Freespire moves back to Debian

 
Freespire Returns to Debian Roots
 
Debian Lenny-Based Freespire 5 to Spearhead Xandros Consolidated Desktop
Strategy
 
NEW YORK, NY - August 6, 2008 - Xandros, Inc. <http://www.xandros.com/> ,
the leading provider of custom OEM Linux solutions, next-generation Linux
desktop and server products, and advanced cross-platform Windows-Linux
management tools, today announced that the Xandros
<http://www.freespire.org/> Freespire 5, arriving Q4 2008, will be based on
the Debian GNU/Linux "Lenny" release. The move spearheads a new Xandros
consolidated desktop strategy following last month's acquisition of
Linspire, developer of the CNR software distribution facility, and the
Linspire and Freespire Linux desktop operating systems currently based on
Ubuntu Linux.  <http://www.freespire.org/> Freespire 5 will be followed by
Xandros <http://www.xandros.com/>  Desktop Professional 5, built on the same
open source code base with additional commercial elements primarily for
enterprise customers.
 
"This is really some of the most exciting news since the announcement of
Freespire 1.0," said Tom House, Freespire community organizer. "This is what
we were hoping to see.both the continued development of Freespire and a
return to a Debian base."
 
"A combined Xandros/Linspire development effort will return Freespire to its
Debian GNU/Linux roots and put it in sync with Xandros Desktop Pro," said
Andreas Typaldos, Xandros CEO. "We will have a leading edge code base while
preserving our commitment to Debian, stability, Windows interoperability,
and ease of use. This commitment allows us to meet the needs of a wide range
of users, from open source enthusiasts to demanding enterprise clients. In
addition, we are intensifying our commitment to the Freespire open source
community, which will now help to drive both the Freespire and Xandros
products."
 
About Freespire
Freespire <http://www.freespire.org/>  is a community-involved, Linux-based
operating system that combines the best that free, open source software has
to offer with the freedom to choose proprietary codecs, drivers and
applications as users see fit. For more information, please visit
<http://www.freespire.org/> www.freespire.org.
 
About Xandros
Xandros, Inc. <http://www.xandros.com/>  is a leading provider of
mixed-environment BridgeWays management tools and intuitive Linux solutions
including SMB and enterprise servers, consumer and business desktops, custom
OEM solutions, and content delivery from digital warehouses. Xandros
<http://www.xandros.com/>  subsidiary Scalix provides the premier e-mail,
calendaring and messaging solutions based on open standards and open source.
Xandros <http://www.xandros.com/>  fosters Linux adoption with graphical
products that leverage existing skill sets while providing seamless
Windows-Linux interoperability. Xandros <http://www.xandros.com/>  server
and management tools feature workflow automation and centralized remote
deployment and administration. The company is headquartered in New York with
research and development facilities in Ottawa, Europe, and Mumbai, and sales
and support offices worldwide. For more information, please visit
<http://www.xandros.com/> www.xandros.com.
 
# # #
 
XandrosR is a registered trademark of Xandros Inc. Scalix is a trademark of
Scalix Corporation. All other trademarks and/or registered trademarks are
the property of their respective owners.
 

You know, I can't help but wonder about these derivative distros that oscillate between Debian
and Ubuntu.  How much of it is really technical considerations versus making noise for press
releases.

Well there was a push to standardize around Debian and ensure binary and package compatibility
between all the players. 

Unfortunately the inability of Debian to get a Stable release out on time combined with the
fact that Ubuntu, which was the big new thing at the time, didn't give a crap about
compatibility doomed that effort.

But now it seems that Ubuntu is having a difficult time getting a stable release so what is
the point of targeting Ubuntu when it's Debian that does most of the work and Debian seems
like it's on track for a timely release?

(I think it's about time that the Ubuntu vs Debian thing goes by the wayside. They need to get
better about merging their efforts with a huge priority on making sure that 'deb' packages
work, with reasonable certainty, on any 'Debian-based' system) 

> Surely if "it's $X that does most of the work", but "$Y didn't give a crap about
compatibility" then it would make it very hard (impossible?) for $Y to do all that leeching
from $X...


No it's not. 

What you do is this:

Start off with a Operating System that is essentially a snapshot of Debian Unstable. 

Then go and modify and break a bunch of existing functionality in order to add your own.  Say
edit some INIT scripts so that they work well with a newer version of Network-Manager (or some
other new sort of thing) without caring of your changes break a bunch of other scripts that
depended on what your editing. 

Then update a bunch of packages, remove a bunch of packages you don't like, and change things
around so that the versioning numbers are different and files sorted with package X on your
custom distro are conflicting with package Y provided by Debian.. which you don't plan on ever
using so it's not a big deal.

And then on top of that while your doing all of this Debian has started introducing newer
versions of packages that you've modified.

----------------

Look at it this way:

It's a hell of a lot easier and you get your work done much faster if you don't care about
compatibility.

Package compatibility takes a lot of extra effort and testing. It's a hell of a lot easier to
break something and have it work for your specific task then it is to modify it while making
sure that it still works.

----------------


If you don't believe me, find out yourself. 

Take all the Gnome packages built for Ubuntu and then install them on Debian Lenny. Get back
to me on how well that works out for you, although I really doubt anybody has to try that out
to guess how well that would work out.

I believe that Ubuntu re-bases itself off of Debian for each release. So this isn't a serious
problem were distributions have diverged off into 2 entirely different directions. I
understand that a lot of work that Ubuntu does goes back into Debian. It just takes extra
effort.


---------------------------

> "[I]t's about time that the Ubuntu vs Debian thing goes by the wayside"; one would hope so.
There are quite a few Debian developers out there, of which 2% happen to be on the payroll of
a certain Ubuntu backer. It must be so annoying to spend 40 hours per week swimming one way
and the other 40 hours per week swimming in the opposite direction.


What ends up happening is that you end up just doing everything twice. Once for Ubuntu and
again for Debian.

What I am trying to point out here is how nice it would be that people would only have to make
a package one time and have it work in both distros.

Oh and I didn't say that Ubuntu was _leaching_ off of Debian.

Far from it. 

Ubuntu using Debian as a base is the best thing they've could of done. Because god knows we
don't need another distro or another person repackaging everything.

If Debian does a good job and Ubuntu can use it they _should_ use it.

That's not different of what Debian themselves do with the software they package.

The packagers pick some version of a program, add and move stuff around so it matches the
distro's policies and put custom fixes on top when problems are found. More often than not,
Debian's fixes are not pushed back to the original code, that has probably moved on. 

That's life. Building a derivative only saves you so much work. 

AFAIK, only Slackware is free from this, because they the software as is.

Well I wasn't trying to make out that Debian is godlike and they do no wrong. I am sure that
they are as guilty at it as anybody.

And the Slackware attitude to dependency tracking with software packages is pretty much
correct. It's a nice feature, but it's not the best way to solve many of the problems faced
with Linux software distribution.


> That's life. Building a derivative only saves you so much work. 

The ideal situation is the upstream developers package the software themselves and then
distributions collect them, test them, and then provide them to their end users.

That should be the ultimate goal for software distribution.. people package their own software
and it 'just works' on all distributions.

All the distributions that get released around the same time all use the same software.
Similar versions of OpenSSL, Xorg, Linux, GCC, etc etc. Having a dozen different people tweak
all these applications in dozen different ways is kinda silly seeing how little it actually
benefits end users. 

I mean, as a end user.. Is the GCC provided by Suse is going to be different enough and
provide compelling improvements compared to the version shipping with CentOS, Fedora, Debian,
Ubuntu, Slackware, etc etc? Is it possible justify the amount of duplicate work that all these
people put into building and packaging their own very-slightly-different versions when there
is so much other important work that these guys could get done by working together?

And if Suse did indeed have some dramatic improvement to their GCC version.. wouldn't it be
nice if Debian users could just copy it off of Suse's FTP server or installation cdroms and
use it? No worries? That the GCC developers could just suck down the patches and use them
without worrying how it may break something in Debian?

I like the idea of high quality binary distribution of open source software. But maybe there
are other ways.. rpm2git seem interesting.

Maybe it can be extended to support debs... I don't know. It could be a useful way to start
smoothing out the trivial differences (that pop up as stumbling blocks for users) between
distributions and lead to higher amounts of software compatibility.

A bit offtopic, but does anyone know how the Xandros takeover will affect the freeness of
Freespire?

I'm not so sure of the current freeness of Freespire, but if anyone knows of Xandros's
statements on this, then that would be insightful.

You don't have to worry. Freespire could hardly add more proprietary components than what is
already present in it.