DistroWatch Weekly

CL: We answered many of the requests we received after the release of Linux Mint 7 and some of the changes we made were quite popular among our users. The Update Manager now allows you to ignore updates for certain packages. The level associated with each package is something we maintain so this addition gives a lot more power to the user. We also improved many aspects of the Software Manager and we implemented numerous little things to make the system more comfortable to use.

CL: Linux Mint isn't just based on Ubuntu, it's fully compatible with it. Unlike what Canonical does with Debian, we do not fork the Ubuntu repositories or break compatibility with our base distribution. We use two technologies to make the most of our package base while remaining independent in our choices and the changes we want to implement: APT pinning (which is well-known to Debian users) and adjustments (which is a technology of our own). When we want something to act differently than it does in Ubuntu we can either maintain the package ourselves or dynamically tell our system to adjust the changes we require. Our repositories are given a higher priority than the other ones, ensuring that Linux Mint users pick the versions we maintain rather than the ones coming from upstream.

Developing our own repositories represents a massive amount of work. If you look at the various distributions on the market you'll notice very few projects which have the resources to both maintain their own packages and develop new innovations on a regular basis. I can think of only a few, and these are backed with corporate funding - with a business model that usually requires them to shift their focus away from what matters to home users and onto more lucrative activities such as business support.

Of course, with our own repositories we would become more independent. I would personally like to slow things down and to be more conservative on the base of the system and when it comes to hardware detection, to ensure more coherence and less regressions between releases, but we're more than happy with what's done upstream, whether it comes from projects like GNOME, KDE, the Linux kernel team or even distributions like Debian or Ubuntu. Every six months our own features shine on top of a large amount of upstream improvements and the result is fantastic. If our goal is to get to a perfect desktop then we're only introducing change when we think we can do better. There's still much to do to improve the desktop and so it's not the time for us to focus away from this. We leave the system to upstream projects and we don't feel the need to introduce changes in that domain.

On the things we do want to change, APT pinning and our adjustment system give us the flexibility we need and so we don't need to duplicate and/or patch every single package in separate repositories.

With regards to server loads, both Linux Mint and Ubuntu are mature projects with mirror networks. For the distribution, a network of mirror hosts is very important. It makes it easy for people to download and use the operating system, it reduces the load on each server and it improves the overall performance for each user. For the mirrors, this is also very important as once they have the resources hosted locally, they can offer them easily to their own audience. Let's take a national ISP as an example. If many people in the country download and use Linux Mint, that creates significant bandwidth and requests from this country to our servers in Germany or Ubuntu servers in the USA. It's in the interest of the ISP to mirror both Ubuntu and Mint so that the local users find the same resources locally. To the distributions, that means fewer things to worry about. To the users that means local servers, to the ISP that means less outgoing requests.

I wish I could answer the question more briefly but there's so much to say about the hosting strategy. To summarize, there's no real advantage in maintaining our own repositories at the moment as it doesn't represent any significant issues when it comes to server loads or our independence as a distribution and it would require a lot of work, attention and focus which would inevitably be shifted away from what matters the most: improving the desktop.

CL: The project was independent from the very start and although the system itself technically qualifies as an Ubuntu flavour (since it's both based on and compatible with it, and since the base system is almost the same) the distribution itself, in terms of direction, structure and ways of working is completely different. We consider Ubuntu as an upstream component and, as prominent as it is within the end result, it's still something we consider as a part, which can be changed, modified, patched and configured to fit in. The same way we're committed to use GNOME as our desktop, we're committed to use Ubuntu as our package base, and the reason for this is simple: these components give us the best results to get the job done. That doesn't mean we're not looking elsewhere though. We're often trying out different desktops, in particular with community editions such as KDE, Xfce, Fluxbox. And we're interested also in porting our technology to other package bases such as Debian (for which there's a project planned) and Fedora. Because of the complexity of these upstream projects and because we're focused on our main task, we're not actively involved in working with them or in developing our own desktop or package base.

CL: We only have a very small number of customers and they're mostly home or small business clients. We're expensive when compared to Canonical, Mandriva or other support offerings in the Linux market and that's because the support is done by the development team itself. We're also careful when it comes to support as we want to remain focused on the distribution itself rather than on commercial activities revolving around it. Our business model is extremely light and very efficient. We're funded by our user base and the on-line activity it generates and that allows us to be successful without worrying about whether or not what we're doing is lucrative.

CL: It's robust, predictable, modern, comfortable, efficient - all the reasons why you'd want to use it at home also make for an ideal workstation. It's quite popular among small to medium companies. Our project is small though, and it lacks independence, long term strategies, marketing, PR and support structures. For these reasons, it doesn't appeal to large businesses, where Red Hat, Novell and, to a lesser extent, Mandriva and Ubuntu are more viable solutions.

CL: Yes, the live CD is compressed with Squashfs. There's about 2.5 GB compressed within these 700 MB :)

CL: It's a bit too soon for me to talk about this but we're planning on two significant developments - a community website with a hardware database, ideas (similar to Brainstorm), blogging, social-networking, support, software portal and many other features. And a complete re-write of our Software Manager. This time we want it to be package-centric (so hopefully it will replace Synaptic) with over 30,000 packages, user reviews, combining the best from the current mintInstall, the GNOME application installer and the Ubuntu Software Center.

CL: We're having a lot of fun making Linux Mint. Whether it's integrating upstream projects, implementing our own ideas, interacting with the community, it's always fun. And it's a pleasure for us to see people getting excited about what we do and users happy with our releases. And then there's also so much more than Linux Mint, so many distributions to try and to download, so many other software applications to install, there's a world of fun for everyone to enjoy. I think that's the beauty of open source, that energy and how easy it is for developers to build on top of what's already there and how exciting the whole thing can be. I hope this will last. There's also important questions to be addressed and conflicts to be resolved when it comes to free software and open source and we shouldn't avoid them, but to all people who bring joy and excitement to us and who keep Linux going, I'd like to say thank you. That's the most important aspect of all and that's what we're all here for.