[conspire] Sidux, Aptosid

[Rick Moen]

Some interesting distro politics, recently.  Let me try to put it in
perspective.

Inspired by the Linuxcare Bootable Business Card, Klaus Knopper created
Knoppix as an occasionally released live CD with advanced automated 
hardware detection, based about 80% on Debian's unstable branch (aka
'Sid').  One of the members of the Knoppix community (around 2003) was
Jorg 'Kano' Schirothke, who maintained custom scripts for Knoppix but
found it limiting, so he started his own KDE-defaulting live CD for i386,
Kanotix, which was more purely based on Debian Sid (unstable), and
unlikely Knoppix officially supported HD installation rather than as an
unsupported afterthought.  

The idea of an i386 live CD closely based on Debian Sid being a good idea
whose time had come, during 2003-2006, 'Kano' gathered an active
community of fellow maintainers, and Kanotix was excellent, a good way
to run live-CD Linux, and a great way to install Debian-Six.

In 2006, 'Kano' had a falling out with... nearly everyone else in
Kanotix, as he thought continually applying stability patches to the
rolling Sid packages wasn't worth the trouble, and wanted to switch to
Debian-stable.  

Pretty nearly everyone else left, forming Sidux as a continuation of the
ideas from Kanotix.  The Sidux people finally got absolutely right, IMO, 
the task of producing a quality desktop distro:  Theirs had a
'rolling-release cycle', where they released a stabilised set of
cutting-edge software quarterly that was able to be maintained
Debian-style rather than getting periodically thrown away and
overwritten (the Knoppix standard).  They got really good at producing
stabilising update packages with relatively minimal effort.

However, earlier this year, some of Sidux's quarterly releases were
late, and we kept hearing that there were major policy disagreements
between Sidux's developers and the sponsoring non-profit foundation in
Germany, 'Sidux e.V.', with the result being this announcement on Aug.
15th, 2010:

  The members have decided to vote for a separation of the project Sidux
  and the Foundation, when it comes to legal bindings [affiliations?].
  This allows the developer team to move on with the distribution, and
  allows the Foundation to also be engaged in other FOSS projects.

  The decision of the members includes the handling of some legal issues
  that need to be dealt with by the Board.  As it cannot happen overnight,
  please give us some time to undertake what is required.

(http://sidux-ev.de/index.php?name=News&file=article&sid=69)


The other shoe just dropped:  All of the parts of Sidux that matter,
which is to say everything other than Sidiux e.V. and the name 'Sidux', 
are going to continue -- without skipping a beat -- under a new name at
a new Internet domain.  That's Aptosid, http://aptosid.com/ .

  As I am sure you are all aware, there have been interesting times for
  sidux recently.   The bad news is that the sidux project is dead.   The
  good news is that aptosid has been aptly born like a phoenix from the
  ashes and will provide a smooth upgrade for sidux systems.   In many
  ways nothing has changed but our name.

  Having declared the last attempt at an election for the e.V. board void
  while it was happening, the election of a new board is now overdue by
  more than a year.   The e.V. has also had no official public financial
  reports and until recently the treasurer was refusing to correspond with
  the other board members.   There is now also even an internal e.V.
  challenge to the decision of their last meeting..

  Each person you ask will provide a different story of what happened and
  they are not all lying.   Most of the problems which occurred seem to
  come back to communication issues, from language problems to personality
  clashes.   Two entities evolved which could not communicate.   Over time
  this communication breakdown caused an ever expanding range of problems.

  The reality now is that the sidux e.V. seems to believe it owns at least
  the European registration of the TM for sidux and we have been unable to
  reach a position where the developers of sidux are able to feel in any
  way comfortable about carrying on under that name.

  So aptosid is born.   Those of you who dist-upgrade will be asked by
  debconf to change your sources.list files.   Despite what some may have
  said, we do care about everyone who uses our software, so your system
  will still be supported there.   You may even find other changes you
  like.

  What has not changed is the primary team behind the development of sidux
  and we hope that everyone who enjoyed our work and used that system will
  come join us at aptosid.

  sidux is dead, long live aptosid.

In a sense, this is a reminder that, with open source, over the long
term, sponsoring companies and their copyright ownerships don't matter 
because you can fork.  Trademarks don't matter because you can rename
and restyle.

Aptosid's kind of a dumb name, but then, so was Sidux.  Some dumb names
wear well, others don't.  We'll see.


I'm rather struck with the contrast between this sort of parting and
resumption, which is common enough in longtime open source projects
(even though the strife is best avoided where possible) and the disarray
that has overtaken the several _theoretically_ open source projects
Oracle acquired when they bought Sun Microsystems:

o  Java.   Hoo-boy.  Even before the Oracle purchase, GPLed Sun Java 
   was so heavily tied down by Java Community Process obligations, 
   patents, and other control-freak measures (if memory serves, you 
   were prevented from using GPLed Sun Java in embedded space) that
   Google was eventually driven to engineer around them by creating 
   Dalvik for Android (after negotiations with Sun failed).  

   Because Oracle's decided it can't permit people to escape from 
   central control over Java even by writing a workalike from scratch,
   it filed the current lawsuit against Google, Inc. (my opinion, yours
   for a small fee).

   People may not have noticed, but the suit has also made open source
   Sun Java basically a dead letter for at least several years, at 
   which time I figure anyone not wanting to be a Larry Ellson vassal
   (not to mention use open source) will have migrated to other things
   entirely.

o  MySQL.  There have been a couple of non-Oracle sort-of forks, but
   man, what a tepid community response.  Monty Program AB got all the
   developers, and has MariaDB as a 'community' fork using the XtraDB
   DB engine instead of InnoDB.  And there are a couple'a others:
   http://en.wikipedia.org/wiki/MySQL#Forks_of_MySQL

o  OpenOffice.org.  This project's sort of sitting around stunned, 
   figuratively, wondering when Oracle's going to lop off its legs.
   In fairness, maybe Oracle will leave it alone.  We shall see.

o  OpenSolaris.  This has been the real travesty.  A _few_ OpenSolaris
   people responded constructively to the leaking of an Oracle memo
   disclosing that they were killing OpenSolaris -- by setting up the
   Illumos Project, which promised to track Oracle's future code drops
   _and_ rewrite as open source the remaining 1% of OpenSolaris that's 
   still proprietary.  I've written about this on the SVLUG list:
   http://lists.svlug.org/archives/svlug/2010-August/033517.html
   But the reaction I've heard from pretty much all ex-Sun people
   matches Alan DuBoff's doom-and-gloom:
   http://lists.svlug.org/archives/svlug/2010-August/033518.html
   I found this surprising, since the Sidux (excuse me, Aptsosid)
   people showed the standard _Linux_ community response of 'OK, 
   we don't need you'.  As I said when I answered Alan:
     Truth to tell, I think the bigger obstacle is psychological:  
     The Solaris community spent such a long time being near-totally 
     dependent on Sun that setting up an independent basis for 
     themselves just probably doesn't come naturally.
  http://lists.svlug.org/archives/svlug/2010-August/033530.html


Anyhow, long live Aptosid, or whatever it ends up being called.


Linux Mint also has a new Linux Mint Debian Edition that, frankly,
sounds a whole lot like Aptosid / Sidux -- except for the bizarre
fundamental basis from Linux Mint, which always struck me as being
'Let's pretend as if patent and proprietary-software restrictions 
don't exist and just throw in everything.'

http://distrowatch.com/?newsid=06263