I interviewed Ikey Doherty a while back but haven’t checked in on Solus, his distribution since then. Linux Voice/Linux Magazine recently did a write-up of it, though, and it looks very interesting—a curated, rolling release, meaning you get the latest and greatest software, but presumably without your system breaking, since a maintainer is watching what gets updated (and when it’s safe). For someone like me, who doesn’t want to worry too much about updates, it’s an intriguing option (although OpenSUSE Tumbleweed has been, knock wood, very stable so far for me).
This is interesting because SUSE stepped in to hlep with hardware. I think a lot of companies are interested in the idea of rolling releases.
Graeme is the person behind FocusWriter, a great writing tool. Like a lot of people I interview, he appreciates the flexibility of Linux, in terms of having options for different types of programs. Some people don’t like to be locked down to a single tool on principle and some people just like to constantly experiment with different pieces of software. I’m always amazed how a new interface can make me think differently. But that could also be how I rationalize that I’m constantly trying out new text editors.
- Who are you, and what do you do?
My name is Graeme Gott, and I like to write open source software. My two
most popular projects are FocusWriter, a word processor, and Whisker Menu, an alternate menu for Xfce.
- Why do you use Linux?
I love the freedom to tinker with everything in my computer setup, from which C library is used all of the way to which window manager is active. The flexibility is incredibly liberating. I also prefer repositories of packages over app stores where every program is an island and needs to include all of its dependencies. And I strongly agree with the philosophy of sharing, upon which free and open source software is built. Also, frankly, after using Linux for 13 years, the other options feel too constricting.
What distribution do you run on your main desktop/laptop?
I use Arch Linux on my desktop. I used to distro-hop a lot, but I settled down after switching to Arch around six years ago. A rolling distro is an absolute must for me, and the package format is so simple that making custom packages is really quick. Arch can be more complicated to use at times when a major change comes down the pike and breaks things, but I don’t mind getting my hands dirty with the guts of my system. To me, the idea of an arbitrary delay to update all of the programs on my computer at once, instead of when they are released, feels bizarre.
What desktop environment do you use and why do you use it?
I alternate between Xfce and KDE, but I end up using Xfce more than KDE right now. The current Plasma release has too many sharp edges that I kept running into, so I don’t use it as much as I used KDE4. I used Window Maker for the first few years after switching to Linux, but I have grown accustomed to a more integrated setup and I usually find straight window managers too plain now.
What one piece of software do you depend upon with this distribution? Why is it so important?
It would be hard to say I depend upon one specific program. Most of my work is done with programs that can be easily replaced with another, such as text editors and compilers, or with small tools I write for myself. I tend to switch between different text editors and terminals on a regular basis, although I vastly prefer KWrite to the new gedit.
The one program I couldn’t live without and that I don’t switch around is actually completely unrelated to what I do on a daily basis, and that program is KMyMoney. I haven’t found any other personal finance programs for Linux that I like anywhere as much, and I don’t feel like writing my own.
What kind of hardware do you run this setup on?
I use a home-built rig with a quad core AMD CPU, Nvidia video card, and 16GB RAM.
Will you share a screenshot of your desktop?
Sure thing. This is my rather basic work setup.
Interview conducted December 6, 2014
From the fake Linux trends file: “Rolling releases are hot!”
There have always been people who use rolling releases. There will always be people who don’t.
This is interesting:
“The openSUSE Project today announced the development version of openSUSE (called Factory) has become
an independent distribution using the “rolling release” development model. This means Factory is no longer just the
development branch of openSUSE – the reliable, modern and easy-to-use multi-purpose Linux operating system – but a
tested and stable fresh-daily bleeding-edge distribution.”
I haven’t used Factory and I’m not sure how it differs from Tumbleweed, but the idea of a very stable rolling release that has more recent packages than Debian Testing is pretty intriguing.
Like a lot of people, I’ve been watching Manjaro rise in popularity (at least according to Distrowatch and Internet chatter). Manjaro, a fork of Arch Linux, is often divisive, one of those distros that people either love or hate, with not much in-between.
Building upon Arch is a bold move, given that it’s a philosophy as much as it’s a distribution. Arch is deliberately complex in order to give users the most control over their system. Manjaro’s goal of simplifying Arch can be seen as compromising that philosophy. But given Manjaro’s popularity, it’s filling a need for users who want a simpler Arch implementation — even at the cost of control over their system.
I played with Manjaro for a few weeks and ultimately, it wasn’t for me. But after exploring the distro and researching it, I was curious about the project, so I reached out to the project leaders, who shared my questions with some other Manjaro team members.
Manjaro’s ultimate goal seems to be creating a user-friendly, stable, rolling distribution. It’s a bold goal, but one that speaks to a lot of users.
- What is your name, title, and role within Manjaro?
Philip: My name is Philip Mueller. I’m one of the project leaders of Manjaro. I work with a lot of aspects of the Manjaro project.
I created ManjaroISO, our install-media creation tool, and work on MHWD (Manjaro Hard Ware Detection tool), which simplifies the driver installation for all devices. One of my biggest jobs is package and mirror management. We have a tool called Boxit which gets all needed packages from Arch Linux and merges them with our own packages on our servers. Also I’m involved with the support team and answer as many questions I can.
handy: Support team member. I provide support where I can to forum users and write tutorials and wiki pages.
Verändert: Support team member. I provide support where I can to forum users.
Quantum is a former support team member.
- How would you describe Manjaro to a new user?
Philip: Manjaro is a user-friendly community around an Arch-based distro. You find people who use and share their experience with and about Manjaro. Everybody is welcome to find his place in our community. And yes, it is an OS for beginners willing to learn and those who want to have full power off their system.
handy: Manjaro uses the Arch rolling release package management system which I consider to be the best package maintenance system available for desktop computer users. It enables a user to install once and then upgrade their system daily if they so choose. This system obviates the need to re-install their OS until their machine has the type of hardware failure that requires re-installation (most likely a HDD failure).
Manjaro uses the packages from the Arch official repos which go through a relatively brief testing and integration phase, where the packages from the Arch repos are combined with the Manjaro specific packages; adjustments are made where required. The packages quickly move from unstable to the testing repo, before being made available in the Manjaro stable repos. All of this is being done in an effort to keep Manjaro as stable and reliable as possible, which is of course particularly useful to those inexperienced in the ways of Linux and/or the Arch rolling release system.
Manjaro also uses tools to make its initial installation, including hardware identification and driver management, much easier than the Arch Linux method. This area, amongst others, is still in the process of being developed and improved.
- What are the plans for Manjaro?
Philip: One of our plans will be to finish our graphical installer, which we are still coding on. The current one is borrowed from Linux Mint.
Also we are working to optimize package management. Graphical tools are planned to improve the user experience with Manjaro. A start is our Manjaro Settings Tool. We will add more plugins in the near future. ManjaroISO will get a nice graphical interface so people can simply click their own system in a few steps to have a custom spin of Manjaro for their own needs. So there is a lot to come before we have a “final” Manjaro 1.0 release.
Verändert: To make Manjaro the best operating system there is and eventually take over the world.
- To what extent is it a fork of Arch? How different are the two distributions? How similar are they?
Philip: Manjaro is Archlinux – with extra spice. You can use the knowledge you already have from Arch Linux and do the same as in the upstream project. AUR support is there. You feel home right from the beginning. We added new features like mhwd, a graphical installer, and graphical tools for common uses. Multible kernels will enhance the support of different systems and additional extra modules for special hardware on each series.
You install Manjaro once and simply update it.
Quantum: Manjaro is deeply enhanced Arch Linux. We take care of all the basic configuration, so you can get right to actual work. If you have a background in Arch, you can leverage that knowledge and do anything you can do in the upstream project,
Verändert: The kernel and the really cool Manjaro hardware detection are made by Manjaro developers. The rest is coming from Arch. There simply is no need to reinvent the wheel here: Arch rocks, it just isn’t (and doesn’t want to be) newbie-friendly. That’s where Manjaro steps in. It takes the best base around and adds tools to flatten the learning curve.
- What are the challenges of maintaining a rolling distribution? Is it realistic to expect rolling distros to work for newer Linux users?
Philip: You have to have a concept to maintain the packages in a small team like ours. We use Arch’s packages since they are mostly stable and up-to-date, so we don’t have to worry much on outdated packages. With BoxIt we create snapshots we use to build our own additional packages and merge them to the whole. Those snapshots will be tested by our community before they got moved to the stable branch. So there is a small delay between Manjaro and Arch. You will find updates later in our stable branch than in the upstream project.
handy: The main challenge as I see it is for users to become familiar with how the rolling release system works. That means how to use pacman in the terminal, how to use the GUI pacman wrapper(s). Work is going on to improve the GUI tools in this regard, so people shouldn’t forget that Manjaro still has a way to go before it reaches release 1.0.
As far as the rolling release system and newer Linux users are concerned, I think that at this stage of Manjaro development, some fresh Linux users will find it too hard (especially if they have some kind of installation difficulty). If the user is prepared to learn how the rolling release package management system works (which as previously stated, will get easier in Manjaro’s future), they are very likely to fall in love with it, as it makes their computing life so much easier.
- There’s been some discussion within the Arch community about the stability and security of Manjaro. How do you respond to the accusation Manjaro isn’t as secure or as stable as Arch?
Philip: We are as stable as Arch Linux is. Due our extra testing we might be even more stable than the upstream project. Having a slight delay might give you the feeling we aren’t as secure as Arch might be. I always tell users concerned about that to use our unstable branch, which we update almost daily, as Arch Linux does. You might find some quirks if you do so since we have to solve them first and test them later in our testing branch. You can choose how stable or bleeding edge Manjaro should be for you. You get stability over security on our stable branch since there is a week or so delay between stable and unstable. Also, there was some talk about our install medias using the same signature master key. This issue is solved since the 0.8.6 release.
handy: There was a potential security problem pointed out in February this year (if I remember correctly) which was remedied shortly after. Beyond that, our stable repos being delayed a week or so from those of Arch, is very unlikely to pose a security threat to Manjaro. We prefer to have stability over instability in this regard. If a serious security threat arose that warranted quick action on behalf of the Manjaro package management team, then such action would of course be taken, pronto.
Verändert: I would say that Manjaro is a tad more stable than Arch because packages are tested for another week before getting released from Manjaro. During that week, Manjaro might be less secure than Arch, because a package fixing a security hole might also been withheld. On the other hand, if some new package in Arch is insecure, it can be withheld by the Manjaro developers. That said, I believe neither Arch nor Manjaro are insecure or unstable.
- Why do you think some in the Arch community (and within the Arch project) have had such a strong and negative response to Manjaro?
Philip: As we “borrow” their packages we seem to be lazy packagers in their eyes. Also we do things they never would do — against The Arch Way — making Arch easy for beginners. Arch was always for experienced users. Now there are third-party projects shaking it up and down as they like it, and even being successful with it. Some hate us and some love us. We are different. Not everybody can be pleased. We love Arch Linux but there is always a possibility to change our base. Not everybody likes Ubuntu, Mint, or even Debian, but they all use the same base. We try to give our community an opportunity to do whatever they want to do with their systems and get support from a friendly bunch of people.
Quantum: Some in the Arch community object to making Arch ‘easy,’ ostensibly lowering the bar for users. But our goal and purpose is to create a distro that is based on solid and current software, which is already set up so that you don’t have to spend days and weeks creating config files, installing GUI tools and every single little app and library, and tweaking everything. Our community has drawn users with all backgrounds in Linux, from beginners to the most advanced.
Verändert: First of all, I’m quite sure that the majority of people that develop or use Arch like us as we like them. In the small Linux universe, you will always have some that think that any spin-off is done by leechers that don’t cherish the work of the makers of the original distribution enough. That’s in no way Arch-specific. People just think that the work that is being put into Manjaro should be put into Arch proper. Which is wrong, since Arch doesn’t want to be beginner-friendly. The developers want you to set up your system yourself and learn by doing so. Which is perfectly fine, it just isn’t for everybody. I had used Arch myself and failed miserably when it changed to systemd. I really tried hard to keep my computer up and running and found that I neither had the time nor the knowledge to do so for a long time. And while I’m familiar with the terminal, I prefer using a GUI application. While I’m thankful for the Arch developers’ hard work that Manjaro relies upon, I prefer to have a bit more of stability. That’s why I use Manjaro now.
This is very disappointing. Fuduntu was gaining a lot of popularity when the team pulled the plug on it. Cloverleaf rose from the ashes of Fuduntu but never got off the ground. Rolling distros are challenging. It’s a shame this team couldn’t make things work — ending Fuduntu broke a lot of hearts.
I was pretty sad when I heard Fuduntu was going end-of-life. It seemed like a very promising distro was being mothballed just as it seemed to be gaining attention within the Linux community. I reached out to Lee Ward, who handles communication for Fuduntu, about the future of the distro, and he had some interesting details to reveal, including the idea that the future distro could be a rolling, curated version of OpenSUSE. It’ll be interesting to see what the new distro shapes up to be.
Linux Rig: What are the plans for the post-Fuduntu distro? Any ideas what
it’ll be based upon? Will it be rolling? What will the desktop environment (DE) be?
Lee Ward: Those of us moving on to the new distro have been discussing and evaluating our options. Right now, we are leaning heavily on going with an openSUSE base. Our devs have been working with openSUSE the last few days to see how viable it will be and things are going well. While a final decision has not been made, that is how we are all leaning at the moment. We do plan to continue with the rolling release in the same fashion we did with Fuduntu. That worked very well and we plan on continuing with that. As for the DE, no decision has been made. We’re looking at all the options to see what will fit best for our goal.
What we want to do is keep the same ideals that Fuduntu had alive. We want to be close to our community and be able to offer things that others have decided aren’t important. We want to help keep the low-end systems going and also to help with the gaming on Fuduntu. Many have said that bringing gaming to Linux would be huge. We were one of the first authorized by Valve to redistribute Steam and we think that was a huge step. We want to keep doing that. Keep bringing the community what it wants.
This Sunday (April 21), we will be having a public meeting on the future and the DE will be one of the things discussed. We are hoping to get participation and input from the community. The community was one of the things that made Fuduntu great and we want to include them as much as possible as we move forward. The meeting will be at 3 p.m., Eastern in #fuduntu on Freenode and we strongly encourage users to come in and help us in making this decision.
Linux Rig: Did you look into keeping Fuduntu going using another DE, like Xfce?
LW: The real issue when it came to the DE was the underlying libraries. Several functions had been deprecated in glibc and glib2 without any consideration for backwards compatibility. In addition, Fedora decided to locate gtk2 headers in /usr/include/gtk-2.0 but left the sources default. This meant building GTK2 packages broke due to the header locations being different than they were installed.
Trying to fix these issues was too much for our small team. It just wasn’t sustainable. Our devs actually started working on it to see what all needed to be done and found that as they were fixing one thing, something else would break. The lack of backwards compatibility hurt us.
Linux Rig: Do you regret sticking with GNOME 2 as long as you did?
LW: We do not. While, ultimately, we were not able to sustain it, we are glad that we were able to give something to the community that was wanted when everyone else had abandoned the wishes of a large part of the community. The popularity that Fuduntu began receiving and the rave reviews are, in part, because we were delivering what was requested. Unfortunately, upstream did not seem to care as much about that and, being a small distro, we were shut out and we had no chance to survive.
Linux Rig: Fuduntu seemed to gaining popularity right as you announced it was
going EOL — do you think it’ll be hard to regain that momentum with a
LW: This is a really hard one to answer since it all hinges on speculation. We’ve heard, in a few places, that people will be keeping an eye out for the new distro. We also have some time. We still have one more Fuduntu release and we still have five months before Fuduntu shuts down. We have an opportunity to say, “Fuduntu is closed, but we’ve got the first release of the new distro ready!” We’ll be able to work on trying to get the new system up as well as packages going and such to the point that we can try to smooth the process out as much as possible. Obviously, there may be some hiccups but we’re going to try to minimize that as much as possible.
One of the important things we want to remind people is we haven’t stopped supporting Fuduntu, yet. Our support team is still dedicated to working with people to get issues resolved and our developers and packagers are still dedicated to getting fixes out there as soon as possible. Asking people to reinstall will be rough, but many other distro users are used to reinstalling every time there’s a new release. We’ve been able to keep it as a rolling release for a long time and, even though this would be a new install, it’s the first time in a while where it’ll be required.
All this to say that I think we have the opportunity to get the momentum back. It’s going to be a lot of hard work, but we’re dedicated. Andrew Wyatt brought the community a very stable distro that was what the people asked for. We want to keep that going and we think we can. While I do expect a small drop, I think we’ll be able to get it back and we’ll be able to show that the new distro is as dedicated to stability and the community as Fuduntu was.
Very interesting post on making Ubuntu more of a rolling(ish) release. Rick makes some very strong points.
This is incredible news. Linux Mint Debian Edition, which is powered by Debian Testing, is essentially going to build a new repository, that will represent the latest from Testing, but also packages that have been tested and/or fixed.
This is going to take a lot of the risk out of Testing (especially toward the beginning of a development cycle, when Testing isn’t as solid as it is toward the end of the cycle).
I’m hoping to buy a new laptop by the end of the summer. Linux Mint Debian Edition (the XFCE variant, of course) is going to be my next distro. I was sold before this project, but now I’m insanely psyched.