On Wed, Dec 10, 2003 at 08:52:03AM -0800, Linus Torvalds wrote:
> > Must be some new management over there... *sigh*
> Never attribute to malice what can be sufficiently explained by
> laziness or stupidity.
Why do you think I mentioned "management"? :)
> But when it comes to old hardware that doesn't bring the company any
> revenue any more, the only reason to maintain documentation (even if the
> "maintenance" is just having people be aware of it, and making it
> available on a web-site and having the proper indexing to find it) is to
> show that you're committed to your products, so that people who buy the
> new ones can trust you.
It's unfortunate that while they still have the sales information and
white papers for their old chips up on their website, it's next to
impossible to get a human being that will even *entertain* your request
for documentation, much less grant it. One would think that for old
hardware, supporting it by posting documentation and code would be
a bigger priority than maintaining product info for something they don't
even sell anymore?
> And a lot of companies don't seem to think that it's worth the pain.
> They'd rather screw their old customers over and try to get them to
> upgrade to the new product. Which works well enough, I guess, since it
> clearly is fairly rare to try to support your old stuff any longer than
> absolutely necessary.
I really wish it were possible to negotiate term limits on NDAs.
Unfortunately, even providing docs under NDA seems like an immense favor
> Major documentation policies usually only exist at the big old-time
> companies. For example, I could buy the "Technical Reference: Personal
> Computer AT" reference from IBM in 1991 - even though the thing was
> written in 1985. Because a company like IBM tends to have all these
> customers that really pay for _support_ more than new hardware. Their
> paying customer base literally cares about the fact that they know that
> they can get support and docs for hardware that may be quite old.
Too bad, really; in all honesty I'd pay for a support subscription for
my hardware because it's the hardware that fits my needs perfectly, and
changing any of it would simply be less economical than buying all new
stuff, not necessarily in terms of raw expense, but in terms of time
spent re-hacking all scripts, changing configurations, etc. And too
frequently the shiny new hardware focuses on the whiz-bang features and
neglects things that you took for granted with the older stuff.
I think support subscriptions would be too much of a trickle effect to
bother with for old stuff, but we will probably see this in the future.
One of the reasons support was given as "part of the package" in the
past was because you had to convince people to buy into your market *at
all*, much less choose your particular product. (Who needs a 3D game
card, there aren't any worthwhile games around? - Me, 1996)
Now that market segments like the video card market have a clearly
defined inelastic demand component (the fanboys who rush out for a new
video card every 6 months), the companies that produce the cards
have less incentive to throw in things like support at no additional
> In contrast, think about the best customer base of a graphics card vendor
> for a moment. The best customers there are the ones that upgrade once a
> year, and if they throw away their old card in disgust because it no
> longer plays the newest games titles, all the better.
I think that's the "conspiracy theory" aspect of it. It's hard to
explain otherwise why the company with a 5 year old product with a huge
install base won't even do its existing customers the perfectly
reasonable favor of posting a few PDF files on their website, or making
printed copies available for a reasonable fee. I think they really do
want to disassociate themselves from any past products in the hopes
you'll just give up and upgrade. And many people give in to the
coercion -- that's the advantage of having a support monopoly on your
product, after all. You get to decide when your customers upgrade.
You only hope they upgrade to *your* product though, so you don't want
to start screwing them *too* soon into the previous product's life
> I hope that will change. I _think_ it will change. But it's likely to
> change only when we get rid of the "new gfx card every six months"
That's probably not the key issue. After all, most hardware is on an
evolutionary cycle now. NVIDIA's new products add the latest whiz-bang
features to a core that's going on six years old now. ATI's base
hardware is younger, but still an evolutionary approach is taken there
too. Matrox has numerous products based on the design of the G400,
which was a step up from the G200 -- only recently did they feel it was
time to make the leap to a new architecture. VIA/S3 chips haven't
changed much since the Virge days. These aren't revolutionary things
that are happening when "new gfx card every six months" comes out.
It's really just a cycle of control; it makes more sense economically to
control people than to give them the information they would need to
compete with you on support. (Or to create a cloned product....)
In my opinion, open hardware is the only solution. We can chase
proprietary vendors till we all turn blue, but in the end, they dictate
the terms under which their hardware will remain viable for use.
I wish that would change, but the trends seem to be going quickly away
=66rom openness in the graphics card market, with no sign of reversal.
That's driven by demand after all -- people want whiz-bang features,
they don't want openness. The few who do want openness are ignored
because they are perceived to be too costly to appease.
In an open software architecture like the DRI, we should do our best to
support proprietary vendors when they give us the means to do so, but
all the pissing and moaning about what they will and won't do should go
either to /dev/null or, more productively, towards opencores.org and a
fully open windowing accelerator and programmable 3D graphics pipeline
core. The technology is there, it just needs the mindshare and people's
willingness to embrace it.
Imagine, an FPGA on an AGP card... upgradable when you feel like it, and
with every internal working available for your perusal. Need a faster
board, pop out the FPGA with a new model. Need more features, upload
a new version of the core. If that prospect doesn't make you tingle, I
don't know what would!
Ryan Underwood, <nemesis@...>