevant.
o I accept that some innovations may cost features (If KWin-composite further inspires from MacOS, I do not see any room for different wWallpapers on each desktop)
o people who had a view on the whole in the past nowadays work on different places within KDE (most on Qt at Trolltech from what I've learned)
o developers come and disppear. New generations have different interests and different views on quality management, maybe even different definitions on that.
o KDE can not reliably supply the service that Microsoft and Apple can offer - we can be happy that they do an awesome job on that.

Therefore, I do accept that there are many unpolished things in KDE, though I am very unhappy with this. It had been better in the past.

In fact, if I consider most new effects in kwin (to come back to the original topic), most of them are nice to look at, but without any use, for example the window flip effect on "Alt-Tab".
I always used Alt-tab for fast finding windows on a desktop (for example when the taskbar is crowded or if a window title was too long to be displayed in the taskbar). But with the new effect I am slower than before. I can only see the current selection, the other windows are visually transformed or invisible at all. That means: The effect looks nice and shows that KDE programmers have the same skills as Microsoft developers, but the work which has gone into that just adds some kilobytes to the download volume and some extra lines of source (which must be maintained!) Whatever: Send a bug report on that and hope that somebody improves this.

by Martin (not verified)

>In fact, if I consider most new effects in kwin (to come back to the original topic), most of them are nice to look at, but without any use, for example the window flip effect on "Alt-Tab".
> I always used Alt-tab for fast finding windows on a desktop (for example when the taskbar is crowded or if a window title was too long to be displayed in the taskbar). But with the new effect I am slower than before. I can only see the current selection, the other windows are visually transformed or invisible at all. That means: The effect looks nice and shows that KDE programmers have the same skills as Microsoft developers, but the work which has gone into that just adds some kilobytes to the download volume and some extra lines of source (which must be maintained!) Whatever: Send a bug report on that and hope that somebody improves this.

Nobody forces you to use these effects. We did not replace the old behaviour. Just disable the effect and you have the normal tabbox. Currently there are four effects for alt+tab in trunk. You can choose which one you want to use or none at all. Where's the problem? Btw. you can turn off the animations in flip and coverswitch effect. So the effect can be as fast as without compositing if you want.

The argument of some kilobytes to download is IMHO not valid. The effects are just 378 KB all together in 4.1. Also lines of code is just BS. If that's an argument we should stop implementing features as this add new lines of code and needs to be maintained.

by Sebastian (not verified)

> Nobody forces you to use these effects. We did not replace the old behaviour.

I know. I actually use the "old" behaviour. And I said, I accept it, but I am unhappy about the fact that it has been integrated at an early development state. And as soon as unfinished things are integrated it sometimes happens that they won't be improved any further.

> Btw. you can turn off the animations in flip and coverswitch effect. So the effect can be as fast as without compositing if you want.

I am not talking about the speed of the animation. The problem is: Using the very old "menu" that appeared on KDE 3.5 I obtained more information (!) compared with any existing effect in kwin-composite4.1 (even without any effect the menu contains less information compared with 3.5). That means: "Using" the "alt-tab" feature is less fast compared with 3.5!

> The argument of some kilobytes to download is IMHO not valid...

This is nut-picking. When you play this card I ask you why KDE includes features that do not create any value? That's my only point.

See, I agree that playing around with neat 3d effects is required and even makes fun. People should do that. But from my user's perspective, it would have been better if these effects where done 'right', that is if they do not only present the new abilities of Qt/KDElib/OpenGL/whatever, but if they provide eye-candy AND improved navigation.

by Martin (not verified)

>See, I agree that playing around with neat 3d effects is required and even makes fun. People should do that. But from my user's perspective, it would have been better if these effects where done 'right', that is if they do not only present the new abilities of Qt/KDElib/OpenGL/whatever, but if they provide eye-candy AND improved navigation.

So please tell us how to do them *right*. I'm always glad for ideas how to improve the effects. If you have an idea how to do the perfect alt+tab window switcher please tell us, so that we can implement it. The right place to do it is http://bugs.kde.org/.

by Danny Allen (not verified)

The flip effect alt-tab switcher *is* improved navigation: by being able to see the contents of each window, it provides more information for the task. And I like how it looks.

As previous replies have said, it is also only one of a choice of 4 options.
Good developers do not only take into account their personal preferences and usage when designing new features...

Danny

by m. (not verified)

I see contents but lose context. Lets see example.

At the moment I have four windows on desktop, when hitting alt-tab in 3.5 I see:

- Eye Candy - Konqueror
- hp12c.pdf
- mikolaj@localhost: ~/kget - Konsole
- E-Mail - Kontact

What this list is giving me?

a) I know whole context of my current work (I use one desktop to group all relevant tasks - not rare scenario)
b) I know that I need to hit Tab two times to get to E-Mail
c) I can even... *gasp* use mouse to quickly select window I want to get - useful with longer lists

by Martin (not verified)

Please tell me where's the difference to KDE 4? You have exactly the same features if you do not use compositing or even if you use compositing and disable all effects for alt+tab.

If you use the box switch effect you have exactly the same with one difference: You see a thumbnail of the window instead of the application name. Personally I think that is a big improvement.

So I do not get your point. If you know the perfect effect for window switching please send us the patch.