[conspire] supported graphics cards

[Rick Moen]

Quoting Tony Godshall (togo at of.net):

> It's not a function of time, it's a function of skill and effort.
> It's only a function of time in terms of how long it takes for someone
> who cares and has the skill to do the work.  Most of us just wait for
> some other good-hearted soul to do the work that we happily use for
> free, and some of us complain about how long it took.
> 
> This is one of the reasons it's good to vote with our wallets and buy
> hardware with open standards, or, better yet, vendor-contributed
> open-source drivers.

Towards that end, what would help a great deal would be if more people
bothered to do some elementary research on open-source driver support
for proposed hardware purchases before paying money.  It's usually just
a matter of Web-searching for the chipset mame/model -- not the
marketing name -- and the word 'Linux'.  

The most certain way of checking chipsets is to run something like
Knoppix or Aptosid (a live Linux distro) and running the 'lspci'
command.  Failing that, pure Internet research works once you get the
hang of it.  E.g., it took me only a couple of minutes of Web-searching
to determine that Ken's 'HP P7-1414' was (more specifically) an HP
Pavillion P7-1414, and that that computer uses a motherboard-integrated
AMD Radeon HD7560D GPU to produce video.

If more people would bother to do one of those two things before
plunking down money with the intent to run Linux on a desktop box, there
would be a lot fewer problem components.

How do I know this?  Because I saw that process at work in the server
market where there were always more-cautious and clueful customers.
Around 2002, Adaptec Corporation reached out to VA Linux Systems for
help, in some degree of management panic, because they had observed a
serious loss of market share for their bread-and-butter SCSI adapters.
Why?  Because word had gotten out that driver support for Buslogic SCSI
adapters was great, while Adaptec drivers were mediocre and buggy
(because Adaptec had been non-cooperative).  

Buslogic, a much smaller company, was eating Adaptec's lunch.  So,
Adaptec tapped VA Linux's help in determining how to better cooperate
with the LInux kernel community and give them what they needed.


Concur with your point that it's a function of time and effort spent by 
someone who cares and has the skill to do the work.   Good drivers exist
for Linux for one of two reasons:  (1) Someone is getting paid to
produce them.  (This goes off the rails when some yoyo in management
says 'But you can release only binaries because otherwise outside
parties will discover secret information about our hardware.'[1])  (2)
Someone writes/improves the driver in order to scratch his/her own itch.

The second point is important.  If you buy hardware likely to be also
bought and used by Linux driver developers, you are likely to find that 
driver support for them is excellent.

I used this strategy in 1999 when I suddenly needed a highly reliable
laptop for work.  I'd noticed and heard that a number of Linux
developers were then using first-generation Sony Vaio laptops, so I
bought a used PCG-505TX -- and it was superb for Linux.

What hardware, generally, is likely to appeal to Linux driver
developers?  As a general rule of thumb, hardware that's high
performance but not luxury-priced, that's been on the market for a while
and not plagued by any peculiar misfeatures or botched implementation.

Conversely, probably few in the Linux technical community are working
very hard to improve support for HP OfficeJets, because instead they
just avoid owning any.