Last month Progeny Linux Systems ceased development on their own distribution in order to focus on selling professional services. In their announcement, the company cited the prohibitive cost of developing and publishing a distro. This move marked another firm in the wave of tech companies, Linux and otherwise, making significant changes to adjust to the market slump. Progeny’s distribution was based on Debian GNU/Linux, and many in the Linux community were closely watching the company because it was founded by Debian creator Ian Murdock. OSNews spoke to the President of Progeny Linux Systems, Steve Schafer, once the dust had settled on his company’s announcement. 1. In August, you stopped development on Linux NOW and you’ve just stopped publishing Progeny Linux. How was morale at the company once these decisions were made?
Steve Schafer: Both decisions made good business sense, but were still a bit difficult to make. Several people inside and out of the company felt very passionate about the NOW project, so it wasn’t taken as well as the Progeny Debian decision. Our commitment to the Debian Project was understood in the latter case, which helped with the reactions.
2. You mentioned the reaction from outside the company, how has it been from your clients and the community?
Steve Schafer: The response has been disappointment for the NOW announcement, understanding and support for the Progeny Debian announcement. There’s no getting around the NOW announcement since the project is, for all intents and purposes, defunct. However, our clients appreciate the fact that we are continuing to support Debian, so they all took the Progeny Debian decision a bit better.
3. Progeny was based on Debian GNU/Linux, which has a reputation for being slow to integrate new software developments. Was Progeny facing the same business challenges that other distributions were facing?
Steve Schafer: Debian integrates new technology just as quickly as other distributions, but the actual version releases are slower. You can buy the latest Red Hat version off the shelf, or run apt-get on the last Debian release to get relatively the same versions of the underlying software. (Note: Software that is not in the last Debian release is posted to an “unstable” and then a “testing” branch of Debian where it is available, but needs to be explicitly included in your apt sources. ) In some cases you would even be getting more up-to-date software than that on the shelf.
As for challenges, the Linux distribution “business” is very challenging today. In order to be a player, that is have actual business success with a distribution, you have to have a retail presence. Two years ago when I helped put the Mandrake/Macmillan deal together (bringing Mandrake to mainstream retail), it was much easier to get in, and stay
in retail. We’ve seen the market go from two prominent distributions on the shelf (Red Hat, Mandrake) to almost ten, and now back down to two or three-including the original two, namely Red Hat and Mandrake. Red Hat, Mandrake, and SUSE have all issued statements on how retail is not making money, and in some cases losing money. Since retail sales have slowed down the retailers aren’t as receptive to new products.
When you add in Debian’s slower development and release schedule, things get really tricky. The only way to release more quickly is to release a separate product, in essence forking from the main Debian distribution. We’ve seen this with every “derivative” distribution -Mandrake, Stormix, Corel. Looking down the road we saw Progeny Debian varying further and further from the Debian core, a path that didn’t benefit customers, Debian, or Progeny.
4. What can you tell us about Progeny’s plans to continue promoting Debian and make it a more viable platform for commercial users.
Steve Schafer: Many of our Debian improvements have been submitted to the Debian Project and will appear in upcoming releases. Others are being revised to be Debian generic and will be submitted ASAP. We also continue to offer support for Debian (as well as other varieties of Linux). Since the main criticisms of Debian are ease-of-use and lack of a commercial entity behind it, we hope we are helping to answer both concerns.
5. Who will maintain the packages that Progeny is folding into the main Debian archive?
Steve Schafer: Our Progeny developers, who are also Debian developers and package maintainers, will take care of the Progeny packages.
6. Will Progeny’s experience be a cautionary tale to others?
Steve Schafer: I think everyone needs to be a bit more wary in the Linux space. It’s still a very good alternative (if not a primary choice) for an operating system, but it’s not the cash cow everyone hoped it would be. Before basing a business on Linux products, one should take a hard look at the economy and the recent experiences of other companies. Also, “different” isn’t always better. Although the Linux Standard Base is doing great work, there are still more Linux distributions in the world than necessary. Why reinvent the wheel or put out something marginally different when the existing distributions will do? Why not spend the energy making the existing distributions more useful, through utilities, application software, services, and support?
7. Can you tell us about your vision for Progeny / Debian in a year from now?
Steve Schafer: Our hope is that Progeny can make a name for itself as being the commercial entity behind the Debian Project, helping Debian reach better market penetration in the enterprise. To that end I also hope we can have a positive impact on the project, spurring development to a faster pace (a release a year would be nice) as well as aiding the technical advancement of the overall distribution.
I think it’s really good that Progeny focuses on it’s business concept.
Progeny is needed to help comercials adopt Debian and provide those clients support.
Why Progeny wanted to maintain a parallell Debian branch I never really understood. If the problem was that they found the current installation program too tricky they could just have made new bootdisks for existing Debian branches.
Debian already has got the best developer network, community and the most stable and secure releases.
What’s really needed is a company that helps other companies with adoption.
Keep up the good work!
thats what it sounds like, they will be a consulting firm that pushes debian and Linux solutions….that could work.
He is really “Steve Schafer”, not “Michael Schafer”, in case anyone noticed the discrepancy over our Fearless Leader’s name.
Fixed. Lousy day today, huh? ;D
I’m probally going to try and organise some other systems administrators in creating a timed-update-pack service for the stable distribution. Many of us dont want the underlying tools to change every 3-6 months, but would welcome user-tool upgrading on a regular basis to keep our users happy.
I used RedHat for several years and always thought that was what I should use. Then one day a friend of mine encouraged me to try Debian. I kept putting him off but then one day my RH box got hosed so I said, “Why not?”
It was a real bastard to install for the first time, but then again, so is any Linux distro when you are not installing straight off a CD. In fact, I had to install it twice. I was feeling fairly annoyed until I used ‘dselect’ for the first time.
I could scarcely believe that I could just select a bunch of packages and that it would calculate and install the dependencies *automatically* – whereas with RedHat, I had to scour FTP sites and my installation CD-ROM and wonder how recent what I was getting was. (Sure, there were X11 tools, but they were merely toys that put pretty pixmaps in front of a basically dumb [as in unsophisticated, not as in stupid] package management scheme, RPM.) It wasn’t enough for Debian to make a package manager – they actually set up a whole system to stand behind it, so that you didn’t have to root around for pre-depended packages.
It was beautiful. No longer was I downloading RPMs, only to find that some stupid library was off by a single minor version number. dselect handled everything. Additionally, the /etc tree was less spaghetti-like – you didn’t have a billion configuration files under /etc, as they were all moved into /etc/pkgname directories; and it actually adhered to the sysv init standard, i.e. /etc/init.d, /etc/rc0..6d, instead of that wierd /etc/rc.d thing in RedHat.
Being able to just type ‘apt-get install task-kde’ and watch it download and install a horde of packages, and then have KDE up and running without any further ado, is a really powerful thing.
I also like the way the default packages are set up – joe, vi, etc. They are configured with better defaults than the RedHat equivalents I have to use on other people’s boxes.
I think that anyone who is considering the use of Linux for workstations and/or servers should seriously look into Debian. Even if you go with something else, you’re seriously cheating yourself out of the chance to evaluate one of the best (IMNSHO, _THE_ best) Linux distributions in existence today. I hear my RedHat-using colleagues saying “Debian, eww” when I talk about the distro I use. I liken this to someone who drives a Kia turning their nose up at a BMW. Stay with what you like, I guess, but don’t knock it until you’ve given it a try!
For the record, I’ve had an experience almost identical to Stu’s. I think the main problem with having so many distributions out there is that people tend to insult other linux distributions when they have no idea what they’re like. Rather than critizising things you don’t know about, why not tell others who may not be so familiar with your favorite the reasons you like it?
I’ve seen and heard tons of people attempt to insult other distributions making wildly inaccurate claims. This is true of everyone, not just redhat users, not just linux users. =)
One thing that keeps people away from Debian is the lack of isos.
I finally tried debian via Progeny, because they provided the isos! you see, my floppy drive is broken. I have no desire to fix it or obtain a new one, beause I never use a floppy drive… and the only way to install debian, as far as I could tell, was to get a few 1.44 mb inmages and put them on a floppy, and boot with that to get to real installation. I had the same problem with FreeBSD, by the way.
Anyhow, this Mandrake user was thrilled with Debian. It reminded me more of my long lost Amiga: Red Hat really did make some sloppy decisions that Mandrake inherited, such as the whole init rc.d deal. And witness the recent article by Mosfet decrying the /usr directory’s glut of programs. And then, debs really are better than rpms… rpms don’t always uninstall, and things like that. ick. To top it all off, being able to go to a command line and type apt-get install <insert practically any program here> and have it grab the file off the network is amazing. I wonder how much bandwidth this takes ??
Thts one thing that could eveolve into a great systme for Linux, whatever distribution: there is a sort of central program source (rpmfind, apt-get, whatever) where a user can get everything from the gimp, to xchat, to licq, to apache, simply and easily. Windows has nothing like it.