lkml.org 
[lkml]   [1998]   [Oct]   [4]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: UDI and Free Software


On Sun, 4 Oct 1998, Richard Stallman wrote:

> UDI and Free Software
>
> Intel has a project called UDI: Uniform Driver Interface. The idea is
> to define a single interface between operating system kernels and
> device drivers. What should the free software movement make of this
> idea?

Here we go again...

First: it's not an Intel project, from what I've seen of the websites and
the UDI guy's posts to Linux-kernel it would seem that Intel has seen the
logic and jumped on the bandwagon of an already rolling project.

> If we imagine a number of operating systems and hardware developers,
> all cooperating on an equal footing, UDI (if technically feasible)
> would be a very good idea. It would permit us to develop just one
> driver for any given hardware device, and then all share it. It would
> enable a higher level of cooperation.
>
> When we apply the idea to the actual world, which contains both free
> software developers seeking cooperation, and proprietary software
> developers seeking domination, the consequences are very different.
> No way of using UDI can benefit the free software movement. If it
> does anything, it will divide and weaken us.
>
> If Linux supported UDI, and if we started designing new drivers to
> communicate with Linux through UDI, what would the consequences be?
>
> * People could run free GPL-covered Linux drivers with Windows systems.
>
> This would help only Windows users; it would do nothing for us users
> of free operating systems. It would not directly hurt us, either; but
> the developers of GPL-covered free drivers could be discouraged to see
> them used in this way, and that would be very bad. It can also be a
> violation of the GNU GPL to link the drivers into a proprietary
> kernel. To increase the temptation to do so is asking for trouble.
>

I can't imagine UDI drivers being written by Linux developers since I
can't imagine UDI ever being as fast as a native driver.

> * People could run non-free Windows drivers on GNU/Linux systems.
>
> This would not directly affect the range of hardware supported by free
> software. But indirectly it would tend to decrease the range, by
> offering a temptation to the millions of GNU/Linux users who have not
> learned to insist on freedom for its own sake. To the extent that the
> community began to accept the temptation, we would move to using
> non-free drivers instead of writing free ones.
>

The manufacturers who don't provide specs will simply have a slower
product, people will naturally lean to wards the faster, right now I'm
tired of telling people "sorry Linux doesn't support that" I'd rather say
"yea it works but it's gonna be slower then the competition"

What I find frightening about this whole argument is that so far the only
person who has even mentioned a technical viewpoint is Alan Cox.. Why is
this ?

> UDI would not in itself obstruct development of free drivers. Soif
> enough of us rejected the temptation, we could still develop free
> drivers despite UDI, just as we do without UDI.

> But why encourage the community to be weaker than it needs to be? Why
> make unnecessary difficulties for the future of free software? Since
> UDI does no good for us, it is better to reject UDI.
>

That's bogus, all we ever need or should do is provide the interface, if
that is all we do then it can do nothing but benefit us.

> Given these consequences, it is no surprise that Intel has started to
> "look to the Linux community for help with UDI." How does a rich and
> self-seeking company approach a cooperating community? By asking for
> a handout, of course. They have nothing to lose by asking, and we
> might be caught off guard and say yes.

> Cooperation with Intel is not out of the question. We should not
> label Intel, or anyone, as a Great Satan. But before we participate
> in any proposed deal, we must judge it carefully, to make sure it is
> advantageous for the free software community, not just for proprietary
> system developers. On this particular issue, that means requiring
> that cooperation take us a step further along a path that leads to the
> ultimate goal for free kernels and drivers: supporting *all* important
> hardware with free drivers.

> One way to make a deal a good one could be by modifying the UDI
> project itself. Eric Raymond has proposed that UDI compliance could
> require that the driver be free software. That would be ideal, but
> other alternatives could also work. Just requiring source for the
> driver to be published, and not a trade secret, could do the
> job--because if that driver is not free, it would at least tell us
> what we need to know to write a free driver.
>
> Intel could also do something else, outside of UDI, to help the free
> software community solve this problem. For example, there may be some
> sort of certification that hardware developers seek, that Intel plays
> a role in granting. If so, Intel could agree to make certification
> more difficult if the hardware specs are secret. That might not be a
> complete solution to the problem, but it could help quite a bit.

> One difficulty with any deal involving UDI is that we would do our
> part for Intel at the beginning, but Intel's payback would extend over
> a long time. In effect, we would be extending credit to Intel. But
> would Intel continue to repay its loan? Probably yes, if we get it in
> writing and there are no loopholes; otherwise, we can't count on it.
> Corporations are notoriously untrustworthy; the people we are dealing
> with may have integrity, but they could be overruled from above, or
> even replaced at any time with different people. When making a deal
> with a corporation, always get a binding commitment in writing.

It Seems to me that Intel's payback comes with our success, I've noticed
that Linux tends to show the difference between good and bad hardware
more then windows does, and it seems to make much better use of newer CPU
designed as well. It seems to me that is a much better approach then
wasting CPU cycles as is done on certain non free Operating systems.


> Given Intel's involvement in I2O, a broad plan to keep hardware
> specifications secret, it is not likely that they will accept a deal
> that gives us what we need. In fact, UDI seems like a plan to make it
> easier to keep specifications secret.

Once again I20 would blow if Linux couldn't handle it. How good would it be
if all they could get is slower than pci ?

What UDI would do for us is level the playing field in our favor, right
now no one has to support Linux, and I run into that problem when I try to
advocate Linux. Once they are a Linux user then it simply becomes a speed
issue.

Also given Linux importance to Intel atm, it seems to me some people
(Linus) could push for changes in the way I20 is done.

If UDI is feasble to implemnt, it would allow us to grab more of the
market, once that happens everthing else becomes a performance issue.

Gerhard



> Still, there is no harm in keeping the door unlocked, as long as we
> are careful whether we let Intel in.
>
>
> Copyright 1998 Richard Stallman
> Verbatim copying and distribution is permitted in any medium
> provided this notice is preserved.
>

PS Why are all my GNU man pages way out of date?

--
Gerhard Mack
irc-admin skyline.starchat.net

gmack@imag.net
InnerFIRE@starchat.net

As a computer I find your faith in technology amusing.



-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:44    [W:0.108 / U:0.168 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site