Kubuntu 11.10 Sneak Peak

Last week the Ubuntu project met in Budapest for the Ubuntu Developer Summit. The Kubuntu team discussed an incredible amount of cool things, of which I’d like to present a number of generally interesting topics.

A very strong focus of the 11.10 release will be continuing innovation in the area of embedded systems such as mobile phones, but also for the first time on tablets. Since the work in embedded systems is quite extensive, information on that will be posted separately in a special ‘Embedded Sneak Peak’.

Boring things first. Kubuntu 11.10 will come with KDE 4.7.2 as well as Qt 4.8.


Accessibility always was a bit of a sore spot in Kubuntu. Mostly for two reasons:

  1. Getting the average developer to care about it is hard
  2. Qt did not support some important features

While the first problem is difficult to fix, the second one is perfectly fixable.  Some weeks ago Qt developer Frederik Gladhorn presented the initial code for Qt AT-SPI 2, which essentially allows screen readers such as Orca to work with Qt (and thus KDE) applications. This is absolutely great news for everyone who needs to use a screen reader, as the better part of the free software desktop world (Qt and GTK+ applications) will be accessible.

In cooperation with the awesome Ubuntu accessibility team we are doing our best to get this into Kubuntu 11.10.

We will also improve the out-of-the box accessibility experience. Primarily by integrating existing KDE accessibility software stronger into the Kubuntu system.

To find out more about Ubuntu accessibility check out their blog.

Low Fat Kubuntu

As a result of ongoing efforts in the area of embedded systems, the Kubuntu team has acquired a lot of knowledge in slimming down the default KDE Workspace to a bare minimum, freeing up more resources for other use. This makes it possible for use to introduce a new set of default configurations enabling a low fat Kubuntu setup.

While we realize that many of the technologies that make the KDE 4 workspace the incredible amazing product that it is, we also see a demand in reducing resource consumption on systems with limited hardware resources. The low fat configurations of Kubuntu 11.10 will enable users to run the KDE workspace on setups where it is essential that the base system uses as little resources as possible.

Touch This

Needless to say from all the work on “touchable” systems, we have plenty of awesomeness related to touch up our sleeve. Kubuntu 11.10 will by default (though probably not activated) come with an application called Touchégg.

Touchégg enables you to assign certain input gestures from your touchpad or trackpad to certain actions. For example if you tap with two fingers at once it could mean a right click, or if you swipe with 3 fingers from left to right it means ‘switch to next track in Amarok’.

After having tried this on my netbook I must say, this is one of my personal favorite new features, though one needs a sufficiently large touchpad to use it properly.

To find out more about Touchégg please visit the project website.

Package Manager

For quite some time we have been using KPackageKit as default package manager in Kubuntu. While it has seen constant improvement over time, a very strong competitor appeared meanwhile: Muon. Muon is developed by one of the Kubuntu Developers and uses, unlike KPackageKit, the native APT libraries. As it is using APT directly it enables Muon to have a much tighter integration into Debian-like systems (such as Kubuntu) as well as expose specific functionality of APT/DPKG more directly.

As a result of these and other advantages the default setup of Kubuntu 11.10 will feature the Muon Software Center instead of KPackageKit, The Muon Software Center has an interface very much like the Ubuntu Software Center, it is however not a clone of it.
For more advanced users there is also a Muon Package Manager which has an interface much like Synaptic and thus allows for more involved “management” of packages. Whether the package manager interface of Muon will be part of the default installation of Kubuntu or not is not yet decided, in any case it will be easy to install.

To find out more about Muon please have a look at Jonathan Thomas’ blog about Muon 1.2.

44 thoughts on “Kubuntu 11.10 Sneak Peak

  1. Sounds like some good times ahead. I really like the idea of trimming the system down (low fat). Releases have been feeling heavier and heavier making me wonder how long my current laptop will last even though it is perfectly sufficient today (or rather, yesterday)!

  2. These are all great news, but I just wish, for now, that the Ku team reach a good basis of stability. Natty came with a lot of bugs — most for Intel chipsets — and annoying, hard to find workarounds. Unfortunately, this release is definitely not good to go for the average user. 😦
    Even I, as a Ku and KDE lover, have considered temporarily switching for Gnome or another environment until all this bugs are fixed, but hopefully I won’t need it (right?).

    • There is little we can do about graphics driver constantly breaking. Generally speaking though we are trying to consolidate towards LTS releases, yet we cannot avoid breaking some things in the process of getting there.
      We do however try to fix stuff ASAP once we find out about it. Excessive testing of stuff before releases is very welcome 😀

  3. is that the new kubuntu theme in your screenshots? I like the unity-like plasma theme a lot. IMHO kubuntus design should be more ubuntu-oriented. Not because of branding or community but simply because ubuntus designers are doing a great job. When it comes to unity, wallpapers, colorthemes or splash screens ubuntu is way ahead of kdes air plasma theme and horos splash screens/wallpaper.

  4. Awesome. Low fat kubuntu sounds like a perfect idea!

    Now tell me, why isn’t KDE Ubuntu’s default DE? Isn’t it possible for them to port Unity to KDE? It’s a compiz plugin, and it’s possible to run compiz on KDE, so shouldn’t it be possible? All the other technologies that are left in a DE seems to be much better in KDE than Gnome. Qt is from what I’ve heard better (easier and more extensible) than Gtk and the applications are generally better.

    • I cannot answer this as I am not a Ubuntu desktop strategist.

      But from a puristic POV: If you take Unity and Compiz and combine them with KDE applications you’d not have a KDE workspace as it is largely defined by Plasma and KWin (which in such a setup would be replaced by the two aforementioned apps). Essentially you would end up with a hybrid, not that this would be bad, just saying that it would not be a KDE workspace anymore. It is however entirely possible right now, of course Unity would have some odd integration issues with the rest of your applications here and there, but other than that it would just work as well.

      For that matter, you could also go the other way around and ditch Unity and Compiz in favor of Plasma and KWin on a Ubuntu setup. Whether that would then make Ubuntu a KDE workspace is a discussion I will leave for people who are into philosophy 😉

      As for Qt: http://www.youtube.com/watch?v=VisI7VWwL_g

      • Also, Unity 2D is a Qt port of Unity 3D (without the 3D obviously) and screams. Based on Qt I’d personally say the two will work well together, though I do prefer the main KDE desktop 🙂

    • I don’t regret the idea of a low-fat Kubuntu, but the amount of fat cutted with Akonadi 1.5.3 and Nepomuk in KDE SC >4.6.1 is massive. Compare SC 4.6.0 and 4.6.3 and you’ll get me. Add Strigi 0.7.3 to the mix, and it’s even more fat gone. Nepomuk will now work happily with 80 MB of RAM. Strigi 0.7.3 is 2x more efficient than 0.7.2, IMHO. About Akonadi… we are all waiting for the transition to Virtuoso, to free the memory used by the embedded MySQL server.

      You can do even more, like moving the Digikam and Amarok databases to Virtuoso. Those seem to be better places to do real fat cutting.

  5. Muon definitely looks interesting!

    There are some basic stuff I miss in Kubuntu. One of them is a solid backup system. While there are candidates out there, I have not found anything I really have fallen in love with yet. A solid Qt frontend for rsync, with two levels of configurations (standard and advanced) should cover it. But it has to be well integrated and stable as well as still being in development.

    Another thing is a little bit of a sad chapter. All operating systems/distros needs a solid PIM if they want to have any kind of reach in the non-geek environments. Kontact is in many ways perfect – it has all the components needed, the idea of having Akonadi as a backend is also good – meaning a lot of functionality can be added by adding resources directly to Akonadi. The problem is manpower – and part of this is how Kontact is treated in the KDE environment. It needs a BIG push, and Kubuntu could help push this in the right direction. It needs to get up, front and center. Developers needs to be encouraged to step in and help, if more assistance is needed in managing the whole project, then it should be added from the highest level. In short – Kontact is really a mission critical application.

    If you want to know how critical, then take a look at Windows. A lot of companies are locked in to Windows today ONLY because of Outlook. There is nothing on the market that can be integrated with the ecosystem of business management applications in the same way as Outlook. This is the single, biggest reason for not trying something else (with the complete Windows Office as a second). Thunderbird is nice, but it does not cover the same areas that Kontact cover. Without a doubt, Kmail in Kontact is my favourite mail app. But as long as it is not complete and stable, I can not recommend it to anyone.

    So get a good backup app that will do backup to a pendrive, USB harddrive, Samba/NFS drive, ftp or cloud storeage. Then get 20 developers to hammer on the Kontact code and get things sorted out. Even better – why not arrange codesprints around the world with focus on Kontact?

    I am sorry this comment is getting rather long, but better get it all in… 🙂 I still have a couple of more points

    It is a good thing that Samba is starting to shape up. I still have some problems connecting to a Samba share on a WinXP computer and miss a point under Properties when right clicking on a share – a place to add a different user and password. But what I am missing more is a way to add and connect to NFS shares that is just as easy as Samba! Why does it have to be a ton of command line editing to get this to happen? Here is what should happen:

    Right click on a folder in Kubuntu.
    Choose Sharing
    Click on the tab saying Security and set the access for users and groups
    Done – this should take care of all the stuff a normal user do not need to know.

    To connect to it should be as easy – we already have a wizard for network connections, we just need to add NFS. Choose it and you should have the choice of entering the name/IP address to the computer, the name of the share and optionally a username and password in case you are not connecting with the same user and pass you logged in with.

    My last point for now is something that will help anyone that has two or more computers with Kubuntu. We are using apt-cacher ng here and I simply LOVE it. Even if you have a fast internet line, you can never beat the speed you get from a local server over your lan. But during installations, unless doing a bit of jumping through hoops, you can not use this local cache.

    It would be great to see an option (could even be under advanced etc.) where you could specify the IP address of a local cache. This would really speed up an installation, and you would not need to add it later. But for later changes, there should be an option in Muon where you can set the address to the local cache, maybe the same place where you can choose what server to use for the updates.

    I thank you for reading this far down. I know it was a lot, but I think it is needed to move Kubuntu forward.

    • Most of these things are better said in bug reports than in a comment here TBH.

      You will however be glad to hear that we are looking into backup solutions for 11.10, whether we come up with something sane is the question though.

    • As for backup I use Luckybackup that seems to fulfill your requests?

      As for Kontact – well I sort of agree though being an avid user.

      Muon looks good! Rekonq is frankly pretty useless and what I sorely miss is some sort of mobile devices/phones comm-backup system. For now I use Wammu – it aint pretty but it works for the basics.

  6. I have nothing but appreciation and gratitude for the great work that went into Natty and what I see coming for Oneric. Just great, can’t wait!

  7. The inclusion of Moun would be a great step forward! I also support the comment above about a backup system… I know many linux users (myself included in the past) frequently just use rsync and/or archive their home folders. I’d definitely appreciate an inclusive graphical solution with encryption and cloud storage.

    I don’t want to sound like a pessimist or in any way disparage the amazing work many developers have done on the mobile end of kde/Kubuntu, but I also think it’s time to reconsider the viability of porting a desktop platform to a mobile platform. As Apple has demonstrated, it seems more intuitive to approach it from the other way around. With netbook sales plummeting and tablets taking over the market, Kubuntu isn’t exactly in a great position to invest so much time in a mobile endeavour. Let iOS, Android, QNX, WebOS and Meego fight for market dominance there. Two of those are open source projects, so it’s not as if the option for an open mobile OS is lacking. Rather, Kubuntu has a ton of potential in exploiting the tablet market, and I think it would be advantageous not to be distracted by specialized use cases that are extremely unlikely to be very popular for mainstream use.

    BTW, unlike Ubuntu Natty, I absolutely love Kubuntu Natty! Great job guys!

    • From a technical point of view, I would not say that desktop->mobile makes more sense than mobile->desktop or vice versa. If we reduce the KDE workspace to Plasma (that is libplasma) and KWin, then we have to realize that libplasma was from the very beginning designed to power various user interfaces for various device form factors. So, essentially what you need to do is build a new Plasma user space experience upon the Ubuntu foundations (i.e. Kernel and X etc.) to get a native enough experience on almost every type of device. But this reinvention of the user space is a given requirement anyway, as only marginal form factor changes can already change the requirements of the software.

      And the Plasma team is doing a awesome job in adopting their platform to new form facotrs. In fact all of our mobile work would not be possible without the great work of KDE in creating simply amazing software.

      We already saw this change of requirements with netbooks, and quite frankly a netbook is not so different from a desktop. Yet the reduced screen size required more sensible usage of screen real estate and eventually the invention of Plasma Netbook, which tries to meet this use case as good as possible.
      So, building up on the assumption that the workspace is only Plasma and KWin, then it is entirely possible to build a Plasma for devices such as smart phones and tablets and cars (media centers for example).

      In conclusion: We have the ability to make the heart of the KDE workspace user experience suitable for just about every device type and we have a great Linux foundation provided by Ubuntu. Put the two together and I personally do not see why going from desktop to mobile makes a technical difference at all. Especially bare in mind that for the foundations there is no such thing as desktop version anyway, the Kernel, X11 and all that magic is being deployed in embedded setups long before the term smart phone even existed (in fact X11 was thought up in a time when the regular work space computer had less processing power than a cheap smart phone today). What I am trying to indicate is that my original assumption about Plasma and KWin being the crucial components for expanding to new platforms was entirely true ;). What I left out of course is the creation of touch friendly application interfaces. If software was designed in a way that you can do this with minimal effort (and software is going more and more in that direction with the appearance of Qt Quick and what not). BUT, if you do not provide the application developers a platform, then they won’t have much reason to add a touch friendly interface to their software, but if you do you enable greatness.
      The better part of free software is much more scalable than we sometimes realize, so why should we not embrace this scalability and at the same time create a *completely* free mobile operating system.

  8. “Rather, Kubuntu has a ton of potential in exploiting the tablet market”
    Sorry, I meant Desktop market.

  9. It would be *fantastic* if the prelinking feature was implemented for Kubuntu!


    Speeding up Kubuntu should be a priority, as it creates an impression of quality when programs just snap open. I’ve often wondered why on a modern system it still is laggy to open Kate.

    Too bad that it says of this topic, “(Deferred as person behind it sort of disappeared)”. The notes with this suggestion make it sound simple to implement…

    • Simple to implement. But from the discussion I figured it is not terribly easy to get it right (i.e. not to break stuff). Especially due to the possibility of breaking it requires rather a lot of testing on a large scale.
      Also there were some concerns about actually being able to do it concerning what prelink does and how that actually conflicts with default Ubuntu security features.

      But yeah, would be nice to have snappier starting :/ Then again for KDE applications it might not make all that much a difference as there are various similar techniques already in place, what makes startup slow is usually not so much the library loading but rather the actual work involved *before* the GUI is displayed. Something that is perfectly improvable for each application though 🙂

  10. Koffice for Kubuntu please:) With a better MS Office (xls and doc) support. And then you get rid of LibreOffice while saving space on CD and having native Office Suite. With a nice possibility to show off to MSOffice users;) I really hope this is possible…

    • To be honest – I hope that never happens until Koffice has gotten a total reworked interface.

      I have tried to use Krita a bit, but it is not intuitive at all and I can barely get anything done at all with it. And I see the rest of the package has things in the same way.

      There is one thing developers of Linux software needs to remember – just because there is a product on Windows or OSX that is more known and made by a big company does not mean that it is bad and has to be re-invented. They have done a lot of research in getting to the interfaces they have today – learn from it. Don’t put things in places where you would never expect to look. And even if you really think everything should be turned upside down, make it easy for users to switch to a compatibility mode where shortcuts and menu placements mimic the competitor on the other platforms. Then make some strong tutorials that support why the new interface would be better to use.

      Reason is usually better than force. Re-learning is something that scares users and managers. They do not want to spend time on it. So make a version of the interface that have things exactly where people would expect to find it. The let them go for the new, advanced interface later. Bait and switch, anyone? 🙂

      • I love kubuntu its whole layout just makes so much sense. I actually was thinking about switching to kubuntu when 10.10 came out. At the time though I had my Ubuntu install working just the way I liked it and had had it on two computers pretty much the same way (just swapped the hard drive when I bought my current computer in December…) When Ubuntu 11.04 came out with Unity, I ended up switching back to classic 2.3.x until I realized that many of my old favorite themes and packages were broken by Unity. So ticked of was I that I decided to install the kde-full pack through the package manager just to see if I would like it. And for the most part I had no issues, until I decided to show a friend who was interested in Linux what unity or even classic gnome looked like. Both were broken by my install of Kde. Not kde’s fault but Unity’s. This was the straw that broke the proverbial camel’s back. I immediately downloaded an ISO of Kubuntu 11.04 backed up all my files and wiped my hard drive then re-installed. Other than minor issues:

        my graphics card’s driver worked in Ubuntu even with Kde over top, but with just a straight install it only installs the fglx driver but not the catalyst control center and the monitor settings doesn’t like my dual screen setup without the ccc I found this out on Ubuntu even…

        the default KPackageKit – definitely not as good as Ubuntu Software Center though it does show updates separate which I never did see any in USC

        Can’t find a regular package manager (like synaptic) so I can see if I have broken packages…

        I am most definitely loving the appearance, feel, and customization of Kubuntu.

        BUT, the one thing that does kind of irk me is re-inventing the wheel…

        There is plenty of software out there that is awesome and works great. The only LibreOffice program that really needs any work is Draw… from what I understand it is supposed to replace MS Visio but does a poor job… 😦 However, to me I really don’t think that extra time and effort needs to be taken to create a completely new (albeit its not exactly new) software to replace it. Ubuntu (which I have used for quite some time) has plenty of great software that already fills these roles… If something needs work rather than starting over and creating your own, it makes more sense to help out the already existing software and maybe create a theme for it, rather than an entire new program. I don’t mean to offend anyone… this is just my opinion…

  11. I’ll love Kubuntu forever if I can get a working install on my old Winbook M Series from ’03. It has less than a gig of memory and some kind of sis video chipset that isn’t properly supported (regular Ubuntu used the vesa driver, I believe), but I like to keep it knocking around as an “Ohey, if this old thing can run Linux, anything can!” computer.

  12. i wonder when can i get a stable realease of kubuntu with kde hitting it big with 4 series,i used kubuntu 10.4 and it is totally breaking my system into pieces i dont get my mobile for detected as like ubuntu does.also installing codecs is very difficult in 10.4

  13. Hello.
    Are you involved in Kubuntu project, right?
    Please include UFW KControl Module and Quick Usb Formatter by default on Kubuntu 11.10. I find no explanation for these two important components are not included. Especially kcm_ufw that was developed for Kubuntu. In Kubuntu forums an easy firewall and a quick memory stick formatter are two of the most required.
    Thanks for all your work people!.

  14. First sorry for my English, I hope you understand and I understood what you’ve said before.
    >Incidentally not even Ubuntu ships a GUI by default IIRC.
    I hope that this does not mean that Kubuntu users have to wait for Ubuntu first do something good for copying. I thought we could implement good things regardless of Ubuntu.
    Believe me, I’m just an user and UFW KControl Module is very simple to use even for me.
    Anyway UFW KControl Module could be disabled by default for any user who does not want to use.
    Any decent distribution (OpenSuce KDE, Mandriva, Fedora) has a GUI for the firewall.
    >About the memory formatter you may file an idea …..
    Quick Usb Formatter for KDE exists and works very well:
    >(please note that latter needs a proper rationale and sufficient background information … license, source repo, maintainer…).
    Oh, I’m just a user and do not know much about it. Anyway I’ll try.
    Thank you very much

    • Why no, we certainly would not have to wait for Ubuntu. That was just an example that we are not alone in the believe that there is not sufficient need for this. Also the fact that other distributions deem it necessary to have a firewall (GUI) is no reason enough to have one. A rationale would be like: we have openssh-server installed by default and users might want to block the port so that intruders could not abuse an unpatched issue in the server to gain access to the system (which is of course not valid because we do not have openssh-server installed by default, so once again just an example).

      About the bug report for quick formatter… you might really want to go for one of the other two choices. A bug report that does not present a rationale for inclusion will most likely get closed as invalid since it then isn’t a bug. The other two options are much less formal (mostly because they are centered around discussion ^^).

  15. hey thanks, sounds like “K” 11.10 is gonna lose a bit of weight, and dump “reKonq” for Firefox instead.

  16. The “Low Fat” is most interesting IMO, I am dying for performance improvements, though we need to know what exactly was changed or cut out for that to happen!

Comments are closed.