By Brian Proffitt
Managing Editor
Apparently, lot of people are thrilled with the whole idea of
Canonical
and Linspire’s technology partnership announced yesterday,
citing source code consolidation, non-redundant development
efforts, and just good karma about the whole deal.
I, for one, am not one of those people.
Initially, my big problem was worry for the Debian Project.
Under the terms of the new partnership (which, for the purposes of
this article I will deem “Linbutu” because I am getting lazy in my
middle age), the Linspire and Freespire distributions will be based
on Ubuntu, not Debian GNU/Linux. Ubuntu, in turn, gets supported
access to Linspire’s Click n’ Run (CNR) install/update toolset.
It sounds like great synergy, doesn’t it? And it is. From a
business standpoint, I think this is one of those matches made in
heaven and I would not be surprised if we actually do see a
“Linbutu” distribution for real some day. For now, just the
partnership seems like a win-win for everyone.
But out of the gate, my first thought was this seemed like a
huge vote of no-confidence for the Debian Project. Again.
For instance, the whole Dunc-Tank plan, where Debian developers
tried to raise funding for some of the release managers could get
paid and thus could work more on their projects and get them out in
a more timely manner, seemed like a good idea at the time. But
several developers did not share that opinion, and when Dunc-Tank
went forward after a Project vote approved the idea, there were
boycotts that ended up delaying the next release of Debian (Etch)
anyway. Etch was supposed to be released in December 2006, and now
there’s talk that it won’t go live until March 2007.
Based on that, and Debian’s historical tendency to run late, I
could not help but cynically wonder if the move by Linspire to an
Ubuntu code base was an effort to just get to a code base that had
a more predictable schedule.
Oh, but my cynicism had yet to start, and I got concerned about
a lot more than the health of the Debian Project.
By separating itself from Debian proper, Linspire not only gets
the advantage of less chaotic source of code, but it gets the added
advantage of the community that’s built up around Ubuntu. I don’t
mean that suddenly Ubuntu coders will drop what they are doing and
go code for Linspire; but what may happen is that suddenly a lot of
the anti-proprietary vitriol that’s been lobbed at Linspire will
die down a bit. Because it would be a bit hard for the Ubuntu folk
to toss such accusations around now.
But make no bones about it; the real PR winner in this
partnership is going to be Canonical. By throwing in with Linspire,
they get more exposure to the slowly-growing desktop Linux market,
they get the CNR technology, and they get a blessing from the Linux
community for “going proprietary.” In recent months, whenever
someone from Canonical mentioned that there would be proprietary
software available in Ubuntu 7.04 and beyond, there was quite a bit
of squawking about it… but never enough to really worry
Canonical. The company’s street cred amongst its community is
huge–more than enough to handle the occasional disgruntled
freedom-loving developer who, at worst, would just go back to the
Debian Project.
Now? Now the situation is different. In one swoop, Canonical has
thrown in with the “props” crowd that is so prevalent in the open
source community these days. No more hints anymore, they are now
committed to “proprietary is okay.” By doing so arm in arm with
Linspire, they just gained an ally that has–thus far–effectively
countered the “all Free/all the time” arguments.
Like I said, it’s a win-win–for them.
The “props” movement within open source has most recently been
codified in Eric S. Raymond in a DesktopLinux
article posted on Linux Today earlier today.
“I know [there’s a] camp that thinks allowing proprietary codecs
into Linux distros will corrupt our vital bodily fluids or
something. My view is we need to get majority market share so we
can crush the proprietary codecs out of existence. If that requires
a temporary compromise, I’m for it,” Raymond said. It should be
noted that Raymond himself has been a member of the Freespire
Leadership Board since last September.
But in all of the celebration about how brilliant Linbutu is, am
I the only one who’s wondering if this may not be such a good
idea?
Linux, as we all know, is superior technology than most anything
to come out of Redmond in recent years. It is free as in beer and
Free as in Freedom, and reflects all that is good about the
creativity of humanity. And yet I worry that if it goes down this
road of inclusiveness with proprietary software too much more, it
may find itself relegated to a field of OS also-rans that tried and
failed against the Microsoft juggernaut.
Before you jump on me, I am not channeling Richard Stallman
today. While his arguments that Free Software is a moral imperative
are noble and attractive, I do not share them. At the end of the
day, software is a tool for me, and I will use the tools that do
the work I need. My preference is to use Free and Open tools. But
it’s not a moral line for me.
No, I am going to make this argument based not on hifalutin
moral stances–this one’s going to be as pragmatic as they
come.
One of the things that drives Microsoft absolutely nuts is the
fact that Linux is Free. Because it’s Free, they have discovered,
you can’t kill Linux. There’s no one to sue, nothing to buy…
heck, there’s not even a crazy corporate executive to vilify in the
media. Not only that, the very Freedom of Linux and its constituent
applications becomes a selling point on its own. I, for instance,
could not code anything to save my life, but the fact that Linux is
Free appeals to me because I know there’s no way Linux will ever be
unsupported or abandoned. Freedom, from a very real and tangible
business standpoint, is a good thing.
But if the Free nature of Linux gets more diluted, I am
concerned that suddenly one of our best defensive and offensive
weapons will get marginalized. Proprietary software can be bought.
Or sued. And, because of its license, proprietary software is
usually not lower-case free, either. Which means the price-point of
commercial Linux distributions that include it will go up or not be
able to ever really go to zero. (Another Linux advantage that could
be lost.)
Linspire, and now Canonical, are arguing that the ends justifies
the means. Get enough users on board and the hardware and software
vendors will be writing Free and Open code so fast Linux developers
won’t be able to keep up with it. I hope they’re right, but it
honestly seems unlikely. I don’t see any incentive to write Open
software if Linux desktop share goes up, not if the proprietary
code was allowed in the first place. Sure, they’ll write
more code if the market share climbs, but what incentive
would they have to make that code Free if some Linux distros have
been letting in the proprietary stuff already? If one distro
maintainer says I can’t come in until I open my code, I’ll just go
to another one that will let me keep my code closed.
I don’t think that all of this signifies the End of Linux. But
it may be a transition point–the end of Linux as we know it now.
We shall see. Good-faith efforts on the part of the commercial
distros to really get those vendors to cough up free code will
help. Projects like the kernel developers offering to write driver
code gratis are a plus as well.
My fear is this: the more like Windows Linux becomes, the better
Microsoft will be able to handle Linux on their terms, not ours.
Let’s keep playing by our rules, not Redmond’s.