The 2006 Linux Client Survey closed today. I included one question for
the desktop architects in the survey and here it is:
Where should the Linux desktop community focus their efforts in 2007?
Top Priority Important Good to have Lower priority
Audio/Multimedia 36% 35% 23% 6%
Printing 37% 40% 18% 5%
Wireless 46% 35% 14% 5%
Power management 21% 33% 34% 13%
Open source drivers 56% 27% 13% 4%
Linux desktop 51% 29% 16% 4%
standards
(cross-distro)
Fonts and document 31% 37% 25% 7%
fidelity
Application 35% 35% 22% 8%
packaging
Cross-desktop 36% 35% 23% 6%
environment tools,
services, and
interfaces
Developer 15% 34% 40% 11%
documentation
Development 22% 35% 32% 11%
tools
Accelerated 37% 32% 24% 7%
graphics
I don't like to draw conclusions based on shotgun exposure to a survey
like this, but it is very interesting that (1) open drivers, (2) Linux
desktop standards (cross-distro) and (3) wireless are topping the
survey.
Congratulations to the printing and power management developers. These
are NOT the top on the list this time around.
In a recent discussion on dtl_tech_board, Bryce had the following
observations and comments regarding open drivers. When the topic of
open drivers comes up, it usually degrades to "there is really nothing
that can be done" or "it all depends on the market".
-----------
Working from the assumption that the 4 options above what we have to
work with, then it would seem logical to select one to invest focus
into. Some thoughts:
1. Highten visibility
Honestly, I don't think there's much more to be gained from
increasing visiblity here. Everyone who cares probably knows about
the problem at this point.
2. Reverse-engineer
The main downside to this is that it's a lot of tedious work.
However, it's been a time-proven solution, and is entirely within
the control of the community. So the best way to apply resources
here would be to either a) increase the amount of workers, or
b) make the work easier to do.
* Assemble training materials on how to reverse engineer. Perhaps
organize talks from guru reverse engineers at Open Source events.
* Provide tools, hardware, places to collect info, etc. to people
doing reverse engineering work.
* Provide (automated) testing resources. I.e., create a "desktop
farm" here at OSDL loaded up with "problem" hardware, that people
can load different (closed/open) drivers on for experimentation.
* Host a 'hack-fest' one weekend to get reverse engineers together
to focus on specific drivers that we need open drivers for.
Give the hardware vendors the opportunity to attend as well.
3. Create wrappers around windows drivers
Pretty much same ideas as for #2, except that for video drivers, I
would worry this starts to sound like a shim layer, which I believe
has already been ruled out by the community, so I think it'd be much
less likely to be accepted.
4. Encourage vendors to cooperate with community
This seems like it's a long shot. For the drivers that really
matter, I expect the vendors have probably already evaluated the
situation and made a firm decision. However, stranger things have
happened this week, so who knows. Some ideas:
* Pay or trade for access to the specifications.
* Invite employees from the vendor companies to attend OLS and other
open source events, to increase opportunities for vendor to
cooperate, or at least so they gain some visibility into the
community.
* Send key community members to the conferences the vendors are at.
* Get reps from each of the competing companies, and get them to
all agree on releasing the same info at the same time, thus
minimizing the competitive advantage issue.
* Research into who the open-vs-closed decision maker at the vendor
is, then have one or more of the OSDL reps from a
$BigMemberCompany call this person and chat with them about how
nice it'd be if they could work with the community.
_______________________________________________
Desktop_architects mailing list
Desktop_architects@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/desktop_architects
