Since we’ve written several posts recently about open source communities, let’s highlight one more example of community members seeing a problem and trying to solve it.
KDE is the oldest graphical desktop environment for Linux, and I’ve used it since the day I installed Corel Linux in 2001 (forgive me if I’ve offered those two facts a hundred times before). It’s a big, complicated software collection (with 300+ software repositories), now undergoing its third major overhaul to KDE Frameworks 5 providing the technical underpinnings of the accompanying Plasma 5 Desktop. In all that time, there are going to be bugs that remain unsolved, and applications that grow stale.
Enter the gardeners
Spanish KDE developer Albert Astals Cid came to the annual Akademy conference with an idea: Put together a team to name and find people to fix longstanding bugs and important, but unmaintained projects. What became the KDE Gardening Team.
The Gardeners are different from the project’s quality assurance team, though it chooses a “Bug of the Month” that needs some attention. It’s really kind of a triage or rescue squad for KDE applications. As described in both Cid’s introductory blog post and the Gardening Team’s main page:
The mandate of the team is to:
Find *really* important bugs and ping people to fix them
Find stale reviewboards and ping people to review them
Those bugs often raise endless discussions from frustrated users about how KDE developers do not care. The truth is, most developers are not even aware of them, because the issues do not happen on their system.
The current “Bug of the Month” is a fun one, dating back to 2011, with 65 comments: “When I opened my laptop from sleep, and … logged in and saw my desktop this crash report was there.”
First sign of progress: K3B has a new update
The Gardeners’ first “love project” revived the venerable CD manager, K3b. Version 2.0 was originally released in 2010, and v2.0.2 came out a relatively short time after that. Since then, developers had worked on v2.1, fixing some bugs plaguing existing users, but never getting released.
After the Gardeners’ applied some love to the project, K3b v2.0.3 came out a few days ago!
Next in line for some love is KRecipes. This recipe manager works pretty well by all reports, but was last released in November 2010. Incidentally for any technical writers reading this: the KRecipes Handbook (user guide) is not yet complete for the KDE 4 version of the software. Should you be inclined to help, see the current text here.
Once this project makes progress, KTorrent is likely the leading candidate for the next Love Project.
I’d like to spotlight other communities’ smart activities here at Notes from the Metaverse in the future. If you’re participating in something cool, or know of a similar project to the KDE Gardeners, let me know, either by email, or commenting on this post.
Tonight is update night, when I open up all my virtual machines and get all the latest and greatest software. Back in dialup days, I updated my Linux partitions on Saturday mornings. Nobody would call me, and I figured the remote servers (especially the openSUSE servers in Germany) would be less stressed than during the weekday, speeding the download ever so slightly.
One advantage of having a terabyte of storage on my laptop is that setting up new “systems” is incredibly easy. VirtualBox can set up 150GB hard drives in a few seconds, and installing a new Linux OS with a set of default applications on that empty drive takes about a half hour. So I have too many machines, and clearly not enough time to use all of them effectively. Trying to figure out what to keep. Here’s my current list of client operating systems:
openSUSE Tumbleweed (32-bit): This one is my “everyday” Linux system, but not for much longer. When I first created this VM a year or two ago, VirtualBox didn’t support 64-bit client operating systems. Now they do. Another reason to junk this one: Until now, Tumbleweed represented a stable rolling release. With openSUSE 13.2, it remains a rolling release, but with a few more cutting edge apps that may not be quite as stable as the old Tumbleweed.
Ubuntu 14.10 (32-bit): This connects with my Nexus 7 tablet dual-booting Android and Ubuntu Touch.
Kubuntu 13.10 (32-bit): I should try one of the other flavors of Ubuntu, but what can I say? I’m a KDE guy. I originally installed this after the kerfuffle over Canonical not paying Jonathan Riddell to work on Kubuntu anymore. I wrote about that here and here too.
Ubuntu 14.04 Test: I think I broke this one.
openSUSE 13.1 (KDE:Current, 32-bit): This has unstable KDE apps available, but I think I broke this one too.
openSUSE 13.1 (KDE:Current, 64-bit): May become the ‘new’ everyday system
openSUSE 13.1 (KDE Plasma 5 Preview): This is the next version of KDE, not yet ready for prime time. See Post #201 for that history.
Kubuntu KDE Plasma 5 Preview: See above.
openSUSE 13.2:This is a clean install of the latest openSUSE, and is the other candidate for “everyday” system. This version of the distribution supports seven desktop environments. I want to get them all installed and play with them a bit.
Kubuntu 14.10 (64-bit): Did I mention my affinity for KDE?
It’s all fun, and all good. What does your system look like?
While writing yesterday’s milestone post, I realized I failed to mention the most popular post (by far) in the history of this blog, also known as the day I was linked to by DistroWatch Weekly.
Past as prologue
Curiously enough, that post topic echoes Tuesday’s post on the openSUSE 13.2 release. You see, openSUSE 11.0 was released in June 2008 to much fanfare, in part because the new version of the KDE desktop was included in the release. KDE 4.0 was what we now commonly call a “technical preview,” not really ready for production. Nonetheless, openSUSE allowed users the choice to install the “old” KDE 3.5 desktop, the new KDE 4.0 desktop, or the continuing GNOME 2 desktop. Despite some warnings that KDE 4 was not quite ready for everyday use, some folks installed it anyway. Pandemonium ensued on the support lists.
In this post, “openSUSE 11.0 and KDE 4,” I offered my take on the “crisis.” In brief, people needed to think about their systems before installing major new components. There was blame to be shared, but just because Linux folks were used to working with applications with version numbers of 0.4 didn’t mean that everything would always work perfectly.
Apparently someone at the DistroWatch site (a place to learn about Linux distributions) thought I had something to say, and two days later, hundreds of folks were dropping by. The post sparked some interesting conversation in the comments as well. For a blogger, things rarely could get better.
Eventually, KDE 4 became eminently workable, and pretty darn cool. While some folks never got over the shock of the bad rollout (and still complain about how terrible KDE is now), KDE development continues apace. Sometime in the last few years, the Trinity project launched to recreate KDE 3.5. I’ve heard they found it difficult to reproduce on modern systems. So it goes.
KDE Plasma 5: Don’t say you weren’t warned
Much has changed in KDE since 2008. With openSUSE 13.2, another new KDE desktop, based on the KDE 5 Framework is included with the distribution. But it’s not one of he options in the install. Many will set it up separately to play with, but much of the code won’t work side-by-side with KDE 4. A lesson learned.
As the month of November continues, I’ll have more to say about all these topics (openSUSE, KDE 4 and KDE 5). I’m still planning to try all the other supported desktops. Maybe I’ll have a new favorite by the end of the month. Stranger things have happened!
Got any memories of the KDE 4 rollout? How about GNOME 3, which didn’t go a lot better for some users? Let me know in the comments!
The road to Plasma Workspaces 2 has been laid out as the Plasma developers recently met in Nuremberg, Germany, to discuss their open issues around future developments – it will be based on version 5 of the KDE platform and Qt…
Mike McCallister‘s insight:
Key points I take away from this: KDE 5 Plasma Workspaces will be designed to function exactly as the current KDE 4 does.
To do that, it will take at least a year of development. Thus we should see a production-worthy release in late 2014.
Penz originally wrote Dolphin in 2006 as a “small and fast” version of the Konqueror file manager that shipped with KDE 3.x and earlier (as he explains in this blog post announcing his “retirement”). The KDE development team soon invited him to join them in working on KDE 4.0. Since that release, Dolphin has been the default file manager for the KDE desktop (now known as the KDE Software Collection).
In his June 26, 2012 post, Penz offers a few reasons for leaving the Dolphin/KDE project:
He really wasn’t expecting to be working on this project for six years running (always a fair point)
“The time required to keep Dolphin in good shape increased during the last years. I’m doing this project in my spare-time and usually have spend around one evening per week on Dolphin. Especially during the last 2 years this time has increased.”
As a user, he suggests that desktops from Apple and Microsoft have now become at least as “efficient and comfortable” as the KDE desktop. KDE is “not competitive anymore.”
To my mind, however, it’s the point about the complexity of development that may prove more significant for the free/libre/open source model generally, and certainly not just KDE. He writes:
The user interfaces tend to become simpler and easier to the eye, while the functionality of the application itself has increased. Hiding a complex functionality behind an easy to use interface are not known strengths of “typical” developers 😉
The complexity of the non-user-interface-parts of applications has increased a lot. Web-browsers are a good example: While the interface got simplified during the last years, the engines showing web-pages got really complex and are maintained mostly by fulltime-developers in the meantime.
Now beyond a little PHP code now and again, I’m no software developer. I just explain the fruits of software developers to the rest of us mere mortals. But it’s long been a truism in the industry that “the simpler the interface, the more complex it is underneath.” Penz is restating that truism here.
Earlier in the post, he talks about the next big project he sees for his brainchild: giving Dolphin a face lift using a new “view engine” called QtQuick2. The problem is that “porting Dolphin to this components will be a very time-consuming and boring task: All the settings-pages, the URL-navigator, the information-panel, the search-interface, the tooltips, … – this is just not doable anymore in my spare-time.”
Maybe what he’s really trying to ask here is whether there are any more “fun” projects for the volunteer developer. If there aren’t any such projects, there’s a reason to worry about the long-term health of the “scratch a personal itch” FLOSS model.
Working on the non-user-interface parts of applications can be challenging and this is not something that most freetime-contributors are striving for. But if there are not enough contributors for the complex stuff behind the scenes and if no company is willing to invest fulltime-developers to work on this… – well then we are losing ground.
Penz admits that “Probably my explanation/guess/theory is nonsense and utterly wrong.” Maybe I’m reading too much into this as well. One reason to doubt the theory from the beginning is that one reason that openSUSE 12.2 is going to be late is that more volunteers are showing up and making contributions.
As a user of free and open source software, I hope there’s still room for the coding hobbyist in our movement. Please tell me I’m wrong, or losing my mind. Or answer these questions in the comments:
If you participate in free-software development, especially for desktop apps, is it still as much fun as when you started?
If you’re just starting out in free-software coding, is it easy to find a project that’s both fun and challenging?
The good news is that Blue Systems has been sponsoring a variety of KDE projects and distributions in the last few months. They’ve pledged to keep Jonathan Riddell on its payroll working on Kubuntu (or whatever it may be called in the future), and is offering marketing support too. But it’s hard to know from this distance how much money they actually have to back those pledges up. A WhoIs search on the Blue-Systems.com site pointed to a German reseller, http://itwu.de/, as the owner. That’s pretty much all outsiders know.
In some ways, the news is not that different from when Attachmate surfaced last year as the company to buy the SUSE brands from Novell. People rightly questioned what the company had planned for the distribution. So far, it appears that Attachmate has largely left the community alone to make its own plans. Plus openSUSE enthusiasts in the Americas now have their own conference to attend this fall. So I think we can say that up to now, that deal has worked out pretty well.
As I said before, best wishes to Riddell and the Kubuntu community. I’m confident this is good news for my favorite desktop environment, and Linux overall.
If you know anything more about Blue Systems and their work with the community, please let me know. Whatever you think about Kubuntu, KDE and its future, feel free to comment too!
You may have heard that Canonical is formally dropping support of the KDE-based version of the Ubuntu desktop come October. This is kinda old news now, but it seems that at least some folks want to make a big deal of it. So I’m feeling the need to talk about it too.
Today I bring the disappointing news that Canonical will no longer be funding my work on Kubuntu after 12.04. Canonical wants to treat Kubuntu in the same way as the other community flavors such as Edubuntu, Lubuntu, and Xubuntu, and support the projects with infrastructure. This is a big challenge to Kubuntu of course and KDE as well.
A few days later, Riddell’s counterpart at openSUSE, Will Stephenson responded to the situation on the KDE Contributor’s Blog in a way that may seem a little cynical, but struck me as a completely valid response.
After a week had passed, TechRepublic’s open source pundit Jack Wallen said he read Riddell’s announcement “with a heavy heart.” The money quote here:
try to find a major Linux distribution that ships with KDE as the default desktop. You’re going to be hard pressed to do so.
He said the best way for KDE to survive this blow was to develop its own distribution, which he named KOS. In the poll accompanying the story, 51% of his readers agreed with that strategy.
Bruce Byfield at Datamation (a fine writer for a variety of Linux publications) may have overhyped Wallen’s article just a little, and turned it into part of a wave of “KDE Death Watch” commentary. The story does effectively dispute the idea of KDE disappearing, but does again raise the question of “Just what is a major KDE distribution.”
Younger folks in the audience may not remember Rodney Dangerfield, the comic who built his entire career on the theme that he “got no respect at all” (Check out the films Caddyshack or Back to School to learn more about the Dangerfield persona). One of the most striking things about all these stories is how Kubuntu is (allegedly) the last major Linux distribution with a KDE desktop. Perhaps I’m biased, but when did openSUSE cease being a major distribution? While it has never been dominant in terms of mindshare or installations (admittedly both hard to quantify), the little green Geeko with the outstanding system administration tool just chugs along.
The SUSE distro has been famously associated with KDE throughout its life. Some historical notes:
Novell bought the original German company that produced the distribution around the same time that they bought one of the main development teams for the GNOME desktop. When openSUSE planned to make GNOME the default desktop on installation, a massive uproar from the user base left the desktop choice to the person doing the installation.
openSUSE was the first distribution to switch to KDE 4. That was certainly a bad idea or miscalculation, but certainly a commitment to the KDE desktop.
openSUSE is not going away, and will continue to be a “KDE-first” distribution for a long time to come. I’m looking forward to seeing how big our community is at the openSUSE Summit this fall.
Byfield certainly makes an excellent case for why KDE is not dead, and certainly won’t be on life support anytime soon. Even Wallen (who doesn’t really use KDE anymore) concedes that “KDE is one of the most polished, professional desktops available for the Linux operating system and deserves to be made available through some official channel.”
I disagree that KDE needs to put out its own distro to succeed long-term. The world probably does not need many more Linux distributions. Linux users should always have a choice of desktop environments and associated applications. I love that I can run apps designed for GNOME on my KDE desktop, and want to continue to do that.
KDE certainly needs to attract more developers, volunteer or otherwise. More users and platforms will follow. As I noted last week, the Spark tablet is a great way to build pathways to the future.
Things may not be entirely rosy for KDE today, but I can heartily raise both hands when Wallen says “Linux without KDE is simply not the Linux I’ve known and loved since the mid-90s.” May that continue to be true.
It’s real: Tablet PCs have arrived. According to a recent DePaul University study, one in every dozen airline passengers is using a tablet PC or e-book reader at any given moment.
Like many of you, I got a tablet (a Nook, if you’re interested) as a gift this last December (thanks Jeanette!). It’s pretty nice. I read Wired on it now, check news, post tweets occasionally. But it’s moderately frustrating that I can’t really do anything worthwhile on this machine.
The problem with tablets is that they are designed for consumption: of movies, books, websites and the like. People want to be productive while on the go. The size and weight of the average tablet is perfect for productivity almost anywhere. But the software isn’t there to support a productive worker.
What if there was a tablet with a real operating system and a collection of software that lived in the tablet (not in the cloud)? What if you could work on a presentation without worrying whether your carrier had an affordable wi-fi connection today? What if you could then use a USB port to plug your tablet into a projector when the time came to deliver that presentation? At under $300, that’s a purchase even a cash-strapped employer could justify. This machine is the Spark.
Project developers are working on building an app store, but you’ll also be able to use the Open Build Service (OBS) from the openSUSE Project to obtain apps for your Spark. This is the “productive” part of this tablet, as you could run most (if not all) applications that could run on desktop KDE.
The main initial problem with Spark is that it’s not an especially powerful machine. The 7-inch Zenithink C71 tablet has just a 1GHz processor, 512MB of memory, and 4GB storage space. The display is 800 x 480 pixels. One hopes that future models will have a little more muscle. The good news here is that it has two USB ports and a microSD slot to help you get work done!
This is where I should be telling you how you can get this marvelous device, but I’m late. Thousands of pre-orders at MakePlayLive.com last week reached the capacity of machines able to be built by the May release. You can (and should!) still put your name on the waiting list, though.
For more complete information on this device, and the philosophy behind it, reading through lead developer Aaron Seigo’s blog posts on the Spark is really exciting.
The Spark is a beginning. The prairie fire will hit when more people realize that a tablet doesn’t have to be a toy.
What do you think about the Spark, and open tablets generally? What tools would you like to see in the Spark? What problems are you seeing in the tablets you use? Leave a comment!
Thrilled to bits to report that for the first time in the Americas, openSUSE users, developers and folks who might want to be in those categories will be gathering in Orlando, Florida this September. This community conference doesn’t have a name yet (more on that later), but is sure to be informative and exciting. As with all openSUSE activities, participants will certainly have a lot of fun!
The story is that the annual corporate SUSE conference is happening September 18-21. This is where system administrators, developers and other people who make their living using SUSE Linux Enterprise gather. Just speculation on my part, but I’ll guess that Attachmate/SUSE got a better deal from the hotel if they reserved the entire week. The beneficiaries of this arrangement include the scruffy brigands of openSUSE.
Planning for the event began last Wednesday on Internet Relay Chat, with a dozen or so active participants, including your humble scribe (see the full transcript) (see a summary). We want to make this a conference that is comfortable for both basic users and the developers who make openSUSE the great distribution it is.
The first item on our agenda, though, is naming this first ever conference. Quite a few names were suggested at the kickoff chat, and a poll is being conducted at openSUSE Connect. Choose your favorite before Saturday!
If this conference excites you, you can help make it happen. Visit the conference wiki and sign up for one or more of the task teams.
Watch this space for more news as things move forward.