Community: A Rebuttal to SJVN on Forks and Desktop Linux Direction
Mar 17, 2006, 17:30 (15 Talkback[s])
Re-Imagining Linux Platforms to Meet the Needs of Cloud Service Providers
[ Thanks to Zeek
Greko for this article. ]
Pardon me Steven, but I must have missed your article of equal
"MS/Linux" (Mark Shuttleworth/Ubuntu Linux) effectively forked
from the Debian proper, core distribution by breaking binary
compatibility with them. As you must know, there are now
third-string distros promising binary compatibility with Ubuntu
instead of Debian.
These disparate cores are what is holding desktop Linux back.
The desktop GUIs that you seem so worried about are, to the
technical observer, just the veneer.
In order for Linux to become fully accepted as a viable
replacement to Microsoft as a desktop for the masses, all of the
system configuration tools and installers will need be graphical
and the same--regardless of the end product distribution. That
makes the next logical step a GNU/Linux-wide standardization the
"widget" cores: GTK and Qt.
This would leave us with two core distros "GNU/Qt/Linux" and
"GNU/GTK/Linux" that anyone could easily build upon and configure,
regardless of the desktop veneer. A GNU/Qt/Linux core distribution
would have configuration tools and an installer based on the Qt
widgets, but one could still run all GNU/Linux software simply by
adding the GTK libraries later, and vice-versa.
Community-owned GTK and Qt cores are the right direction to move
Linux in, and right well you know it. You have probably thought of
it yourself already. The commercial Linux interests are actually
slowing desktop Linux's growth. They won't willingly go down this
path because of the "Linux-end-game," which is an end-run around
the GPL for not "actual" but "effective" ownership and control of
GNU/Linux through market momentum and tipping point market
Microsoft does not "actually" own the desktop market, but they
have oh-so "effective" ownership and control of the desktop market
through market momentum and a tipping-point market share they
achieved long ago.
My white box builder/system admin friends and acquaintances
continue to sit on the fence where GNU/Linux is concerned because
of the disparate core distributions that they don't have time to
support. Or are unwilling to make the time to support. They also
don't like the fact that GNU/Linux desktop distros don't have a
universal/general purpose installer for third-party software. And
that the only unassisted "end user easy" way to install third party
GNU/Linux software is by turning what any four-year old can do for
free on Windows with a couple of clicks, into a vendor locked-in,
anti-competitive paid-for subscription service like Click-and
The general package repository model common with all GNU
distributions is also "broken" because it is untenable. Debian's
last release was delayed, and is good proof of this fact.
If all Windows-compatible software had to be packaged,
maintained, and distributed by Microsoft--besides being an
anti-competitive nightmare controlled by those maniacal brown
shirts--those repositories would be so staggeringly huge that they
simply could not be maintained. Distributions need to become a
distributed effort. Let the third-party software vendors package
their own products and deal directly with their clients by
standardizing the installers. In this case, using two core binaries
"dot Qt" (.qt) and "dotGTK" (.gtk) packages.
You are a recognized and respected pundit in the computing
industry, Steven, with a voice that will be heard. Why not be an
elder statesman and rise above the commercial pressures? Represent
everyone in the community by publishing a manifesto of where
GNU/Linux really needs to go to be successful on the desktop!
Repeat After Me: No More Linux Desktop Forks(Mar 17, 2006)