OpenBSD Journal

BSDCan 2006 - Why should I care where my device drivers come from?

Contributed by ben on from the last-bsd-hippie dept.

Deanna Phillips wrote:

BSDCan 2006 was my first conference, though I've been using Unix (professionally and as a hobbyist) for over ten years now, participating in political and tech activist groups all the while. The conference left me pretty well disillusioned, and the idea of principles when it comes to software has been a theme with me ever since.

I used to run into a lot of people who were very political about the software they used, and I've always been one of those types myself, so it was a bit crushing to see so many Macs and Blackberrys and Windows laptops running at that conference, not to mention the crowds of FreeBSD people and their followers dominating the entire scene.

I spoke with many people, mostly users but not all, and their attitudes about software were distressingly similar: they'll use anything as long as it does what (they think) they need it to do, regardless of the licensing or reputation of the commercial vendors they are supporting and (often ostentatiously) advocating.

What I'm left with is the overwhelming feeling that many people, users and developers alike, have become more selfish and pragmatic about the software they use and distribute; it seems as if the idealists have all but died out.

They used to be clustered around Debian, and maybe they still are, but I am not a fan of the GPL or Linux in general. However, I've always had a deep respect for Richard Stallman and his passion for keeping software free from commercialization. So to further demonstrate my disillusionment with the current free software scene: if I had a loonie for every smug comment about Mr. Stallman's personal hygiene, coming from some OS X user (at a technical BSD conference!), I'd be funding this year's OpenBSD hackathon.

Why not follow Stallman around and help his movement instead? Well, it's firstly a matter of taste and secondly a matter of politics. Whereas Stallman's method seems to be that of peaceful protest, I prefer the tactics used by the OpenBSD project, which bear more resemblance to what's often called `direct action' in activist groups.

OpenBSD feels very strongly about the freedom to write and distribute whatever software they wish, for whatever reason, and when something gets in the way, they don't wave signs; they do something about it.

Some may already be up in arms about my reference to FreeBSD with regard to political disillusionment. Well, this is due to their decision to ignore the current OpenBSD campaign to stop binary blobs from gathering momentum and being incorporated into otherwise or formerly free operating systems. Some of these projects, including FreeBSD, have slipped the binary drivers right in; ostensibly so that the user may get right to work on her hardware. No need to worry about the political ramifications of giving money to a vendor that refuses to supply source code or specs for the hardware that they sell. What a hassle that would be! Users just want to get things done, right? Who cares, as long as it runs?

Well, I care. Other people used to care, and OpenBSD seems to care. The waves of blob-free drivers, being written to support closed hardware show how much work they are putting into this campaign.

Does anyone else care? I overheard one of the OpenBSD devs asking the BSDCan organizers for a table. Yes, he had to ask for a table, and the response wasn't apologetic: it exhibited obvious annoyance.

Thus, my disillusionment. Maybe it's time to stop caring about software; maybe it's gone too far into the realm of mass media, and my complaints are about as silly as someone turning on broadcast TV and complaining about the commercials. Sadly, I suspect that the majority of the people I ran into at BSDCan would respond to this with: let them use TiVo!

Note: This article was accidentally deleted. Unfortunately, any comments that were associated with it were deleted as well. We apologize for any inconvenience this may have caused.

(Comments are closed)


Comments
  1. By Daniel Hartmeier (62.65.145.30) daniel@benzedrine.cx on http://www.benzedrine.cx/dhartmei.html

    While the story could be restored, we unfortunately lost all comments posted to it so far. I'm very sorry about that. As far as I remember, they were all non-trivial and valuable. If you remember what you posted, feel free to reconstruct and resubmit from memory. My apologies.

    There are currently also some (unrelated) networking problems (pages load slowly and temporarily stall for several seconds), which are being worked on. Thanks to everyone who noticed and wrote in about it.

    Comments
    1. By Anonymous Coward (70.74.75.200) on

      > While the story could be restored, we unfortunately lost all comments posted to it so far.

      About 24 hours ago, I only saw 8 posts, but I don't know when this loss occurred. However, you can recover the 4 posts from Google cache, if you want.

      http://72.14.209.104/search?q=cache:n_cmQJtFgCYJ:www.undeadly.org/cgi%3Faction%3Darticle%26sid%3D20060518224039%26mode%3Dflat+%22BSDCan+2006+-+Why+should+I+care+where+my+device+drivers%22&hl=en&gl=ca&ct=clnk&cd=1&ie=UTF-8

      or

      http://www.google.ca/search?hl=en&ie=ISO-8859-1&q=%22BSDCan+2006+-+Why+should+I+care+where+my+device+drivers%22&meta=

    2. By Deanna Phillips (70.48.229.109) deanna@sdf.lonestar.org on

      Thanks for restoring the article. It's too bad that the comments were lost, but I remember the sentiments very well, and found them encouraging.

      One thing I'd like to say, for those who took my words personally and got defensive: not every can be a saint -- it takes effort that could (should?) otherwise be spent on oh, writing code. :)

  2. By Doctor Coldfoot (204.212.175.30) drcoldfoot@hotmail.com on

    For One, binary drivers (blobs), in an OpenSource OS completely undermine the validity of the OS, and actually is pretty scary to me. What is so secret in the code for a Video Card, NIC, SCSI Adapter, or Wifi Card? COuld it be Intellectual Property? Or could it be more seedy? Maybe a keylogger, or other for m of information gathering. My Point is, Hardware should be Hardware. Drivers that should run them should be Open Sourced. Propriatary engines should be on the Chip firmware and not on the OS in my Opinion.

    Comments
    1. By Anonymous Coward (68.104.1.58) on

      workarounds for bugs in the hardware mostly

      Comments
      1. By David Gwynne (24.42.21.250) loki@animata.net on

        > workarounds for bugs in the hardware mostly

        I'm yet to read a vendor supplied driver that didn't have bugs of its own and occasionally drop support for large revisions of the hardware. And that's the open source drivers. I hate to think what is hiding inside their binary only stuff.

    2. By Charles (216.229.170.65) on

      > For One, binary drivers (blobs), in an OpenSource OS completely undermine the validity of the OS, and actually is pretty scary to me. What is so secret in the code for a Video Card, NIC, SCSI Adapter, or Wifi Card? COuld it be Intellectual Property? Or could it be more seedy? Maybe a keylogger, or other for m of information gathering. My Point is, Hardware should be Hardware. Drivers that should run them should be Open Sourced. Propriatary engines should be on the Chip firmware and not on the OS in my Opinion.

      * * *

      Competition is one factor, so are resources.

      Frequently the hardware on competing products is identical, or very close. In the case of OEM equipment, you will frequently have several products that are the exact same thing except the logo and the drivers.

      In the video card world it is well known that drivers are a big differentiator. You can take identical hardware (GPU, RAM, bus, clock, board layout, etc.) with different drivers and get wildly different results. The same goes for networking adaptors, which can vary widely on CPU utilization and overall thruput.

      If a new player wants to enter the game, they have to spend some bucks on software R&D and tweak things to differentiate. If all of it was totally free then you'd have a bunch of newcomers pop in, reselling reference hardware that is fabbed at the exact same Chinese factory that they sell for just $1 or $2 lower than everyone else because the only thing they do it rebox reference hardware/software. Hell, a lot of that goes on now.

      Competition gets a lot more cutthroat between vendors, who now focus almost exclusively on lowest cost and not quality. "Good enough" becomes king.

      As for refernece drivers, which most graphics card vendors use, ATI & nVidia don't want to tip their hand. They are usually very close in performance and every little bit helps. While the respective engineers do know what they are doing, they quite probably don't know the BEST way on everything. If nVidia got a hold of ATI's source it is quite possible there would be several "oh, they do it THAT way" moments.

      Take for example -- a very primitive example -- sorting code. CS101 has people code up shell, bubble and quick sort to show the difference in speed. Each does the same thing, sort a list, but they go about it in algorithmically different ways and there is a BIG difference in speed. Each also consume different amounts of CPU horsepower, which means heat. While the engineers at nVidia may know about shell sort, the ones at ATI may have evolved to quick sort which is a big advantage (in most cases). ATI doesn't want nVidia to cherry pick that code and thus improve the performance of their competing product.

      With notebook components, especially wifi, think power consumption, heat and battery time. The more efficient the code, the better all those are.

      Now, from the *end user* perspective, OSS is the best simply because such cherry picking could go on and improve the performance and quality of all the code. Other people could step in and clean up the code for readability (aka maintainability) and then improve stability, reliability, etc. Everyone wins.

      But from a hardware vendor perspective that is one more step in reducing their product to a commodity, with the cooresponding drop in profit margin. Why do you think nVidia and ATI always push those high end, $400+ video cards? Margin, margin, margin.

      Comments
      1. By tedu (69.12.168.114) on

        > In the video card world it is well known that drivers are a big differentiator. You can take identical hardware (GPU, RAM, bus, clock, board layout, etc.) with different drivers and get wildly different results. The same goes for networking adaptors, which can vary widely on CPU utilization and overall thruput.
        >
        > If a new player wants to enter the game, they have to spend some bucks on software R&D and tweak things to differentiate. If all of it was totally free then you'd have a bunch of newcomers pop in, reselling reference hardware that is fabbed at the exact same Chinese factory that they sell for just $1 or $2 lower than everyone else because the only thing they do it rebox reference hardware/software. Hell, a lot of that goes on now.

        this is a load of crap.

        1. nobody really wants a big vendor driver, since they tend to suck (usually try to share code between too many platforms, meaning they don't take advantage of common 802.11 code, or mii code, ...). or the nvidia driver that means you have to replace all the X opengl libraries.

        2. "write 4 to draw a red pixel, write 6 to draw a green pixel" cannot be some sort of secret.

        3. reference hardware? nobody sells "generic" video chips. nvidia will not go out of business because they revealed how to program their chip.

      2. By Anonymous Coward (68.104.1.58) on

        hardware documentation should be fine then?

        Comments
        1. By dlg (24.42.21.250) loki@animata.net on

          > hardware documentation should be fine then?

          that's an awesome question (and no, for once i am not being sarcastic). who cares if a vendor only supplies a binary driver if we had documentation as well? we could write our own in that situation.

    3. By Anonymous Coward (216.135.89.5) on

      > For One, binary drivers (blobs), in an OpenSource OS completely undermine the validity of the OS, and actually is pretty scary to me. What is so secret in the code for a Video Card, NIC, SCSI Adapter, or Wifi Card? COuld it be Intellectual Property? Or could it be more seedy? Maybe a keylogger, or other for m of information gathering. My Point is, Hardware should be Hardware. Drivers that should run them should be Open Sourced. Propriatary engines should be on the Chip firmware and not on the OS in my Opinion.


      Meanwhile, you most likely run a BIOS that uses a similar blob.

      When is OpenBIOS from OpenBSD going to get serious?

      Comments
      1. By Anonymous Coward (24.42.21.250) on

        > Meanwhile, you most likely run a BIOS that uses a similar blob.

        Uh...

        The BIOS is there to bootstrap the machine and then hand control over to an operating system. Once the machine is up and running we don't really use it anymore. We're talking about drivers here, which are executed as part of the operating system and are relied on for proper operation of the system. Asking when we're going to replace the BIOS is irrelevant to this discussion.

        Comments
        1. By Anonymous Coward (216.135.89.5) on

          > > Meanwhile, you most likely run a BIOS that uses a similar blob.
          >
          > Uh...
          >
          > The BIOS is there to bootstrap the machine and then hand control over to an operating system. Once the machine is up and running we don't really use it anymore. We're talking about drivers here, which are executed as part of the operating system and are relied on for proper operation of the system. Asking when we're going to replace the BIOS is irrelevant to this discussion.

          I am very aware of what a BIOS is. My point is that the entire blob campaign is about having the source to what you run on your machine. The BIOS runs at very early stages and could do evil things if it wanted to. I just don't see the difference in not having source for the bios vs a hal.

          Comments
          1. By Anonymous Coward (68.104.1.58) on

            > I am very aware of what a BIOS is. My point is that the entire blob campaign is about having the source to what you run on your machine. The BIOS runs at very early stages and could do evil things if it wanted to. I just don't see the difference in not having source for the bios vs a hal.
            >
            You might as well ask for source for your whole computer, including
            full details of your processor(s). Good luck with that.

          2. By tedu (69.12.168.114) on

            > I am very aware of what a BIOS is. My point is that the entire blob campaign is about having the source to what you run on your machine. The BIOS runs at very early stages and could do evil things if it wanted to. I just don't see the difference in not having source for the bios vs a hal.

            the bios came with my computer. i'm fairly certain it's compatible with my cpu. when's the last time you got a windows driver to work on a sparc?

      2. By Anonymous Coward (68.104.1.58) on

        > Meanwhile, you most likely run a BIOS that uses a similar blob.
        yes BIOS code is a blob, but its not part of the operating system, its
        a part of the hardware (like firmware for various devices)
        >
        > When is OpenBIOS from OpenBSD going to get serious?
        you first.

  3. By Anonymous Coward (24.34.57.27) on

    Could you site where FreeBSD ships with binary drivers?

    Comments
    1. By Anonymous Coward (68.104.1.58) on

      there's the nvidia ethernet blob for one, lots more as documented earlier.

      Comments
      1. By jsg@ (210.15.216.215) on

        > there's the nvidia ethernet blob for one, lots more as documented earlier.

        As mentioned previously:

        http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/hptmv/
        highpoint ide "raid"

        http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/rr232x/
        highpoint sata "raid"

        http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/contrib/dev/nve/
        nvidia ethernet

        http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/contrib/dev/ath/public/
        atheros wireless

        http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/contrib/dev/oltr/
        olicom token ring

        They support loading closed windows networking drivers so the sky
        is the limit. http://marc.theaimsgroup.com/?l=freebsd-questions&m=114803204727296&w=2
        People even say to do this when we have freely available drivers like wpi, rtw, atw, atu etc that could easily be ported to freebsd.

        Their raid drivers are designed to act as bridges for binary only
        linux management programs so you might as well include all of them
        as well.

        There are probably more, it is hard to keep track.

        Comments
        1. By Brain (71.232.119.246) on

          > As mentioned previously:
          >
          > http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/hptmv/
          > highpoint ide "raid"
          >
          > http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/rr232x/
          > highpoint sata "raid"
          >
          > http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/contrib/dev/nve/
          > nvidia ethernet
          >
          > http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/contrib/dev/ath/public/
          > atheros wireless
          >
          > http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/contrib/dev/oltr/
          > olicom token ring

          http://cvsweb.netbsd.org/bsdweb.cgi/src/sys/contrib/dev/ath/public/

          It looks like NetBSD has some binary-only files for the Atheros driver as well. I heard stuff like this, but I never really bothered to check it out for myself. It's very disheartening :/.

          I didn't bother looking for any of the other binary-only stuff from FreeBSD in NetBSD, but the point isn't really how much there is, it's that it's there at all.

    2. By Anonymous Coward (68.104.1.58) on

      > Could you site where FreeBSD ships with binary drivers?

      i imagine the point of this question was to "prove" (by lack of replies) that there are none. didn't really work did it?

      <nitpick>and it's cite</nitpick>

  4. By Anonymous Coward (70.20.136.187) on

    > Well, I care. Other people used to care, and OpenBSD seems to care.

    If your sense is that none of these people seem to care about
    promoting good, stable, open software, why do you suppose that they
    were at the conference? I wasn't there, and don't have a clue why.

    > have become more selfish and pragmatic about the software they use
    > and distribute

    So is pragmatism such a terrible thing? People have to earn a living,
    and if everyone that "cared" insisted on working in companies that
    used only OpenBSD, their main contribution would be to the ranks of
    the unemployed. Most people can't afford to be anything but
    pragmatic: they work to earn their paycheck on the platform that their
    job requires, and promote good, free software whenever they can.

    Many people would consider those that drive Hummers to the mall to be
    environmentally insensitive. By the same reasoning in your post, one
    could conclude that anyone that ever drives any car would also be
    "selfish and pragmatic".

    > a bit crushing to see so many Macs and Blackberrys

    Maybe I'm out of the loop. I don't own a Crackberry. I don't need to
    be that connected, but for people that do need something like a
    Crackberry for work, what's the alternative? I'm not a professional
    videographer. I use Apple's moron-proof video editing tool for
    sending movies of my kid to his grandparents. It's simple. It works,
    and it's incredibly inexpensive. Are there free tools that are as
    simple and effective? There weren't only a few years ago. By choice,
    I use OpenBSD just about every day. I purchased 3.9 CDs, and I
    promote it to people whenever I can, but I don't consider myself a
    traitor when I pick up a Mac.

    The glass half-empty view would be that there are perhaps fewer
    Richard Stallman's in the world. This may or may not be true. The
    more optimistic view would be that there are more suits that view open
    source as a viable alternative than ever. The fact that we can even
    have a dialog about open drivers seems like a great deal of progress
    to me. Could we have even had this conversation 5 years ago? Linux
    users were looked upon as hippies only a few years ago. Now Linux is
    commonplace. Who knows, maybe in a few years people will come to
    their senses and realize that they can even do better than that?

    Comments
    1. By Anonymous Coward (67.64.89.177) on

      5 years ago there was more documentation. It is getting worse, not better. Linux and FreeBSD are doomed if they continue down this path, worse is that they'll take others with them as collatoral damage. The sentiment of the post is accurate, passion is dissapearing fast, videos of your kids is all that matters.

      Comments
      1. By Anonymous Coward (218.214.194.113) on

        > 5 years ago there was more documentation. It is getting worse, not better. Linux and FreeBSD are doomed if they continue down this path, worse is that they'll take others with them as collatoral damage. The sentiment of the post is accurate, passion is dissapearing fast, videos of your kids is all that matters.

        I fear you may be correct.

        George Santayana once said words to the effect that those who do not remember the lessons of the past are destined to repeat them.

        For those who don't remember.....

        OS/2 was a joint venture of MSFT and IBM. The last couple of versions ran windows apps more efficiently that widows (3.x) did. IBM proudly boasted that OS/2 was a better windows than Windows.

        As a result app coders wrote for windows only and when windows moved on from 3.x lots of apps didn't work on OS/2 and eventually OS/2 mostly died, largely due to app starvation.

        In the present parallel universe we are being white-anted by "free" software that runs windows drivers for "must have" hardware.

        The hardware makers then see less reason why they should open up the APIs and the downward spiral is entrenched.

        The word free in regard to some operating systems reminds me of the word Peace in Orwell's 1984.... (The Ministry of Peace).

        Comments
        1. By Jason LaRiviere (142.161.99.64) on

          > > 5 years ago there was more documentation. It is getting worse,

          > I fear you may be correct.

          Maybe so, but the doomsday scenarios all share the same fundamental supposition error, namely that we will never produce our own hardware.

          History *also* tells us that at one time, unix itself was closed. The folks at Berkely and GNU fixed that for us.

          The long term solution is not to continue to beg for source from uncooperative vendors, but to make your own hardware; ie. to `write your own unix', as it were.

          I hear you laughing, and see you pointing out that a raid controller cannot be made with any combination of vim and screen. But consider that once upon a time in America, some guy started a business manufacturing stuff.

          Also consider that certain Asian vendors, by the team's own admission, have proven themselves very cooperative as of late. Provided they recognize the growing market for unencumbered hardware for what it is, there's no reason partnerships can't be formed.

          Finally, I hear your complaints about cheap Asian hardware. But recall that given a market and a little profit, the asian vendor in *any* given market is the heavyweight contender.

          You are the decider,
          jsunn.

          Comments
          1. By Clay Dowling (12.37.120.99) clay@lazarusid.com on http://www.ceamus.com

            > Finally, I hear your complaints about cheap Asian hardware.

            Who's complaining about cheap Asian hardware? At least in my local Linux user's group (which has a fair number of BSD users) cheap Asian hardware is extremely popular. Nothing about inexpensive, reliable, well-supported hardware to be complaining about.

          2. By tedu (69.12.168.114) on


            > History *also* tells us that at one time, unix itself was closed. The folks at Berkely and GNU fixed that for us.

            not really. source licenses for unix were pretty readily available. how do you think sunos, aix, hpux, xenix, and all the others came about?

    2. By Anonymous Coward (70.74.75.200) on

      > If your sense is that none of these people seem to care about promoting
      good, stable, open software, why do you suppose that they were at the
      conference? I wasn't there, and don't have a clue why.

      I wasn't there either... Anyhow, "Open Source" isn't good enough, if you
      haven't been paying attention to OpenBSD. If you need proof, please
      search for ipf, apache license, xfree86, and so forth.

      > So is pragmatism such a terrible thing?

      The conference is about promoting freedom in software right? BSDs always
      claim freedom over GPL due to the fact the BSD license is free for all,
      while the GPL is not. So, why not show a little more respect than
      pragmatism at the conference, for a few days of the conference compared
      to the rest of the year?

      > Now Linux is commonplace. Who knows, maybe in a few years people will
      come to their senses and realize that they can even do better than that?

      Just because GNU/Linux is growing doesn't mean hardware vendors are
      providing documentations. The fact that GNU/Linux is continually adding
      binary blobs and NDA drivers (and now FreeBSD) should be a big alarm to
      everyone who cares about hardware freedom. Maybe in a few years people
      won't think twice about registering their OS and software with "Big Brother".
      Oh wait, that happened already. Hardware is next?


      p.s. Please forgive my long Google cache link in replying to Daniel.
      I'll use HTML for long links next time.

    3. By deanna (70.48.231.49) deanna@sdf.lonestar.org on

      > Many people would consider those that drive Hummers to the mall to be
      > environmentally insensitive. By the same reasoning in your post, one
      > could conclude that anyone that ever drives any car would also be
      > "selfish and pragmatic".

      That's really quite funny. Where do you live, Detroit? :)

      Most people outside of the USA (and many in the larger cities there) would say that driving is selfish.

      > Who knows, maybe in a few years people will come to
      > their senses and realize that they can even do better than that?

      Why wait? Why not raise some awareness now, before it gets worse?

      Comments
      1. By Anonymous Coward (155.247.166.29) on

        > Most people outside of the USA (and many in the larger cities there) would say that driving is selfish.

        "larger cities" pretty much says it all. If you live in a densely populated area, especially one with a good public transportation infrastructure, there's no need for a car. This isn't the case for many people. Have you ever seen a map of the USA?

        Comments
        1. By Anonymous Coward (66.207.218.19) on

          Back on topic.. Another _business case_ against binary drivers:

          "People should care about open drivers because it opens up the market for support for your hardware. If the vendor decides to stop supporting a peice of hardware and you need to track a stable release of an operating system on that hardware, what happens if the vendor doesnt move with you? You're screwed. With an open driver you can find or hire a smart person to help you out."

      2. By Clay Dowling (12.37.120.99) clay@lazarusid.com on http://www.ceamus.com

        > Most people outside of the USA (and many in the larger cities there) would say that driving is selfish.

        They have clearly not tried to get to work in the U.S. outside of a major city. Driving is necessity where I live, not luxury. There are no employers within a human powered distance of my house that would allow me to actually live in that house. And I don't own a huge house. I just don't want to walk 10 miles to get to work, and I'm not going to live in a neighborhood where it isn't safe for my family to be on the streets.

        You seem keen to cast aspersions on people without any significant understanding of their circumstances. It might be that I'm driving because I'm a selfish prick. Or it could be that I need to feed my family and keep a roof over our heads.

  5. By Anonymous Coward (67.64.89.177) on

    I found this a good read regarding this issue.

  6. By Dan Langille (70.26.229.230) dan@langille.org on http://www.langille.org/

    Does anyone else care? I overheard one of the OpenBSD devs asking the BSDCan organizers for a table. Yes, he had to ask for a table, and the response wasn't apologetic: it exhibited obvious annoyance.

    I am the person that Bob Beck asked for table. On Friday, after I saw your post, I spoke with Bob Beck about this incident. He has no issues with what transpired.

    Your statement implies that a table should have been provided to him without asking and that I should have apoligized because there was no table sitting there with an OpenBSD label on it.

    In fact, a table was there. Waiting for him. Same as last year. What's wrong with that? Did you not notice the empty tables? Some groups ask for a table in advance. Some do not. So what I do is order EXTRA tables to cater for those that do not ask in advance. Would you prefer Bob have to wait for me to call the University and deliver another table? Please tell me how you think the exchange should have gone.

    I was not annoyed with Bob's request. It had been anticipated well in advance, as evidenced by the waiting table. If I was annoyed I would have said something directly to Bob about it. We're both happy with what happened. Why aren't you?

    Instead of providing a negative spin you could have just as easily praised the incident. If I didn't care, there would have been no spare tables.

    --

    Dan Langille
    BSDCan

    Comments
    1. By Deanna Phillips (70.48.229.109) deanna@sdf.lonestar.org on


      > I am the person that Bob Beck asked for table. On Friday, after I saw your post, I spoke with Bob Beck about this incident. He has no issues with what transpired.

      Hi Dan,

      I did not overhear you speaking with Bob Beck; it was someone else. And I was not hallucinating. This may explain the annoyance though; maybe you didn't know who the person was, that was asking. And who would expect you to know all 4 of the OpenBSD people by name? :)

      Comments
      1. By Anonymous Coward (70.26.229.230) on

        >
        > > I am the person that Bob Beck asked for table. On Friday, after I saw your post, I spoke with Bob Beck about this incident. He has no issues with what transpired.
        >
        > Hi Dan,
        >
        > I did not overhear you speaking with Bob Beck; it was someone else.

        Who was it? I do not recall being ask for a table by any other person. Realise that BSDCan is a pretty busy time for me, oddly enough.

        > And I was not hallucinating. This may explain the annoyance though; maybe you didn't know who the person was, that was asking. And who would expect you to know all 4 of the OpenBSD people by name? :)

        How about you tell me what happened since I clearly have no clue.

        Without straining, I can think of Bob, Reyk, and David.

        Regardless of WHO asked me for that table, please address my other points. Please go back to my previous post and answer my questions and explain what you think should have transpired.

        Comments
        1. By Deanna Phillips (70.48.229.109) deanna@sdf.lonestar.org on http://deanna.freeshell.org

          > Regardless of WHO asked me for that table, please address my other points. Please go back to my previous post and answer my questions and explain what you think should have transpired.

          Dear Dan,

          I have no interest in getting into a point-by-point debate with you about this. It is a minor issue as far as the article is concerned. I was at the conference as a user and a casual observer, and this article is pure user commentary. I heard what I heard. Let it go.

          If I jumped to any conclusions, perhaps you should make an effort to fairly represent all of the *BSD projects at next year's BSDCan. If you'd like some suggestions about how to do that (as it seems to be a problem) I suggest you start a discussion about it on bsdcan.org.

          Yours,
          Deanna

          Comments
          1. By Dan Langille (70.26.229.230) dan@langille.org on http://www.langille.org

            > > Regardless of WHO asked me for that table, please address my other points. Please go back to my previous post and answer my questions and explain what you think should have transpired.
            >
            > Dear Dan,
            >
            > I have no interest in getting into a point-by-point debate with you about this. It is a minor issue as far as the article is concerned. I was at the conference as a user and a casual observer, and this article is pure user commentary. I heard what I heard. Let it go.

            Pardon? You claim on a public forum that I behaved inappropriately and then do not wish to discuss it? How cowardly. BTW, you have still not told us exactly what you heard, and who was involved. I told you who asked me for a table, and what happened. As far as I can tell, your version of events never happened. I'm keen to be correct however.

            > If I jumped to any conclusions, perhaps you should make an effort to fairly represent all of the *BSD projects at next year's BSDCan.

            I reject the premise that no effort has been made
            .
            > If you'd like some suggestions about how to do that (as it seems to be a problem) I suggest you start a discussion about it on bsdcan.org.

            I'm always happy to hear your suggestions. What' wrong with here? If you'd like to raise it on bsdcan.org, it's always welcome.

            As far as discussion here is concerned, until you address my original points, it reflects poorly upon you and blocks any attempt on my behalf to fix any perceived slights.

            >
            > Yours,
            > Deanna
            >

            Comments
            1. By Dan Langille (70.26.229.230) dan@langille.org on http://www.langille.org/

              > I'm keen to be correct however.

              Nice. I meant "corrected", but a nice mistake.

  7. By anarcap (70.53.128.39) on

    Lets come back to the original title of the OP "Why should I care where my device drivers come from?" for a minute.

    There was an interesting article on Newsforge a week ago about the Kororaa live CD Linux distro. http://trends.newsforge.com/trends/06/05/15/1451229.shtml?tid=138&tid=130&tid=150

    Seems that they got an email from a kernel developer who wants them to stop distributing the Live CD because it contains proprietary (closed-source) nVidia and ATI driver blobs.
    http://kororaa.org/comments.php?y=06&m=05&entry=entry060512-160752

    The part that struck me as interesting is that the debate seems to be over the semantics of the GPL and whether the fact that the source for the nVidia "kernel interface layer" (i.e. the glue that talks to the blob driver) is provided is enough to make the blob GPL-friendly. That's completely missing the point. The debate should be over why there are blobs in the first place. None of the people running Kororaa get it.

    Sure there are a few people who make noise about it in the FOSS world, obviously Theo and RMS come to mind and so does Greg Kroah-Hartman http://www.kroah.com/log/2005/11/03/ , but most just don't give a damn if there are non-free and non-open blobs in their OS.

    So what's the definition of a free and open OS now? Once you start signing NDAs and including proprietary, binary, closed-source, non-open, non-free crap in your distribution, your OS might be "free as in beer" but not "free as in freedom" (don't get me wrong, I like free beer). Should an operating system even be allowed to refer to itself as "free" under these circumstances?

  8. By Clay Dowling (12.37.120.99) clay@lazarusid.com on http://www.ceamus.com

    Open source drivers are great. I certainly won't begrudge somebody their politics regarding the source of their software. But at the end of the day, I have to ask myself how open drivers help my business.

    I'm running on OpenBSD and I buy the CDs regularly. I like the quality of the os and the drivers.

    But I can't run my operations on OpenBSD alone. The market realities are that I have to deal with a lot of operating systems. I can't turn up my nose at them just because they use closed source drivers. My download logs suggest that Windows is a very popular platform for my users. I think that I'll keep on selling to them, and keep on using that operating system. After all, I too need to eat.

    Please don't run us down because we aren't manning the baracades with you. We appreciate the software that's being written; it makes an excellent platform for development and network services. But free software isn't a means until itself for most people. Families, homes, those are the reasons we write software or run servers. Free software is just a tool to help us, and it isn't the only tool available.

    Comments
    1. By anarcap (74.12.77.195) on

      Hi Clay,

      I agree with you that many of us cannot run our operations on OpenBSD alone. Listen to dlg's BSDTalk interview. He says outright that OpenBSD is not for everyone, and that there is a place for Windows and Mac users. There are a lot of things that just work better on those platforms for a whole slew of reasons, the availability of drivers being one of them.

      My work-supplied laptop has Windows on it, as there are a number of proprietary tools I need to do my job and feed the wife and kids (of course the first thing I did was shrink the Windows partition and installed OpenBSD on it but that's beside the point).

      I don't know how open drivers can help your business, but I know that they do help mine. My employer is primarily running Windows, however I've been dropping OpenBSD servers wherever they are more appropriate. I'm replacing Windows boxes for security and stability, running blobs on the BSD machines would defeat that purpose. Like you, I want the best tool for the job, and a "free" OS running non-free blobs is not it.

      That's the point of my previous post: there is no place for closed drivers in a free and open operating system. There are philosophical and practical reasons for that that have already been discussed here, on misc@, and elsewhere. You can add "it's good business sense" to that list.

      //mts

    2. By Anonymous Coward (143.166.226.17) on

      I run windows at work daily because of outlook, hardware design tools, firmware flashing devices, fpga programmers etc. That stuff simply only exists on windows and I use it every day and just like you I feed my family. There is nothing wrong or weird about that.

      What is weird is that FREE and OPEN operating systems are allowing NON-FREE code inside them. That's what is wrong with this picture. By all means use proprietary stuff whenever appropriate. It is however not appropriate in a community developed OS.

    3. By deanna (70.48.231.49) deanna@sdf.lonestar.org on

      > Families, homes, those are the reasons we write software or run
      > servers. Free software is just a tool to help us, and it isn't the
      > only tool available.

      Since when do free software programmers do their work in order to pay for their families and homes? AFAIK most do it in their spare time, to the detriment or exclusion of these things. The philosophy of the BSD licence is that you may profit from their work, and lucky you, that you may, but when they get frustrated enough with their inability to even _write_ free software that will run on whatever hardware is on the market, my guess is that they'll give it up. Lucky for you that you have other tools, I guess.

      Comments
      1. By anarcap (74.12.78.202) anarcap at gmail dot com on

        >
        > Since when do free software programmers do their work in order to pay for their families and homes? AFAIK most do it in their spare time, to the detriment or exclusion of these things.

        You're dead-on there Deanna. Over the last couple of years I've found more money in the couch than made from working on/with free software (but I guess I should be more selective in my choice of clients and employers since it is possible to make a pretty good living running only FOSS). Yet I'm still keeping at it, purely on principle.

        This goes back to the main theme of your original post, which is that most people don't have any principles when it comes to their software and hardware. That's fine for the masses, who really don't care which OS/platform/browser they run as long as they can get their email and the odd bit of pr0n. However you would expect the attendees of a free software conference to care at least a little.

  9. By Louis Bertrand (69.192.176.206) louis()bertrandtech*ca on

    One reason vendors want to keep drivers and firmware proprietary is the rampant cross-licensing of patents. These arrangements have resulted in a fully meshed network of mutual dependencies. A single company that wanted to open its IP would have to get permission from all of its licensors.

    I like the idea of open hardware. Go Taiwan!

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.]