OpenBSD Journal

OLPC hurts wireless documentation efforts

Contributed by deanna on from the let-them-eat-binary-blobs dept.

The public clamour to free wireless chipsets has heated up significantly in the days since Theo de Raadt accused Intel Corporation of being an Open Source Fraud.

Today he decided to step forward and make public his criticisms of Red Hat Software, speaking out against the OLPC agreement to build the laptops with millions of non-free wireless devices manufactured by Marvell. De Raadt quoted Jim Gettys of Red Hat/OLPC:

"Free and open software is a means to an end, rather than the sole end unto itself for OLPC."

A copy of mail he'd sent to Gray was posted to the OpenBSD misc mailing list (misc at openbsd.org), where the discussion continues. Feel free to join in.

To: Jonathan Gray
cc: dcbw at redhat.com, jg at laptop.org, mtosatti at redhat.com
cc: rms at gnu.org
cc: deraadt
Subject: Re: Marvell 88W8388 documentation
Date: Wed, 04 Oct 2006 12:38:34 -0600
From: Theo de Raadt deraadt at cvs.openbsd.org

> Please correct me if I am wrong but it seems that documentation
> for Marvell's 88W8388's is not publically available without
> signing an NDA?
>
> If this is the case why did a project that seems to pride
> itself of openess agree to deal with such a company?
> Drivers written under NDA tend to be full of magic numbers,
> near impossible for others to properly maintain and
> totally against the spirit of open projects.
>
> I really think you should push for Marvell to give out
> documentation without them forcing NDAs onto people.
> Failing that I'm sure there are other vendors
> who would be willing to be more helpful.

Jonathan showed me this mail he sent you about your NDA "cooperation"
with Marvell for the wireless chip that you want to use for the OLPC
project, so that Marvell will write you special hacks to do low-power
mesh networking while the main cpu is powered off.  This does not
gaurantee Marvell is going to be open and release documentation for
mesh networking while the main cpu is powered off.  This does not
gaurantee Marvell is going to be open and release documentation for
their chips though.

When large players like you make such private agreements with such
secretive vendors, you work against our common goals of getting more
open documentation for devices.  It is only with open documentation
that OS groups can increase device support, and later -- keep the
device drivers reliable after the device is EOL'd by the vendor.

I've heard claims that you (OLPC members, Red Hat employees) think
this relationship with Marvell will eventually prompt/teach them to be
more open in time.  Do you not realize how much of a DELUSION the
history of free/open operating systems shows that point of view to be?
Very few chip vendors have ever opened up unless they were pushed, let
alone Marvell (who I  am led to believe also has NDA's with Red Hat
employees for the Marvell Yukon/Yukon 2 gigabit ethernet chips --
again one of the few closed chips).

It is clear that your choices are not about opening up Marvell, but
simply commercially expedient and hurtful to our common cause.  You
came to Marvell with potential sales of millions of units, and then
completely wimped out in demanding ideals that you say you share with
the community.  Now other companies like Intel, Broadcom, and TI can
say to us "Why should we open up, Marvell did not have to".

So I must say I am extremely dissapointed you have chosen to work
against the very obvious goals of "open", and I hope that in time you
are made to feel ashamed of the choice you have made.

------------------------------------------------------------

A strong statement was also issued by OpenBSD developer Bob Beck.

"No, it's real simple.

Red Hat (and a number of other linux distros) are morally
bankrupt.

By that I mean the sit under the linux banner touting the GPL,
and yet this is not how they act. They act in a way that helps
to ensure that GPL'ed software can not continue to be written.

I am not a GPL fan, but I'll defend someone's ability to write
such software agressively. I consider it the same thing as defending
freedom of speech - it's defending your ability to buy something and
use it in the way you see fit, as opposed to buy something and use it
only where and when the manufacturer tells you you can.

The only reason you see only OpenBSD doing this is because the
mass market and media out there is too busy being a linux
fanboys to notice and ask the questions they should. All the
media is seeing is "we can use this cool new thing in linux"
and they are missing the point of "you have just been sold
out". That's not a diss of Linux in general, it's a diss of a
number of short sigheted developers who support that, and a
diss of the techincal media who ignores the fact that your
freedoms go down the tank by making these compromises. The
attitude that the end (hardware support) justifies the means
(complete sacrifice of the principles the thing was written
under in the first place) has to stop.  The fact that Theo can
end up being a professional shit-disturber and find these
things so easily is a huge inditement of the community and the
media reporting on it that we read.

Allowing developers to sign NDA's with companies to add
support to an OS that purports to be free is letting them have
a Munich conference with your freedoms. You aren't invited -
and they're carving you up while doing a Chamberlain and
saying "look - device support in our time - they'll be much
better behaved now." We all know how well that worked out, and
this is no different."

-Bob

(Comments are closed)


Comments
  1. By Tassilo (90.6.22.102) on

    Well - the only thing I want to say is that I'm really happy that people like Theo exist - a big "thank you" for continously bugging these "pseudo do-gooders" and not shutting up like the linux people did by don't caring about their own ideologies.
    Thx !

  2. By Anonymous Coward (69.70.207.240) on

    Glad to see this - isn't this just too true and a shame on RedHat and others...

    I don't see Linus Torvalds standing up for OSS like this (OSS in general), hell, not even for Linux itself! What a shame there too if you ask me.

    OpenBSD is really paving the way for OSS and standing up for all OSS as a whole, we should all step up and stand together for what we believe in and for what OSS is meant to be! For those who don't have the balls, go out and buy a strap-on dong with some (no names mentioned)...

    Hope to see this original post on /. too.

    OpenBSD rocks!

    Just my $0.02

    Comments
    1. By rmg (208.181.115.2) on

      > I don't see Linus Torvalds standing up for OSS like this (OSS in
      > general), hell, not even for Linux itself! What a shame there too
      > if you ask me.

      Not to detract from Theo's work, but Linus has done his own form of protesting. It's really just a difference in goals. For example, refusing to use GPLv3.

      I'm grateful for all the outspoken open source idealists with more spine and weight than I. Keep up the good work!

      Comments
      1. By Anonymous Coward (66.11.66.41) on

        > > I don't see Linus Torvalds standing up for OSS like this (OSS in
        > > general), hell, not even for Linux itself! What a shame there too
        > > if you ask me.
        >
        > Not to detract from Theo's work, but Linus has done his own form of protesting. It's really just a difference in goals. For example, refusing to use GPLv3.

        Linus has done shitall. Refusing to use the GPLv3 is irrelivant, choosing one not very free license over another not very free license has nothing to do with getting open hardware. Why isn't he brought nvidia and ati to court for violoating the GPL? He *says* that those binary drivers are against the GPL, yet he does nothing about it. Why doesn't he speak out against linux distros that do everything they can to include binary shit and lock users out of their own computers?

        Comments
        1. By Anonymous Coward (151.188.247.104) on

          > > > I don't see Linus Torvalds standing up for OSS like this (OSS in
          > > > general), hell, not even for Linux itself! What a shame there too
          > > > if you ask me.
          > >
          > > Not to detract from Theo's work, but Linus has done his own form of protesting. It's really just a difference in goals. For example, refusing to use GPLv3.
          >
          > Linus has done shitall. Refusing to use the GPLv3 is irrelivant, choosing one not very free license over another not very free license has nothing to do with getting open hardware. Why isn't he brought nvidia and ati to court for violoating the GPL? He *says* that those binary drivers are against the GPL, yet he does nothing about it. Why doesn't he speak out against linux distros that do everything they can to include binary shit and lock users out of their own computers?


          The reason that he and other kernel developers haven't taken nVidia, etc. to court is because nVidia, etc. use kernel shims that call some user-space "cross-platform on x86" binary. That's how they get away with it.

          Now, why doesn't he publicly object to it? Simple; he's too busy wanting to be "an Oppenheimer" and not caring about what hardware manufacturers do. What he seems to forget is that if all the hardware manufacturers do this, then his kernel won't mean jack, nor will any other Free Software kernel. Why? Because it won't run anywhere!

          At least RMS is publicly, and loudly, opposing this kind of crap.

          Comments
          1. By gardion (24.87.50.74) on

            > > > > I don't see Linus Torvalds standing up for OSS like this (OSS in
            > > > > general), hell, not even for Linux itself! What a shame there too
            > > > > if you ask me.
            > > >
            > > > Not to detract from Theo's work, but Linus has done his own form of protesting. It's really just a difference in goals. For example, refusing to use GPLv3.
            > >
            > > Linus has done shitall. Refusing to use the GPLv3 is irrelivant, choosing one not very free license over another not very free license has nothing to do with getting open hardware. Why isn't he brought nvidia and ati to court for violoating the GPL? He *says* that those binary drivers are against the GPL, yet he does nothing about it. Why doesn't he speak out against linux distros that do everything they can to include binary shit and lock users out of their own computers?
            >
            >
            > The reason that he and other kernel developers haven't taken nVidia, etc. to court is because nVidia, etc. use kernel shims that call some user-space "cross-platform on x86" binary. That's how they get away with it.
            >
            > Now, why doesn't he publicly object to it? Simple; he's too busy wanting to be "an Oppenheimer" and not caring about what hardware manufacturers do. What he seems to forget is that if all the hardware manufacturers do this, then his kernel won't mean jack, nor will any other Free Software kernel. Why? Because it won't run anywhere!
            >
            > At least RMS is publicly, and loudly, opposing this kind of crap.

            You seem to forget that the non-windows/ non-macintosh market is very small. I use Ubuntu linux on all my home machines, but I am aware that most people don't use anything other than Windows or Macs. Quite honestly, at this point it would be very easy for Nvidia and ATI to stop any support of Non -windows/ non-mac environments. I hardly think it would make much of a dent in their market share.

            However, if the linux or bsd desktop market starts making significant inroads on then, I think they can begin to call the shots more.

            As for Linus, he does his part by constantly changing the kernel making it very difficult for people to supply binary only drivers. He sees no reason for having a non-changing api as that encourages people to write binary drivers. This makes hardware that have open-sourced their drivers or given the documentation an advantage over those who do not.

            Ultimately Linus, thinks customers need to make the choice about free hardware. If they refuse to buy products without open driver support then, those products won't succeed. Intel is opensourcing it's graphics drivers and people are talking about buying products that use intels graphics cards rather than nvidias or ATI's.

  3. By Anonymous Coward (68.167.146.78) on

    Actually, I'm one of those GNU/Linux afficionados that you so derisively talk about so often. Like you, I too am appalled at this. Red Hat knows better than this. I wonder, now, if they're getting some kind of kickback from Marvell; I hate to think this, but that's the only logical explanation that comes to my mind. Does anyone have another, better one? I sure hope so.

    What you are seeing is the difference between the Free Software movement and the Open Source movement. The Open Sourcers, like Eric Raymond, are willing to make "pragmatic", as they call them, compromises with proprietary, closed vendors. Free Software advocates, like us, do NOT. We know better, and we act on it.

    As acidic as Theo can be personally, he is absolutely right on this and has been for several years. Fortunately, Richard Stallman also agrees on this point; there is no good reason to keep programming specs hidden, especially for something as noble as OLPC is supposed to stand for.

  4. By Charles Hill (76.16.134.135) on

    Is there something like a hardware compatability list, except instead of listing "it works" listing "it is free and properly documented"?

    I'd be interested in a list of what types of hardware need blobs and what types don't. I know network controllers, both wireless and wired, need blobs. What about north/southbridge chipsets and other components?

    I'd also be interested in knowing which ones are truely open because they provided proper docs, and which ones are truely open in spite of the manufacturers. The ones that were reverse engineered, and to what degree.

    There are some comments in the OpenBSD Crypto page about hardware accelerators, but some of that info is real old and I'm not sure if it is 100% accurate. For example, was the public-key accelerator on the HiFn 795x chips ever supported?

    Comments
    1. By Charles Hill (76.16.134.135) on

      To partially answer my own question, the FSF has a good deal of this information available. Granted, it is biased towards GPL and GPL-compatible licenses, but it is a start.

      http://www.fsf.org/resources/

      The Free BIOS and Hardware Resources sections are a good place to start.

    2. By Terrell Prude', Jr. (68.167.146.78) on

      > Is there something like a hardware compatability list, except instead of listing "it works" listing "it is free and properly documented"?
      >

      Another one is http://www.vendorwatch.org/. I've been updating some of the pages with exactly this information.

      --TP

  5. By Anonymous Coward (204.13.236.244) on

    Well... seams the devil showed his face once again. ;]

    There where some "hackers" who stole the MS-Source-Code and the Cisco-Source-Code...
    Maybe some day such a hacker (who belive sin the ethic..) may hack INTEL or MArvell or or or... and forks the Docs, the Firmware(includes the Source of course ;]) and all other stuff to a P2P-Network.

    Who knows...
    As consumer you`re the gay if you write to INTEL....
    As developer you`re the nobody as long as your adress doesn`t end up with @microshit.com

    Well.. sometimes words are not enought....
    Propably those in the scene who may have the knowledge to "change" something should think about the situation and "change" something...
    And I`m not telling you to hack the Servers to get what you need... or?

    Comments
    1. By Anonymous Coward (89.57.20.139) on

      ...
      > There where some "hackers" who stole the MS-Source-Code and the Cisco-Source-Code...
      > Maybe some day such a hacker (who belive sin the ethic..) may hack INTEL or MArvell or or or... and forks the Docs, the Firmware(includes the Source of course ;]) and all other stuff to a P2P-Network.
      >
      ...

      So, use the stolen docs to write drivers and let them sue you?

    2. By Anonymous Coward (69.207.171.114) on

      > As consumer you`re the gay if you write to INTEL....
      > As developer you`re the nobody as long as your adress doesn`t end up with @microshit.com

      After reading this post I think you`re the crazy.

  6. By Terrell Prude', Jr. (68.167.146.78) on

    I'm another "Linux fanboy", and I too am appalled. I just now wrote and sent an email to the OLPC folks listed in Theo's email. Here it is.


    Folks,

    I just recently heard about the Marvell 88W8388 wireless chipset to be used in the "$100 laptop", and it is also my understanding that the programming specifications for this chipset are not publicly available without signing a restrictive Non-Disclosure Agreement (NDA). I am in amazement at this, given the OLPC project's stated goal of helping out children through the open proliferation of information. It is especially surprising, since you're planning to run a Free Software platform (GNU/Linux) on these same laptops!

    Including this Marvell gear for OLPC goes totally against the spirit of your own stated goal. Please do not go forward with this, even if you think that it can be "expedient". Mussolini and Hitler were also very "expedient", if you recall. It is also "expedient" to conduct illegal wiretapping, which my own government's administration was recently caught doing. However, it does not promote freedom at all. On the contrary; Marvell's act is simply another form of "lock-in", akin to what Microsoft does with its closed MS Office file formats (hence OpenDocument). Marvell's act of refusing to allow Free Software programmers to write truly Free (as in Freedom) device drivers for this wireless chipset should immediately disqualify them from something as noble as OLPC is supposed to be.

    Jonathan Gray has already outlined the technical problems with this. Theo de Raadt eloquently explained the ethical problem with it, and I hear that Dr. Stallman agrees. Well, so do I. Mr. Gray and Mr. de Raadt are correct in that there are other vendors who would gladly cooperate with the Free Software community and publish the programming specifications for their wireless chipsets. Why? Simple; they want those millions of sales, too, just like Marvell does!

    Marvell will of course try to argue, "oh, we need to protect our 'intellectual property', our 'secret sauce', from competitors." Releasing the programming specifications does not "reveal" their "secret sauce." It merely allows Free/Open Source developers to write device drivers for the chipset so that it can do what it was designed to do. That's all. Nothing more. There is no valid reason for Marvell to keep these programming interfaces secret.

    Additionally, some devices require "firmware images" to be loaded into the wireless card itself in order to function at all. Perhaps Marvell's wireless chipset does, too. The source code of this "firmware image" is not required, nor are we asking for it. We are asking only for rights to redistribute the firmware image, if Marvell's chipset indeed requires one to function. Doing so does not reveal the manufacturer's "secret sauce", either. Ralink and Realtek do this all the time. Even Intel allows this redistribution for the various firmwares for their wired FastEthernet interfaces.

    Please get in contact with Marvell and demand, in polite, but direct, language, the following:

    1.) Publish the programming specs for any of their chipsets used in the OLPC laptop.
    2.) Grant a non-revocable license to distribute any "firmware image(s)" that the chipset requires to operate.

    Tell them that, if they don't do these two things, they're out.

    Please do not allow a repeat of the well-known Broadcom wireless situation. Give Ralink or Realtek a call. I'm sure that they would LOVE to talk to you.

    Sincerely,

    Terrell Prude', Jr.

  7. By Anonymous Coward (66.46.215.210) What is the official Red Hat reaction? on

    What is the official Red Hat reaction? For sure, I will follow this one.

    Comments
    1. By Anonymous Coward (206.59.235.113) on

      > What is the official Red Hat reaction? For sure, I will
      > follow this one.

      Redhat don't do "public reactions", they prefer to skulk on private mailing lists and hide behind individual developers who work for Redhat but "aren't speaking for Redhat".

    2. By Peter Hopfgartner (217.199.20.139) on http://www.hopfgartner.it

      > What is the official Red Hat reaction? For sure, I will follow this one.

      It seems that Theo's letter has forced some public reaction: http://www.gettysfamily.org/wordpress/?p=27

      IMO, worthwhile reding.

      Regards,

      Peter

      Comments
      1. By CODOR (67.158.69.254) on

        > It seems that Theo's letter has forced some public reaction: http://www.gettysfamily.org/wordpress/?p=27

        I don't understand; that article seems to talk almost entirely about the firmware, while Theo and Bob Beck are looking for documentation for the Marvell chip itself.

        Comments
        1. By sthen (85.158.44.149) on

          > > It seems that Theo's letter has forced some public reaction: http://www.gettysfamily.org/wordpress/?p=27
          >
          > while Theo and Bob Beck are looking for documentation for the Marvell chip itself.

          well it's not exactly docs, but better than we had before...
          http://dev.laptop.org/git?p=olpc-2.6;a=tree;f=drivers/net/wireless/libertas

        2. By Charles Hill (76.16.134.135) on

          > > It seems that Theo's letter has forced some public reaction: http://www.gettysfamily.org/wordpress/?p=27
          >
          > I don't understand; that article seems to talk almost entirely about the firmware, while Theo and Bob Beck are looking for documentation for the Marvell chip itself.

          The article does mention that most likely the documentation they're asking for doesn't exist, is scattered in many places or is tied up with future products. That last one is a killer.

          Considering the Marvell chipset is NOT a commodity and is UNIQUE in the way it operates, I can see them not being interested in disclosing all the documentation.

          Commodity items are one thing. Everybody and his dog in China can make an ethernet card, or a standard WiFi card, but what Marvell has no one else does.

          The article DOES mention that they ARE working on a free firmware that will most likely be released GPL, LGPL or MIT. They are also working on a GPL driver called Libertas.

          [Please insert BSD vs GPL rant below.]

          Comments
          1. By Timothy Schmidt (71.205.223.8) timschmidt@gmail.com on

            > Considering the Marvell chipset is NOT a commodity and is UNIQUE in the way it operates, I can see them not being interested in disclosing all the documentation.
            >
            > Commodity items are one thing. Everybody and his dog in China can make an ethernet card, or a standard WiFi card, but what Marvell has no one else does.

            Huh? The Marvell chipset has an embedded ARM core and some ram (96 or 32k IIRC), which allows it to do whatever you can code into that much space. There are several other wireless chipsets on the market with _very_ similar configurations.

  8. By Anonymous Coward (128.171.90.200) on

    So they are planning to ship these laptops to kids in developing countries under the guise that the children will be able to get-their-hands-dirty with the machine and then they close off parts of it. Looks to me as if they have been peddling false promises.

Latest Articles

Credits

Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original deadly.org with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]