Wednesday, March 28, 2012

KDE Telepathy news

The need for a modern IM client that provided seamless desktop integration in KDE has been clear for some time, so about a year ago, some KDE developers set out to create KDE Telepathy. Since its inception, the soon to become default IM client in KDE was more than just that, it was more about creating a 360º view inside the desktop and integrating real time communication as deeply as possible, albeit providing the foundations for other bits and pieces to be built on top. They call it Real Time Communication and Collaboration, and frame the project goals as follows:

"Real time Communication has traditionally been a detached feature of Desktop Computing, provided via stand-alone Instant Messaging clients with poor integration into the desktop experience. One of the primary goals of the KDE 4 series is to tighten integration between different components of the environment. The Realtime Communication and Collaboration (RTCC) project aims to tackle just this."


Click on image to enlarge

It was not long ago that KDE Telepathy 0.3 was released and it can already be found in some distros, taking over from Kopete. In my experience, KDE Telepathy works fairly well, with an easy and intuitive user interface. It is a departure from the Kopete approach in that it is part of KDE, not so much a stand-alone application you open or shutdown. KDE Telepathy is always there, users simply get to create accounts and manage their connection status through a plasma widget, which makes the whole desktop integration concept more of a reality.


Click on image to enlarge

When users get messages, they will see a plasma notification with the message content and replying is as simply as clicking on it.



The plasma notificiation eventually disappears, but a small icon remains there in the system tray until the user acts on it.



Needless to say, KDE Telepathy 0.3 is a very early release, not yet fully stable and missing several important features, but work is ongoing towards significant improvements. Many of those will see the light of day come version 0.4, as explained by David Edmunson in this POST of his. Some of the highlights include:

  1. Log Viewer: Keeping conversation logs was a feature already available in KDE Telepathy, but with this new release, a log viewer will make its debut. Merging of Kopete logs (relevant for users who were using Kopete) is a work in progress.

  2. Chat Plasmoid: Easily the most exciting new feature in my opinion. By default all conversations will open in this plasmoid (if installed), on top of the notification area. This approach is similar to the one we see in GNOME Shell and ensures that instant messages will not get in the way of what users are doing, offering an extremely convenient way to reply if desired, avoiding switching to another application, etc.

  3. Contact List Plasmoid: Still a work in progress, the team is asking for ideas for a fancier interface.

  4. KRunner Integration: Thanks to a newly created plugin for krunner, searching for contacts and starting chats should be even easier.

  5. Video Support: Probably a feature we all want, but the chances of seeing it there in this upcoming release are slim at best.

    EDIT: Sounds like video will finally make it to KDE Telepathy 0.4. Check this POST for the latest update.

So there you have it, KDE Telepathy is coming together nicely. At this rate, it will be a first class instant messaging client by v0.5. Very much looking forward to that!

Thursday, March 22, 2012

Jonathan Riddell on Kubuntu

Jonathan Riddell was the dedicated developer Canonical funded for Kubuntu. As you probably know, Canonical cut such funding and Jonathan will now work on other projects, but he is still heavily involved in Kubuntu (and will probably be for a long time, to one extent or another).

Jonathan published an extremely interesting post in his blog yesterday touching on a number of topics, from open source software to KDE and Kubuntu. I want to share his article here because I feel very aligned with his point of view, even his open criticism (which I actually consider realism, but judging by some comments I read, some people will think otherwise). There are, of course, some bits and pieces I don't share, but that doesn't change the fact that I consider this a superb post and I wanted it to get more reads.

Should you want to read the original article, you can find it HERE.

Now, on to Jonathan's words...

KUBUNTU AND THE STATE OF FREE CONSUMER SOFTWARE

Since I had to announce that Canonical was dropping support for Kubuntu from 12.11 (and then had to announce two days later they were dropping support for 12.04) I've been getting lots of people asking "is this the end of KDE?"

Of course it isn't, KDE is a vibrant community of people making useful and fun software.

Recently Gnome have been noticing they're not winning either. There is a growing realisation that Canonical dropped Gnome some years ago. [This is melodramatic overstatement, there are still a bunch of Gnome programmes used in Ubuntu Desktop but the workspace and webbrowser and e-mail client aren't. Canonical is also using more Qt and less GTK.] Articles like GNOME 3: Why It Failed don't really help the impression.

All this just highlights that we've been making free software for users for over 15 years and still not got out of the geek market. This comment from mpt, canonical designer highlights some reasons why: third party software, marketing to users but importantly to OEMs and their supply chains, online services, an SDK etc.

...SO HOW CAN KDE REMAIN RELEVANT?

Better design? This was one of the comments from Ettrich when I first met him at a trade show ages ago. We can call it "usability" but design is a slightly broader term of stepping back and working out why certain tasks are hard to do.

New markets? Aaron and his Make Play Live company is making hardware devices with the Vivaldi tablet, that's exciting. That's a small company. Canonical is a large company and may well do the same, will be interested to see if either work.

Fill in the gaps! There is a common meme that "we've achieved what we wanted 15 years ago", well free software in general has but KDE is really nowhere near a usable desktop. We miss a decent web browser, our office suite is looking promising but still isn't much used, the plasma media centre has never got past an alpha stage, Kontact is losing popularity due to a bumpy transition to Akonadi. There's lots to be working on!

App shop? It's what users expect now. Ubuntu has one from Canonical. Muon in Kubuntu is decent and Plasma Active are working on one but they need to link up to third party suppliers.

Server! We should be welcoming in OwnCloud and Kolab to the KDE community. So far we've failed to do that.

Modularisation. KDE Frameworks 5 is a great project, it might mean developers like Canonical start picking up bits of KDE technology as well as Qt.

Take advantage of Qt project. It's there for the using, has some bumpy areas in the infrastructure (you can't download a patch without a whole clone) and social structures but we can help them.

...AND HOW CAN KUBUNTU REMAIN RELEVANT?

Kubuntu has the world's largest Linux desktop rollout. I'll say that again. Kubuntu, often mistaken as a mere derivative of Ubuntu, has more spread than any other desktop Linux. Since I had to announce Canonical moving to focus on Unity I've been contacted by plenty of people saying they rely on Kubuntu. Fortunately Kubuntu isn't going anywhere, that's the advantage of Free Software, when you have a cool community (and we have about the most active community of any part of Ubuntu) then it carries on. We may even find new sponsors.

We like to show KDE at its best. We are the only regularly released distro to ship only KDE Software on the desktop(*), others fill it in with GTK tools or their own config tools, we want to be all KDE. [For the benefit of those journalists who don't understand 'regularly' we ship every six months, just like KDE SC]

(*)There is one bug in the above which is shipping LibreOffice, I think the time is right to move to Calligra, they are doing great stuff and need our support to get it to users. They also are reputed to have better MS Office format importers than LibreOffice thanks to the work of KO.

We must remain part of Ubuntu, they are a great community for distros and we couldn't survive without them. Kubuntu is often incorrectly called a "derivative" of Ubuntu but we are part of the Ubuntu family and we are one of their flavours which is just where we should be.

New markets? Kubuntu Active is taking shape. I'd love to have TV friendly media centre support too for example.

But do we need a new name? Kubuntu has never been a great name, it was actually a joke name made up by the original Ubuntu developers for the KDE side. I wonder if a new name would give us a new lease of life like Calligra has. Suggestions welcome :)

Monday, March 12, 2012

REVIEW: KDE SC 4.8.1

Just a few days ago I upgraded KDE SC to its latest release, 4.8.1. This first dot release is very interesting in that it incorporates a significant number of fixes to elements as critical as Dolphin and KDE PIM. Now that the first round of polishing is there for KDE SC 4.8, and since I have been using it extensively for weeks, I think it is a good time to put together a review and see where KDE stands as of today.

THE GOOD NEWS

There are three main areas where KDE SC 4.8 shines: Stability, Performance and New Features.

Stability: In all honesty, if stability is not there, all the rest is garbage. Because of that, I am glad that KDE developers spent time making this last iteration of their DE even more stable, which is noticeable right off the bat. In fact, this is by far the most stable .0 release I have tested to date.

In my experience during the last few weeks, stability is particularly noticeable around processes and applications that are "core", for lack of a better word. The desktop, panels, effects, system settings, as well as applications like Dolphin, Okular, Gwenview, Kate, Konsole and others, feel solid. Yes, the unexpected segmentation fault still makes the seldom appearance, but significantly less than in previous releases.

Performance: Yup, one of the areas that was historically considered a KDE weakness is today one of its strengths, how about that? KDE SC 4.8 continues to build on the significant steps forward around this area that have consistently been there for the last 3 or 4 releases. Everything feels snappier, faster, more responsive... even Dolphin with its new animations feels faster!

Desktop effects get to a new level never before experienced in KDE using kwin. For the first time, they feel just as fluid as Compiz, but in my experience, even more solid and stable. Moreover, it is surprising how little they impact system resources. Yes, if pure performance and responsiveness if what you are after, you are probably better off turning effects off, but I am surprised of the results I have got on relatively modest hardware.

New Features: Last but certainly not least is the amount of enhancements and new features that come with KDE SC 4.8. I already talked about them in a recent ARTICLE (which I recommend reading in case you haven't already), but let me just talk about those that consistently make my day.

- It was among my favorites already, but with the improvements in this release cycle, Gwenview sits at the very top of my list of picture viewers. Its intuitive and consistent interface is as good as it gets, and its many awesome features just make it even sweeter. For instance, I love its import feature, which is flexible and very reliable (heard that, Digikam?). Throw KIPI plugins on top, and you are in for a treat.

- As is the case with Gwenview, I am completely amazed by Okular. It really doesn't get any better than this. Not only it can chew on a crazy variety of formats, but the fact that it also became a comic reader is the icing on the cake. Things only got better with the improved table text selection features that came with this release. - Like I mentioned in the previous section, the improvements to Effects & Compositing are not only easily noticeable, but also very enjoyable. I believe desktop effects play an important part in making the overall experience more fun, which ultimately has a positive effect on the way I work.

- Although it took me a while to get used to the changes to Power Management Settings, I have to say I am sold. They have been simplified to the point where they are just as flexible, but way more intuitive. Miss the old profile approach? Don't, activities are there to help.

- The improvements to the Semantic Desktop are obvious and welcome. I am particularly happy that nepomuk is finally indexing only when it has to, plus it is clear that the whole thing is less resource hungry than in previous releases. In fact, Sebastian Trüg already ANNOUNCED that further (and very significant) optimisation is on its way to KDE SC 4.8.2.

THE NOT SO GOOD NEWS

Surprisingly, some of the areas that shine in KDE SC 4.8 are also the ones that have more room for improvement.

Stability: Yes, I know, I just said KDE SC 4.8.1 is very stable, but the fact of the matter is that it is only very stable when compared to older KDE releases. Compared to the most solid DEs out there, KDE still feels somewhat fragile. Such lack of stability is most evident around some of the most specialized applications, like Marble, Digikam or Kdenlive.

Here's an example of the Marble routing feature, which not only cannot find a route from Madrid to London, it just happened to bring both cities much closer than they actually are. (Just to clarify, this works perfectly with both Google Earth and Google Maps, so one has to wonder what kind of testing went into this Marble feature if something this basic does not work.)


Click on image to enlarge

Kdenlive and specially Digikam also have stability issues, where segmentation faults or even freezing the entire system are not uncommon.

Unfortunately, that´s not all. There are still gaps around areas as important to KDE as PIM, which simply don't work as they should. As an example, the impact of Akonadi on system resources is still way too high. In my experience, simply starting the Akonadi server makes a huge difference (although the behavior is not 100% consistent across different distros). Here's an example of my CPU activity before and after starting Kmail (which in turn started Akonadi):



There are more examples, but I see no point in listing them here. From my point of view, it would be extremely beneficial to put efforts and resources in place to address these important gaps, so that KDE gets its foundations 100% right. Once the Plasma Desktop is truly a plasma desktop, once the Semantic Desktop truly works as such and is fully reliable, once KDE PIM nails every bit, then it is time to concentrate on the minor details (i.e., Dolphin animations, Marble routing, Digikam face recognition, etc.)

Productivity: I use KDE on a daily basis, have found my workarounds around things that don't work as they should and can manage it reasonably well. It is precisely because I know it that I would not recommend it in a work environment. The lack of stability already discussed is part of it, but also the lack of attention to details that are critical for the average Jane or Joe who sits in front of a computer to get things done. For example, something as relevant in a professional environment as setting up and using a remote calendar in Korganizer can prove very frustrating even under KDE SC 4.8, sometimes even impossible if use of resources is taken into account. Similarly, setting up a remote account using the Kmail wizard is way more complicated than it should be at this stage (GNOME online accounts, anyone?), plus it often fails, forcing users to set up accounts "the old way".

I guess that in some aspects KDE still maintains that "computer geek" vibe to it, like even the simplest things should require a bit of "hacking" for the fun of it. I believe that has got to go if KDE is to have a chance in any kind of professional environment, or even make the cut for users who prioritize productivity.

Overcomplexity: The previous topic kind of nicely leads on to this one, an area that is similarly not getting much love from KDE developers. Yes, we keep getting more and more bells and wistles, but have the basics improved?

I am in favor of flexibility and customization, which is partially why KDE stands out to me like an oasis in a desert of dumbed down DEs. However, I think flexibility and simplicity are not mutually exclusive and I am a firm believer that KDE could be just as powerful using a much more intuitive interface. For instance, it is great to have tons of options and parameters, but how should they be presented to users? At the moment the System Settings tool has an excess of 30 categories... Is that really the best way to go about it? Unfortunately, aside from some very good ideas around power management, KDE SC 4.8 is not addressing this area.

A SOLID STEP FORWARD

Don't let this last bit of negativity confuse you, KDE SC 4.8 remains a successful release by any standards. Miracles don't exist, and it would be unreasonable to expect it to solve all issues in one go. Having said so, KDE SC 4.8 does make significant progress in several important areas, such as performance, plus it adds many interesting and powerful new features. The great thing is that KDE is relentlessly improving and it´s great to know that work is already ongoing to address some of the issues I discussed above.

If you are using KDE SC already, you may be wondering if the jump to this new version is worth the effort. In my opinion, specially if you are on 4.6 or 4.7, it really depends on whether the new features in 4.8 make a difference to you. If moving to 4.8 does not imply risks, or if you are on an older version of KDE SC, then I definitely recommend it!

Tuesday, March 6, 2012

Fedora 17 new features

Not long after discussing the new features that will be part of the latest Kubuntu release, I wanted to take a quick look at another very popular distro upcoming release. Fedora 17, codenamed "beefy miracle", should go live some time early may 2012 and it incorporates a number of interesting features that should build on the fantastic foundations and heavy improvements brought by Fedora 16.

The complete list of new features, as is the case with every release, is published by the Fedora project itself. Along with explanations for each feature/component, there is information on the progress of that particular portion of the project. All of that information and more can be found HERE. If you are not interested in that level of detail, here's a short summary:

  • GNOME 3.4 and KDE SC 4.8: Fedora 17 will rebase to the latest from the two most popular desktop managers in Linux. Both of them bring significant improvements/enhancements, so in both cases it is already interesting to try Fedora 17 just to enjoy the latest from each.

  • Firewalld: firewalld will become the default firewall solution. The services iptables, iptables-ipv6 and ebtables will be replaced by firewalld, which should help simplify firewall management.

  • GIMP 2.8: Fedora 17 will incorporate the very latest from the best open source image manipulation program. There are plenty of nice things coming with this release, but probably the one most people is after is the single window mode.

  • GNOME Shell for everyone: GNOME Shell will no longer require graphics acceleration in Fedora 17. This is obviously great news for those whose hardware couldn't quite keep up with the original GNOME Shell requirements.

  • EXT4 enhancements: BTRFS may be close to becoming the default file system in Fedora, but EXT4 is not being left behind. Fedora 17 will support EXT4 file systems larger than 16TB.

  • NetworkManager Hotspots improvement: Will provide a smoother Internet Connection Sharing/Hotspot feature that actually works consistently and reliably.

  • Single password quality checking: Fedora will have a single point where the system password quality limits are configured, thus reducing complexity and increasing efficiency.

  • Enhanced KDE Plasma Integration: Users will be able to install Plasma widgets (plasmoids) written in scripting languages through "Add widgets… / Download new widgets…" and automatically get the required Fedora packages for those widgets to actually work installed. This gives users convenient access to upstream's huge widget library without manually having to install dependencies such as plasma-scriptengine-*. It looks like Fedora will be the first distribution to offer this convenient Plasma feature, developed by Fedora.

  • Move all to /usr: Fedora is again trying to reduce complexity with this move. While symbolic links will be maintained to ensure backward compatibility, RPM package content will no longer be split in many different system folders, but unified under /usr. This change will have many benefits, such as increased compatibility with other UNIX/Linux systems, a simpler and cleaner system layout, no confusion about tools install locations, no $PATH fiddling, etc.

    Moreover, this change gets things ready for the adoption of BTRFS and its ability to capture system snapshots at any point. In addition, along with some changes to systemd (which inherits more processes, this time from SysV), this change should also help reduce boot times.

  • Memory and Power management: Benefits in these two areas will land in Fedora 17. Thanks to tuned, more flexibility and capability around power profiles will be added. Memory optimisation is clearly benefitial to all users.

And the list goes on and on... Developers will get (as is usually the case in Fedora) updates to many different programming languages and compilers, while system admins will surely enjoy the extensive list of new features, updates and improvements around servers and virtualization.

Very much looking forward to Fedora 17!

A new twist to PIConvert

Some time ago I shared a small and simple bash script that was helping me mass manipulate images. Its main purpose was to help me convert all the screenshots I had for my review articles into JPG format, including a thumbnail for each. However, it didn't take long before I started adding features, such as converting to sepia and B&N, resizing to Blackberry and iPhone screen resolution, etc.

I created PIConvert under GNOME initially, using zenity and imagemagick as the foundation for my script. The script was quickly put together and so the code was far from neat and tidy. Moreover, I realised there was a strong zenity dependency, which not all KDE distros would necessarily support since they already include kdialog. As a result, I decided to revisit my script, strip down some of the unnecessary complexity, improve comments for easier understanding and get it to work with kdialog.


Click on image to enlarge

In the screenshot above you can see how I added PIConvert to the favorites section of my menu. After clicking on the launcher, the following menu appears:



The idea for the most part is that regardless of the option chosen, PIConvert will ask for a few files to convert, as shown below.



Once selected, it will ask for a location to convert them to, again shown below.



Finally, the script exits nicely when done.


Click on image to enlarge

Easy enough, right?

I am happy with the end result because this version of PIConvert now works perfectly natively under KDE, plus the code is simpler and easier to understand. If you are interested in using this version of PIConvert, simply download the script from HERE, review the code to ensure it is safe (always do this whenever you download any kind of code you don't trust!), then grant execution rights and give it a go!

Note that this version of PIConvert depends on imagemagick (it uses the convert command) and obviously kdialog.

Friday, March 2, 2012

Kubuntu 12.04 new features

The Kubuntu 12.04 Beta 1 release announcement is still hot from the oven, and it gives both a glimpse about what we should expect from this upcoming release, as well as about the Kubuntu future itself.

Jonathan Ridell posted: "...And for anyone worried about the future of Kubuntu, Kubuntu 12.04 to be Supported for 5 Years reaffirms that we will be treating 12.04 like any other LTS, only 2 years longer. It also affirms that we will be continuing Kubuntu in the same way I have run it for the last 7 years, as a successful community made Ubuntu flavour."

This comment goes on top of many others from the Kubuntu community to clarify that the project is not struggling and that no concerns appear in the near future.

Great news!

PRECISE FEATURES

Kubuntu 12.04, just like its older brother, was code-named "Precise Pangolin", kind of metaphorically conveying the fact that most efforts will go into stability and precision, not wild or revolutionary features. I am particularly excited about this, because both Ubuntu (specially around the Unity area) and Kubuntu (nothing as big as Unity, but some areas could use some bug fixing) will benefit from this approach. In fact, I think all Open Source projects should slow down every once in a while to dedicate one or two release cycles to stability and bug fixing.

Does this mean that this release is void of exciting features, plain boring? Not at all, there is plenty of stuff to look for. Here´s a brief summary:

  • KDE SC 4.8 rebase: Kubuntu will introduce the latest from KDE, it´s SC 4.8 release, most probably 4.8.1. This in itself already means tons of new features and enhancements, not only from the previous KDE SC 4.7 series, but also from KDE SC 4.8.0.

    A recent post in Planet KDE mentions 50 bug fixes around KDE PIM that will make into 4.8.1. Similarly, other post lists a wide number of Dolphin bugs that are gone for good. This is awesome news, because KDE SC 4.8.0 is the most stable .0 KDE release I have tried to date, so further stability can only enhance an already great experience.

  • Telepathy KDE: The latest KDE IM client will become default in Kubuntu 12.04, meaning Kopete will likely be left out. This is a good way to help Telepathy get the testing it needs to become the great IM client KDE deserves.

    EDIT (06/04/2012): According to Dave Edmunson, a last minute decision was made to stick to Kopete until KDE Telepathy is a bit more mature. I believe Kubuntu 12.10 will most probably default to KDE Telepathy.

  • Amarok 2.5: The latest from Amarok, which incorporates a wide range of exciting new features, will also be part of the default Kubuntu install in about 2 months.

  • Plasma active: Plasma Active is now available to install from the archive for tech preview. Work is ongoing around a Kubuntu Active remix.

  • Rekonq and OwnClowd updates: Rekonq 0.9 is already available and will be once again the default web browser. I hope the new features and bug fixes make it a good enough alternative, so I don´t have to install other more popular names.

    OwnClowd 3 lands in Kubuntu 12.04, with a bunch of awesome features. Check it out!

  • Calligra: In a move that I hope most KDE distros follow, Precise Pangolin will default to Calligra as its de facto Office suite. It is a brave decision, for certain elements in Calligra are still under heavy development and/or polishing, but probably the best way to get it the amount of testing it requires to continue improving.

  • Muon 1.3: The great Muon Software and Package manager get fixes, enhancements and a facelift.

  • Oxygen-GTK 3: GTK 3 apps will now look as native as possible under Kubuntu.

Not bad, huh?

One of the things I was expecting to see in this release was KDE Light-DM, which I believe Kubuntu may default to at some point. For the time being, this piece is only ready for testing, as explained in this POST by David Edmunson.

For some time now I have come to appreciate Kubuntu more and more. I used to consider it a poor KDE implementation, lacking in terms of consistency, stability and performance, but a vast improvement took place in the last few releases. Precise Pangolin looks like another solid step towards setting Kubuntu as a reference in KDE distros. If only those who discarded it a few years ago gave it another chance...