Exclusive Interview with Andreas from Yoper.com
In the past few weeks, there has been a lot of hype and controversy surrounding Yoper, ranging from insults to ruthless Gentoo comparisons. I recently sat down with Andreas Girardet, who is a key developer for Yoper, to dispell all the rumors and discuss the direction in which the Yoper project in which Yoper is headed:
[Andy Kissner] Which OS do you use personally?
[Andreas Girardet] Yoper
[AK] Recently there has been a lot of comparison between Yoper and Gentoo. What do you think that says about the YOS Project as a whole, and the general direction it's taking?
[AG] Your Operating System (Yoper) is really different from Gentoo on both a technical level and a marketing level. I really love all operating systems apart from one [Windows] and this is the reason why we take the best out of each OS and mix them into one [Yoper]. This causes some of the similarities between Yoper and many other distributions, including Gentoo. This also causes similarities between Yoper and say MacOSX or even Windows. For example Yoper contains emerge/portage in its basic form, which is handy for single packages that we do not currently provide. It also contains Sax2 and in an upcoming release we will include the first Yast2 tools from SuSE. I also find the flamewars between different distros very disturbing, since they are counter-productive to my personal mission in life, which is to unseat the Microsoft monopoly.
[AK] It seems that it has become pretty standard to have a package selection step during the installation. Was there anything in particular that made you guys go with a generic install?
[AG] One of the main ideas of Yoper is that its ISOs are purpose driven. Even though we currently only have one ISO, my original idea was, and really still is, to have several seperate ISOs for various pruposes. Each of them is based on the same code base and can easily be turned into one another by a power-user, but in general the purpose should be fulfilled 100%. This allows each ISO to be 100% optimised for that particular purpose. Personally I started on a KDE Desktop, and will build off of that. In a few weeks I'll release YMinimal, which is a minimal install, and YGnome, the GNOME Desktop Package. Also in the works is YLive, which is like a combination of Yoper and Knoppix, and we have some very innovative ideas on how to make a Live CD really fast. Another developer is working on a Server edition, and even more are in the development pipeline.
[AK] I know that there were several techniques employed to make Yoper faster, including stripping* and prelinking**. Would you like to briefly explain them, and are there any other things you did to help speed things up?
[AG] It is not rocket science and if one has the know-how, one could tweak their Gentoo, LFS, or even Debian system to be like Yoper. You would probably spend weeks/months doing it, but after this long, possibly frustrating road, you would get something like Yoper. But instead of a week-long struggle, you can have Yoper ready within 10 to 15 minutes,which to many people is more important than a steep, frustrating learning curve. Some of the "secrets" of turning your distro into Your Operating System are:
0.) Performance patches from Con Kolivas, i686 2.6.7 kernel, reiserfs
1.) All original sources, minimal patches.
2.) Compiled with i686 against latest gcc
3.) Stripping
4.) Prelinking
5.) Latest gcc and glibc and other sources
6.) Keep everything only dependent to what it really needs not what
the ./configure happens to find.
7.) Hdparm on install
[AK] You seem to place a lot of emphasis on keeping the open-source mindset and keeping the community involved. Are there any steps you are taking to ensure that Yoper doesn't go the way of many other commercial Operating Systems?
[AG] Sure there are. I believe that there need to be rules for something like this not to happen, and to ensure that the focus continues to be acheiving the original goal of Yoper, which is to destroy the monopoly in the OS Market. This is a commercial war; a war against the Redmond Giant [Andreas is referring to Microsoft].
[AK] Could you clarify a little?
[AG] Yoper can keep the open-source mindset by following a non-elitist approach when choosing developers (Yoper Team). Everyone has the right to be a Yoper Team Member. Yoper is a user-driven distro and not a developer-driven distro. Being part of Yoper team is easy and ensures that a wide variety of people are part of the decision making process. Of course, it is my call for as long as I lead the project, but as a reasonable individual I work on consensus. It is every Yoper-user's RIGHT to be a Team Member if they:
0.)Use Yoper
1.)Are not scared of the command-line (or at least don't hate it!)
2.)Are very enthusiastic about Yoper.
The top priority of the Yoper Team is to help other users. I think this is novel, because developers are forced to think "user" every day. In the end this is an OS made by users for users. Ultimately having such a good and open development base also assist us in our commercial path to unseat the monopoly. I feel that writing
quality documentation for the enduser, like books, will help ensure that a commercial path for Yoper is possible and brings some real value to the users. In general I will be looking for ways to create such 'real value' and charge money for it. The bleeding-edge software and packages should remain free, because charging for this would go against this community-driven model, since there this software was developed for free. The demand by the increasing amount of users will drive this and users will in the end demand commercial services.
[AK] So you're trying to keep the open-source mindset, but still have some funds flowing to keep things afloat?
[AG] My hope is for even more than that. The idea is to grow dynamically by achieving critical mass. If you look at even 10% of the computer users in the world, that great an amount of users will naturally have some demands that can only be met commercially, and there are plenty who would be happy to pay for a Linux OS that fits their need 100%. The basic idea is a Free OS for Free people. If you have an outstanding product that is free, you can reach critical mass faster and more efficiently. Once critical mass is reached people will demand a higher quality or an additional service that we will then happily provide.
Consider also this: Microsoft could easily give Office and Windows away for free, yet still make enough profits to put 99% of publicly traded companies in their shadow. Anyone in the Linux market who charges for the base Operating System and builds their entire business strategy around a charging based model for the OS would be completely wiped out if Microsoft would ever give Windows and Office away for free, even if it were only for a year or two. This is why Yoper is free. Why should I build my business on such shaky grounds especially if my competitor could just kill my business instantly by giving away their product for free? I rather build up critical mass to the point in which this critical mass will automatically demand a service for which we then can charge.
[AK] And what did you do before Yoper?
[AG] I hold a Masters in Philosophy of Logic,and have worked for ISPs as a system administrator/manager since the dawn of the commercial internet. Linux was my daily bread as an admin since early on, and I am still contracted by companies for work with Linux. Right now I am with IBM, and in my spare time I work on Yoper. Though I have started taking extended holidays more frequently to dedicate myself to work with Yoper.
Andreas Girardet is the lead developer of the Yoper project,and can be found hanging around the forums at
www.yoper.com *Stripping is the process of removing extraneous comments and
debugging code from the final build, which makes execution faster.
**Pre-Linking is the loading of required libraries before they are needed rather than at runtime. A good description and pre-linking
tutorial can be found at http://www.crast.us/james/articles/prelink.php
--Andy Kissner
lakerdonald@gmail.com