Theo de Raadt On Firmware Activism

An anonymous reader writes "KernelTrap has an insightful interview with OpenBSD creator Theo de Raadt, discussing their recent activism to try and open up wireless chipsets. In the interview, Theo discusses what has been accomplished so far, the difficulties involved, and why such efforts are important to all free and open source operating systems."

Comments

All chipsets

[9-bits.tk]

should be open. Really, it's very narrow-minded of the chipset manufacturers to not consider the possibility of people using F/OSS operating systems instead of propietary.

Re:All chipsets

[arose]

If you can copy the money without taking any of mine be my guest.

Why not?

[isometrick]

I don't understand what companies have to lose by open-sourcing firmware or software that goes with hardware.

They make money on the hardware, not the software ... right?

Re:Why not?

[RAMMS EIN]

In fact, they lose money on having to develop the firmware. They would be better of if they released the specs of what hardware is in there, and had the F/OSS community develop the firmware and drivers.

They won't do that, though. I've heard people say that they are afraid this will make it easier to reverse-engineer there hardware, and that is somehow a Bad Thing. Personally, I think it's easier to design hardware from scratch than to look at the interface exposed by another piece of hardware and design your hardware around it - but that's just my naive view.

Re:Why not?

[nbert]

IIRC it's not up to them, because some FCC rules prevent completely OS firmware drivers.
The FCC is basically afraid that someone could modify the code in a way which would lead to a wlan device operating out of spec.
But that's just what I read some time ago...

Re:Why not?

[Ralusp]

Somebody correct me if I'm wrong, but my understanding is that FCC regulations come into play here. These devices are considered wireless radios, and they are approved by the FCC to operate on specific frequencies with specific power levels. Having open sourced firmware could allow modifications to these parameters, such that the devices no longer conform to FCC regulations. This is why several wireless chipsets have open source drivers, but none have open source firmware. Thus, the best we can do at the moment is to pursue "free" firmware, even if it is closed source.

Re:Why not?

[awehttam]

So why are there drivers for Atheros based cards?

Re:Why not?

[squiggleslash]

This honestly doesn't matter. If I lop off the antenna on my cellphone and put some kind of directional thing on it instead, chances are it's breaking FCC rules too. That doesn't mean Nokia or Motorola are breaking the law by making cellphones. On top of which, as has been demonstrated repeatedly, source code is merely a more readable version of code. You can hack binary firmware if you want, just apply a disassembler to it. You need more help understanding the code, but it can be done. So simply failing to provide source does not prevent illegal modifications.

It's up to the users to determine that their modifications are legal. There's nothing stopping companies from putting out the source code of firmware.

Re:Why not?

[Simon Lyngshede]

Would you please stop this. They are not trying to get the companies to open source their firmware, they can't. The firmware determines the frequency of the wireless chipset, open sourcing it means you can change the frequency, which is illegal in most parts of the world. The OpenBSD developers know this and is therefor NOT trying to get the source code for the firmware, Im sure they wouldn't mind having it, but they know that it is impossible.

What they are trying to do is to have the firmware release under a license, which will enable them to distribute it along with the operating system. They're aren't asking for anything but permission to ship the binary firmware. I am amazed by the number of people not getting this.

Re:Why not?

[isometrick]

I originally meant to say open-sourcing or loosening the license. Same question applies with the latter ...

Re:Why not?

[Technician]

and they are approved by the FCC to operate on specific frequencies with specific power levels.

It's like the days of CB radio. Early PLL sets had an easy to access PLL divide by N counter. Feeding it vales other than what the dial provided permitted illegal operation. Later to prevent lawsuits, the divide by N counter had a pre-programmed interface front end. The channel number was input and the divide by N was done internally. It made for more complex chips, but made out of band operation much easier.

Some WiFi chip manufactures may have the same choices. The user interface software may take the chosen channel selection and set the chip PLL to the correct divide by N ratio. The advantage is if later the FCC opens more frequencies, a simple driver update will put the chip on the new frequencies. With OSS, renagades may ditch the FCC permitted channels and find a "channel" without neithborhood interferance and not seen by the wardrivers for additional security. The chip manufacture could be held liable for enabling the out of band operation. The chip manufacture could do like the CB radio chip manufactures later did and do the divide by N table in the chip instead of in the driver software.. Now you have a chip that can become instantly obsolete if/when the FCC opens more bandwidth. The chip costs more to manufacture to boot. In a comptetive market this is a bad thing.

Re:Why not?

[DoctorPepper]

I see your point, and you're half-right. The equipment is type-accepted to operate at a certain power level on a certain frequency, but there is nothing in the FCC regs stating that the firmware has to be closed source. After all, there is really nothing (except the fear of a hefty fine and/or jail time) preventing a person from buying a type-accepted radio at Wal-Mart and modifying it to put out more power, or to transmit on a different frequency band. These radios are not "black boxes", so neither should be the firmware.

My own thoughts on as to why they are closed source leans more towards trade secrets.

Re:Why not?

[nbert]

There is nothing special about OS *drivers* for wlan cards. My Netgear MA101 runs fine on linux with the at76c503a driver, which is under the GPL. But the *firmware* is closed source.

Needless to say that the chip manufacturers put all the interesting stuff into the firmware, so the driver isn't really much worth (don't get me wrong - I'm actually happy about it. Otherwise wlan on Linux would be even more complicated)

Re:Why not?

[nels_tomlinson]

... some FCC rules prevent completely OS firmware drivers.

Let's assume you're right. Your point would be irrelevant, since they aren't being asked to open-source, but to make their closed source firmware freely distributable.

And, I think you're wrong.

Re:Why not?

[Anonymous Coward]

It made for more complex chips, but made out of band operation much easier.

s/easier/harder/

Re:Why not?

[JohnQPublic]

This honestly doesn't matter. If I lop off the antenna on my cellphone and put some kind of directional thing on it instead, chances are it's breaking FCC rules too. That doesn't mean Nokia or Motorola are breaking the law by making cellphones.

If Nokia or Motorola make it too easy for you to replace the antenna, then they are indeed breaking the law, at least in the USA. Radio transmitters here are heavily regulated, and both use and sale of unapproved transmitters is illegal (with certain exceptions - e.g. hams can build their own stuff).

On top of which, as has been demonstrated repeatedly, source code is merely a more readable version of code.

Don't let the FSF hear you say that :-)

It's up to the users to determine that their modifications are legal.

Nope. The US "type acceptance" practice does not permit unapproved modification of key parameters, and holds all parties responsible at various levels, including the manufacturer.

Re:Why not?

[ajlitt]

Firmware for many modern devices is encumbered by license restrictions beyond the hardware vendor's control. Look at the MP3 player. A manufacturer selling a player that can do MP3 and WMA must not only pay the Fraunhofer tax, but also pay Microsoft for the WMA codec. If the firmware is designed around an RTOS, then there's that too. While most copies of firmware are useless on hardware other than the vendor's, the possibility that a freely redistributable binary will provoke the ire of the owners of the software components tends to freak out the vendor's legal department.

Re:Why not?

[Anonymous Coward]

loosening the license.

This is Slashdot. If you write stuff like that people will think you mean the license can't be found.

Re:Why not?

[dmiller]

I think that you have been lied to. see here (and the discussion that ensues).

Re:Why not?

[Anonymous Coward]

I have to tell you one thing I've learned over the years: Just because a device's firmware could be upgraded to allow more functionality, typically manufacturers don't provide the upgrade when the time comes. They'd far rather sell you a new widget than let you keep using your old one with new functionality "for free". I wrote off the concept of getting a firmware upgrade for all but the worst offending bugs years ago.

Manufacturers really ought to get a clue...

Large Concerned User Base

[lukewarmfusion]

"The idea being to let the vendors know that there is a large concerned user base that is going to decide how they will spend their money based on the vendor's willingness to work with open source software."

Not only for themselves, but in how they make recommendations for their companies... TI (and other vendors) may not care about that one extra unit sold, but they'll certainly be looking at the corporations that might purchase thousands of units.

Not so good for linux...?

[julesh]

Isn't there a problem here, that while a closed source redistributable binary is fine for use with *BSD, it cannot be linked into the Linux kernel because the GPL requires you to distribute source for anything that is so linked...?

Re:Not so good for linux...?

[Anonymous Coward]

You'd better get everyone to stop using their NVidia cards then, can't run that closed-source driver.

No, there is not a problem with running a driver that is a LKM, and thus loadable on the fly (not compiled into the OS). And we aren't actually talking about anything thats really *part* of the OS anyways, you're talking about basically a binary image file (in essence) of the code used internal to the wireless card, that you have to send to the card before it will work. I'm not sure that fits into the "part of the OS" mold, although thats an interesting legal line to deal with that didn't exist years ago when the GPL started.

Re:Not so good for linux...?

[Anonymous Coward]

You got the wrong end. The topic is on firmware which is a binary blob loaded into hardware. So this is no x86 code at all (ok probably in rare cases). This blob is NOT executed by the host machine instead the blob is loaded into hardware and then the hardware starts working and can be used in the usual way by the driver.

Re:Not so good for linux...?

[RAMMS EIN]

It's actually much more the other way around, but I see why you're confused.

The BSD license permits the use of the covered software in closed-source projects (it just mandates an acknowledgement). The GPL does not permit this - if you link with a GPLed work, the result must be GPLed, too (at least if you redistribute it).

Now, for the kernels, the situation is a quite different. The BSDs are very puritan in that they only allow BSD licensed code in the kernel, whereas Linux is more pragmatic in that it allows a mixture of GPL, MPL, closed-source, and basically whatever you come up with.

This is not to say that it wouldn't be legal for you to link a proprietary module to a BSD kernel and distribute it (provided that the proprietary part can be redistributed). It's just that the BSDs won't do this in their official distros. Many Linux distros also won't ship non-free kernel modules. And they couldn't legally compile a GPL incompatible module into the kernel and distribute the result.

Re:Not so good for linux...?

[julesh]

You'd better get everyone to stop using their NVidia cards then, can't run that closed-source driver.

Well, why isn't it incorporated into the kernel then? Can't do that. You have to download it when you install your system rather than get it distributed alongside it... because the result of compiling something that mixes closed source and open source code is not actually distributable under the GPL.

No, there is not a problem with running a driver that is a LKM, and thus loadable on the fly (not compiled into the OS).

Do you have a good source for this? Preferably from a lawyer? My reading of the GPL is that this is not allowable. NVidia (and others) get around this by having a small open source driver that talks to both the kernel and their closed source driver, meaning that there is no direct interface between them. Even this is legally dubious, although I doubt they'll have any problems, because they don't distribute a binary version of the closed source section.

And we aren't actually talking about anything thats really *part* of the OS anyways, you're talking about basically a binary image file (in essence) of the code used internal to the wireless card, that you have to send to the card before it will work.

Being part of the OS is irrelevant. The question is, does it link directly to the kernel? And whether the linking is static or dynamic is irrelevant, according to the GPL. I can see one allowable way of achieving this: have it as a data file that must be installed somewhere. This seems OK, because then it won't be linked, it will be loaded at run time.

Re:Not so good for linux...?

[julesh]

I don't think that matters. If it is distributed as part of a kernel module, that module will be linked to the kernel when it is compiled. If you then distribute that binary, according to the GPL, you have to distribute or offer to distribute the source (that is the "preferred form for making modifications") to _everything_ that goes into that module.

Re:Not so good for linux...?

[Anonymous Coward]

You'd better get everyone to stop using their NVidia cards then, can't run that closed-source driver.

Or just use the X.org driver. Hell Operating Systems you've never heard of have nVidia drivers.

Re:Not so good for linux...?

[LiENUS]

Do you have a good source for this? Preferably from a lawyer? My reading of the GPL is that this is not allowable. NVidia (and others) get around this by having a small open source driver that talks to both the kernel and their closed source driver, meaning that there is no direct interface between them. Even this is legally dubious, although I doubt they'll have any problems, because they don't distribute a binary version of the closed source section.

Actually, the word from linus is closed source drivers are perfectly acceptable, the reason nvidia has the open source glue is so that they dont have to distribute 5000 binary drivers for each and every version of the kernel, they simply redistribute the open source glue to translate the current api to their binary driver's api

Re:Not so good for linux...?

[Anonymous Coward]

The GPL talks about linking. You can't link a binary blob from a different architecture. So you just need to distribute the binary blob. So redistribution of the blob is the problem and not linking. If you accept the blob as beeing DFSG free is another matter, but this has nothing to do with the GPL or BSD license.

Re:Not so good for linux...?

[thomasj]

Isn't there a problem here, that while a closed source redistributable binary is fine for use with *BSD, it cannot be linked into the Linux kernel because the GPL requires you to distribute source for anything that is so linked...?

This is not a part that is supposed to be linked into the kernel proper (at least not as executable code), but firmware for a different chip on the board of the WLAN NIC.

Look at it this way: You used to have the firmware on a chip as closed source and you didn't even blink. Now it is loaded by the driver... What changed here?

Re:Not so good for linux...?

[julesh]

the word from linus is closed source drivers are perfectly acceptable

Then either he needs to put an exception into the licence, or provide a way of loading them that doesn't link them directly to the kernel, because it is in contravention of the GPL to do so, whatever Linus says about the matter.

Re:Not so good for linux...?

[julesh]

What changed is that it is incorporated into that driver. If it is stored as a separate file and loaded from the filesystem after it is mounted, it seems reasonable (it would come under the 'mere aggregation' clause of the GPL, as mentioned by another poster), but my understanding is that this is not how the Linux kernel currently works.

The firmware would have to be incorporated into the kernel module, which means that the firmware's source code would have to be distributed.

The offending portion of the GPL is this:

For an executable work, complete source code means all the source code for all modules it contains

The firmware meets just about any definition you care to apply of "module", and is certainly "contain[ed]" within the kernel module.

Re:Not so good for linux...?

[tedu]

the firmware does not need to be linked into the driver. you can store it in a separate file. this is how the centrino driver works. it may be convenient to link it in, so it's all one piece, but that is not a requirement for it to work. a gpl driver can read a file off the disk and stuff it into the card just the same as any other driver.

since the linux kernel is not executing the firmware either, in the context of that program, the firmware is not software. it's only binary data that cannot be interpreted. the firmware doesn't actually link against anything in the kernel, nor does it need the linux kernel to work.

Re:Not so good for linux...?

[swillden]

Now, for the kernels, the situation is a quite different. The BSDs are very puritan in that they only allow BSD licensed code in the kernel, whereas Linux is more pragmatic in that it allows a mixture of GPL, MPL, closed-source, and basically whatever you come up with.

Whaaa??

Anything distributed as part of the Linux kernel is either GPL or constitutes a violation of someone's copyright, and possibly other rights as well.

Worth reading TFA

[Len Budney]

...if only to see a pic of the guy who causes such a ruckus.

Otherwise, the interview is extremely tame. He alludes to the "corporate ways of Linux vendors", but doesn't give vent to any interesting rants.

De Raadt raises one interesting question, though, when he says, "in other cases we have had to resort to activism. An example of this was Qlogic...for a few years we did ship this code without being aware of the issue. But after a few mails to Art at Qlogic, and a threat to remove their code from our upcoming release, they decided to let us include the firmware in our operating systems."

The question is: how is this "activism"? He states that they used Qlogic SCSI firmware inadvertantly, and when asked to stop, threatened to comply with the request. That's "activism"?

Re:Worth reading TFA

[compass46]

http://www.benzedrine.cx/pf2k4/DSC01359-medium.jpg

Re:Worth reading TFA

[I confirm I'm not a]

He states that they used Qlogic SCSI firmware inadvertantly, and when asked to stop, threatened to comply with the request. That's "activism"?

I didn't read it like that: I interpreted it as "Theo et al thought the firmware was free enough to include in OpenBSD. When they became aware that it wasn't they contacted Qlogic, who eventually conceded and now the firmware is free enough."

In other words, the threat to stop was made by OpenBSD, not Qlogic. I think!

Re:Worth reading TFA

[archen]

...if only to see a pic of the guy who causes such a ruckus.

Yeah, for some reason I had this picture of him with "Einstein" hair and his eyes WAY to wide like he had just consumed 15 cups of coffee. He actually looks like a normal person...

Re:Worth reading TFA

[compass46]

The guy asked for a picture so I provided one. Why was that modded down?

Re:100% Free?

[RAMMS EIN]

The legal status of the official ISOs and artwork do not change the fact that the OS itself is 100% free.

You can make your own ISOs and distribute them, do a network install (which, last time I did it, required just one floppy image and was very easy).

It's all similar to Red Hat not allowing you to call copies of the official CD Red Hat, or vendors not releasing the latest version of their software under a free license. It doesn't make other distributions of the same software non-free.

To keep average Slashdot reader on track...

[codguy]

Since most Slashdot readers will not RTFA before commenting, let me clearly point out that this is *not* about wanting the companies involved to open up their source code for use by OSS. It is simply requesting that the existing firmware be freely distributable by OSS without onerous conditions.

For A.D.D. and no-RTFA Slashdot readers/commenters, let me repeat that this is simply about being able to freely distribute an already compiled (e.g. binary) version of the firmware. OpenBSD is *not* asking for the source code.

Loosely speaking, the firmware in question is already freely available--you just go to the website and download it. But that doesn't help when you are loading a distro. If you *only* have a wireless connection, this is a chicken-or-the-egg problem. You can't go to the website to download the firmware because your wireless NIC won't work without the firmware. Yeah, there are many possible workarounds, but by simply allowing the firmware to be freely distributable without onerous licensing terms, the wireless NIC can work right off the bat.

Unless your foresight is amazingly shallow, or simply a Theo-hater, note that this will benefit *all* OSS, and not just OpenBSD.

--codguy

Re:To keep average Slashdot reader on track...

[jhdevos]

... note that this will benefit *all* OSS, and not just OpenBSD.

Well, except for Debian, which does not distribute non-free software at all. It might distribute a script to easily download and install te binary firmware, but that is allready possible today.

Jan

Re:To keep average Slashdot reader on track...

[renoX]

> It might distribute a script to easily download and install the binary firmware, but that is allready possible today.

"possible" in theory yes, but not very easy to do if this firmware is used to allow Internet connection.. Chicken/Egg problem..

Re:To keep average Slashdot reader on track...

[tedu]

except for the fact that you can't download TI's firmware. so your script won't work too well.

I hope that Linux&FreeBSD users will join

[renoX]

Whatever one may think on Theo, I think that he is right here: firmware which cannot be redistributed by distribution are a *pain* for the users.

I hope that Linux&FreeBSD users will join this movement because the more users requests hardware-makers to allow redistribution of firmware, the better!

Also, I think that this movement should not be restricted to wireless HW, I have a speedtouch ADSL modem where there is a similar situation: firmware may not be redistributed.

This is very annoying when you want to install a distribution.. I think that Mandrake managed to get the rights to redistribute this firmware, but they shouldn't be the only one to have this right..

Re:I hope that Linux&FreeBSD users will join

[9-bits.tk]

Getting a SpeedTouch adsl modem working on mandrake is an absolute BUGGER to set up.

It never bloody connects properly.

Re:100% Free?

[julesh]

Why do you want the Official CD-ROM, if you're not going to pay for it? Go download an unofficial one, and stop caring so much.

Re:100% Free?

[myc_lykaon]

Wait a minute, this is coming from an organization that forbids people from sharing ISOs of their "official" CD set?

Isn't that what we are all saying is good about OSS/Free as in speech software? Value added development. We all start from the same base and if you want to make money/kudos/whatever then add value by producing a bundle/ISO that targets a particular market.

They write the software and bundle it up in a way they and we percieve as 'adding value'. You can still d/l the whole kit yourself and value add for re-distribution and give that away free. Just don't reduce the value of anothers work by giving away their hard work if they ask you not to.

Re:100% Free?

[nharmon]

I appreciate that the OS as available through FTP is 100% free, but don't quite agree with you on the comparison to Red Hat.

Instead, I would say that what OpenBSD does would be more like Linus saying that Redhat was the Official CD of Linux.

Re:100% Free?

[bheer]

Nothing stops you from downloading the source, building it on your own, stamping the binaries on CDROM and selling it.

Give them a break, a lot of OBSD fans buy the CD set just to show some support. And remember, at least as far as commercial use is concerned, it *is* free-er than GNU/Linux (not getting into whether that's desirable or not).

Re:100% Free?

[raddan]

Gimme a break. You can download the entire OS via FTP and make your own ISO. Even the OpenBSD installer doesn't care about the files, so you're not dependent on the 'official' ISO at all.

The people who buy OpenBSD CDs don't do it because they're locked in or forced to in any way. We do it because we want to support a high-quality operating system. Considering that OpenBSD has replaced several costly Windows boxes where I work, the $40 for a CD is inconsequential.

And, lest you forget, OpenBSD has a free-er license than Linux (don't get me wrong, I love and use Linux every day). OpenBSD's goal is getting high-quality software out there, not to free the world. You seem to be forgetting Theo's interview on Slashdot:

The licence on our code is pretty clear. We want vendors to use our code. We want commercial operating systems to ship with OpenSSH. Not shipping with an SSH varient causes great grief, and it is time that ends.

Same goes for OpenBSD. We would prefer if companies building commercial network appliances used OpenBSD, rather than writing their own operating systems. Typically, these companies are very comfortable with solving the problems within their application space. Yet, there is a history of these companies writing their own cruddy operating systems, and at the same time writing worse applications.

It would be better if routers, firewalls, telephone switches, fileservers, and whatever else used reliable components, designed by people who care.

So go ahead, use any parts of OpenBSD as parts of commercial systems.

Re:100% Free?

[nharmon]

If I ever do want the Official CD-ROM, I will pay for it.

But don't call the thing 100% free when it isn't.

Re:100% Free?

[Anonymous Coward]

``Instead, I would say that what OpenBSD does would be more like Linus saying that Redhat was the Official CD of Linux.''

That comparison is a very flawed one. OpenBSD is a full OS; Linux is only a kernel. The RedHat CDs include much more than Linux alone, and are composed outside Linus's sphere of influence. Him saying that RedHat is the official CD of Linux would be mostly meaningless.

As you asked - you're utterly wrong.

[Anonymous Coward]

The firmware that OpenBSD wants is the binary code that gets uploaded into the card and run on the card's onboard CPU.

Thet don't want to obtain that firmware's source. They want distribution rights to include the unchanged firmware in their open-source drivers. That firmware is already "free" to download from the web and extract, or extract from the Windows driver on the CD bundled with the card, but OpenBSD wants to cut to the chase and just have drivers that work first time from their install CD.

Re:As you asked - you're utterly wrong.

[Ralusp]

Actually, that's exactly what I said. They are pursuing "free" firmware, not open-source firmware. The remainder of my comment was in response to the question posed by its parent.

Re:100% Free?

[squiggleslash]

OpenBSD itself is 100% free. A particular CD layout of it isn't. I don't think there's a problem with that.

Re:100% Free?

[julesh]

I see your point, but think your reasoning is flawed. This CD is only really 'official' in that procedes from its sale help fund the OS development process. Theo doesn't say that it's any better than any of the other ISOs out there that you can download -- he just says that if you want to help OpenBSD development, that's the one you should buy.

Theo? Activism?

[PornMaster]

Whoda thunk it?

Anyway...

[hummassa]

Firmware is NOT linked with the linux kernel, so the GPL "mere aggregation" clause applies. Obviously, this is MHO, and this has already generated a LOT of debian-legal discussion.

YMMV HTH :-)

Theo de Raadt at its best?

[a_hofmann]

From a story linked in the article:

[Theo de Raadt] said he found it sad that the Linux crowd did not help in the activism at all. "(They) always seem to talk about freedom but are not helping in this activism. It's basically BSD people doing it. That is curious. For instance, do you think Linus (Torvalds) would send a mail to TI? No, I would bet money that he did not. Yet he is aware of what is going on. That's very odd to me."

I honor the efforts of the OpenBSD projects and Theo de Raadt, but this is childish behaviour. I don't think one could deny that Linus does propagate the idea of free software. He created the currently most distributed free operating system, did he?

Speaking like that of the Linux crowd at whole is ignorant. I'd wager that most closed source hardware components where opened by previous Linux efforts, the project that actually initialized the wider acceptance of UNIX like operating systems we enjoy today by (unfortunately still limited) vendor support.

If Theo didn't split camps he might actually find cooperation with Linux developers.

Re:Theo de Raadt at its best?

[agent dero]

"He created the currently most distributed free operating system, did he?"

No, he created a kernel, he created a GPL licensed, monolithic and modular kernel.
Linux is _just_ a kernel

I'll have to stick by Theo on this one, there is a lot of whining about Nvidia binary drivers for their video cards, but that seems to be all it is, whining.

Theo is doing something about it, whether or not you agree with his cause, he is _doing_ something at least.

Re:Theo de Raadt at its best?

[10Ghz]

I'll have to stick by Theo on this one, there is a lot of whining about Nvidia binary drivers for their video cards, but that seems to be all it is, whining.

uh, there are open-source drivers available as well. And what do you suggest that Linux-folks should do? Deliberatly sabotage the kernel in such way that closed-source NVIDIA-drivers don't work?

Disclaimer: I think Theo is an arrogant asshole.

Re:Theo de Raadt at its best?

[Anonymous Coward]

I honor the efforts of the OpenBSD projects and Theo de Raadt, but this is childish behaviour.

Theo is a crackerjack software engineer no doubt, but he's also a flaming asshole. Use his OS and pretty much let anything he says roll like water off of a duck's back.

If Theo didn't split camps he might actually find cooperation with Linux developers.

Yeah, the "splitting of camps" seems to be a common theme with him ala NetBSD, though THEY chopped HIM (as opposed to him leaving) since he was seen as a liability. Failure to keep his mouth under control also cost him the Darpa bucks, which is sad because OpenBSD is so fearsomely secure.

Re:Theo de Raadt at its best?

[agent dero]

"Disclaimer: I think Theo is an arrogant asshole."

Correction, I've worked on a couple OpenBSD forks that can confirm, that yes, Theo is an arrogant asshole.

He does however, stick to what he believes in, a very respectable trait ;)

Re:Theo de Raadt at its best?

[justins]

I honor the efforts of the OpenBSD projects and Theo de Raadt, but this is childish behaviour. I don't think one could deny that Linus does propagate the idea of free software. He created the currently most distributed free operating system, did he?

Of course, de Raadt did not claim that Linus has not done enough to "propagate the idea of free software." He was talking about the firmware fight. Keep knocking over those strawmen!

Speaking like that of the Linux crowd at whole is ignorant.

I guess it's a good thing he didn't do that, then.

Re:Theo de Raadt at its best?

[Mr.Ned]

"He [Linus] created the currently most distributed free operating system, did he?"

No, he didn't. He wrote a large part of the Linux kernel, which is combined with GNU software to produce an operating system. This distinction is what gets RMS hopping mad.

The kernel is only one part of (Open|Free|Net)BSD. The *BSDs are operating systems because they include userland tools.

Re:Theo de Raadt at its best?

[tedu]

if you don't like what he said about linux users, i recommend you prove him wrong by actually doing something about it. say, like trying to contact some people at TI to free up the firmware, instead of whining about mean theo is.

Re:Theo de Raadt at its best?

[a_hofmann]

No, he created a kernel, he created a GPL licensed, monolithic and modular kernel.
Linux is _just_ a kernel

By definition of the word Operating system, you're right. Still, though free kernels and user land utilities existed well before Linux, it was his project that kicked off a lot of development around open source operating systems.

Without Linus there would be no Linux, and that's the simple truth.

As FreeBSD user I am well aware of the diversity of ongoing efforts delivering us the final product of a free operating system. Still all the lot would not be possible without the people fighting at the front.

I'll have to stick by Theo on this one, there is a lot of whining about Nvidia binary drivers for their video cards, but that seems to be all it is, whining.

That's just a statement of your personal opinion, not being backed up at all. Just like Theo's. Insulting on the basis of such arguments is unprofessional and does no one any good. Why not try to get together in a proper forum that actually allows bidirectional communication?

Re:Theo de Raadt at its best?

[Anonymous Coward]

Well, now that you know we (all of Open Source) need help on getting TI to change their firmware license, what are you and the Linux crowd going to do about it? Are you going to be childish and ignore the whole thing just because Theo made a true and "mean" remark against you?

Re:Theo de Raadt at its best?

[mirabilos]

I actually started one of these, and I think
Theo is arrogant, but not totally of an asshole.

Especially the recent interview showed me some
other sides of Theo reminding me of myself:

http://smh.com.au/articles/2004/10/07/1097089476 28 7.html?oneclick=true
(if it wants registration, use a better browser,
such as lynx, then you'll see it directly)

Re:100% Free?

[Anonymous Coward]

Nonsense. You can get OpenBSD the OS without using the "official" (ie blessed by Theo) CD layout. The CD layout is not part of the operating system.

Re:100% Free?

[nels_tomlinson]

... don't call it 100% free. Call it, 97.1% free.

Ok, it's 100% Libre, and 97.1% free.

There is a reason to distinguish between the two, and you've illustrated it beautifully. As others have pointed out, the ISO isn't the distribution.

There are reasons to complain about OpenBSD, but they don't include its ``Libre-ness'' or its quality.

Open Cores?

[cpghost]

[All chipsets] should be open. Really, it's very narrow-minded of the chipset manufacturers to not consider the possibility of people using F/OSS operating systems instead of propietary.

All chipsets should be open. Really, it's very narrow-minded of the chipset manufacturers to not consider the possibility of dust or humidity settling or condensing on the open raw chip. Plastic cases are there for some reason, ya know?

Now Open Cores would be great! But as long as we don't have a home chip manufacturing unit (say, like a printer or so), we won't be able to use the source code anyway (though some of us could find out about hidden functionality etc...).

What we do need now are open specifications, both electrical and functional: What do you need to write to Pins 3-29 and what does the result on Pins 30-35 mean? This kind of stuff ought to be open!

Re:Open Cores?

[runderwo]

But as long as we don't have a home chip manufacturing unit

ever heard of a FPGA? Or even a CPLD. These things are not expensive at all.

Re:Open Cores?

[bigberk]

Now Open Cores would be great! But as long as we don't have a home chip manufacturing unit

Oh come on, you're acting like you don't know about FPGAs (quite cheap from many vendors, my favourite is xilinx). These chips can be totally reprogrammed and many of them now come with soft cores. Awfully configurable. These things are in the process of a revolution in firmware/embedded systems for flexibility and updating. Forget about uploading new data to flash memory; you can easily, cheaply, and reliably update the whole damn processor.

Look at it this way

[Anonymous Coward]

The OS is free. The official CDs aren't. They make their money on the CDs, so they ask people not to distribute ISOs so they can sell their official CDs. But it's not illegal to distribute your custom ISO.
Apparently their web site content isn't free either (cf. MicroBSD). But supposedly the entirety of the OS is (though there may be some non-free stuff in there as well, I'm not sure exactly what the deal was with MicroBSD).

Re:Look at it this way

[Stormgren]

MicroBSD ripped off major chunks of the site, docs and disto by stripping the copyright notices and doing other things which violated the BSD license.

They fucked up by essentially doing a s/OpenBSD/MicroBSD/

Check out this Wikipedia link for somewhat more info.

great

[torrents]

to see people out there really pushing hardware co's to open up and increase their potential client base... you think this would be something they would undertake on their own...

closed-source drivers not acceptable to Linus

[Anonymous Coward]

http://kerneltrap.org/node/view/1735
The only exception is if the driver was not developped specifically for Linux (ie somebody ported it) and then it's not considered a derived work. Otherwise any kernel code falls under GPL.
People who don't like that should use BSD instead. It has no such requirements.

Re:closed-source drivers not acceptable to Linus

[julesh]

Thank you. I thought the OP was wrong, but didn't have a source for that.

I'd also say that even if a kernel module was ported from another OS, any code touched by the porting is likely to become a derivitive work of the kernel.

Re:closed-source drivers not acceptable to Linus

[LiENUS]

And if you'd read that link you'd see i'm not wrong. Yes closed source modules that do all kinds of weird things with kernel internals are not ok, however porting your binary driver from windows to linux is acceptable... and what is it that we are discussing here? using binary firmware from windows (not even ported but directly from) under linux, not even touching the kernel internals. Those emails are primarily dancing around the issues behind filesystem drivers and the like.

Theo and the other Gods of BSD ;-)

[petrus4]

I haven't read all that much about de Raadt...a few interviews mostly, I'll confess. What I have read though I've felt pretty positive about, myself. Yes, he's rather prickly/sensitive, and deeply strange, but those are two characteristics that are commonly associated with people who are abnormally intelligent.

It's true from most of what I've read that the BSD dev crew *do* seem to see themselves as one of the last holdouts of human intelligence. The thing is though, the evidence would tend to suggest that they're almost certainly correct in thinking that. ;-)

Have a look through bsd.ports.mk and its associated files (as one example) at some point if you don't believe me...I found myself being reminded of Wayne and Garth's reaction to Steven Tyler when I did. ("We're not worthy, we're not worthy!" etc)

It's made me think that the old crack about LSD and BSD is true...though not from the point of view of the BSD developers taking it themselves, but from the point of view of them being sufficiently intelligent that the rest of us would need to consume LSD in order to keep up with them. ;-)

Let's push freedom aside for convenience.

[jbn-o]

What other proprietary software would also "benefit" "open source" operating systems? Since this is all being done in the name of convenience for the user, why not admit it and ship more software that the user might find convenient so the OpenBSD system is even more useful out of the box?

As much as the OpenBSD team cares about making software that commercial developers can build on (even to make proprietary derivatives), these proprietary firmware packages will not be available under such terms. There will be vastly different licenses for portions of OpenBSD--one license for each of the proprietary binary firmware files, and one license for the rest of the OS (the new BSD license).

When such firmware is built into the Linux kernel, that variant of the kernel becomes non-redistributable because one can't meet the terms of the GNU GPL (complete corresponding source code cannot be supplied). The OpenBSD system will remain redistributable (except for the oddity concerning distributing verbatim copies of the OpenBSD discs). And "open source" will again demonstrate its ability to trade away the practical idealism that built the free software community in exchange for convenience.

Re:Let's push freedom aside for convenience.

[tedu]

the firmware does not need to be linked into the linux kernel. you can write a driver that reads it off disk. at least if it's free, it can be put onto the same floppy or cd image as the kernel and you can use it to install.

Re:Let's push freedom aside for convenience.

[ComputerSlicer23]

When such firmware is built into the Linux kernel, that variant of the kernel becomes non-redistributable because one can't meet the terms of the GNU GPL

Here you claimed to understand that this wasn't a matter of being a part of the driver.

You are being silly. Go read up on the definition of "derived work" and "linking". Then read the GPL very closely. You'll notice that the firmware of a PCI card doesn't even execute on the same CPU architecture, let alone in the same address space (the rule of thumb amout if you are "linking" or not). When you get done figuring that out, realize that because the firmware was independently developed outside of the Linux kernel, and it works completely independently of the Linux kernel, it's not a derived work. Thus it's pretty much in the free and clear of all GPL issues. The actual binary bits in the kernel you are free to change to your hearts content with full GPL rights.

That's why firmware can be inside of the Linux kernel now. Some people want to move it outside of the Linux kernel (for both technical and political reasons). The technical being, that you can upgrade the firmware without re-compiling your kernel, and it shrinks the size of the kernel to not have it statically compiled in. The political is that so dolts like you don't say "That's a GPL violation". It isn't. The firmware isn't a derived work, and it isn't a linking to a GPL'ed piece of code. It's just data as far as the kernel is concerned. Just like the C code that passes thru a GCC is just data. Yes technically speaking it is source code, but it's just data. It's just like saying, well in order to initialize this card, you have to write a "0x80" to this port to get it configured correctly. In this case, instead of a single byte, it could be a 10-64k chunk of bytes. It's the same thing. They are going to make it blazingly obvious by moving it outside of the source tree so it acts just like the GCC code does in every single way. Then we can finally be finished with this argument. That's why the 2.6 kernel is building all of the infrastructure so that firmware can be loaded from user space. Then as long as the vendors say the firmware can be distributed for free, it's all good (which is what Theo is attempting to make happen).

Even if the OpenBSD and Linux people got the source to the firmware GPL'ed, there's no way in hell they'd ship source you had to compile. You'd still get a binary distributed to you. That would require you to have a development tool chain for whatever language (compiler for the language, assembler for the target architecture, and possibly a linker for the object format). Some or all of which literally might not exist outside of the company. It's not like Adaptec is using an x86 OBJ from C source for writting it's firmware. They might, but I wouldn't be shocked to find out they use a PIC with a propriatry C compiler from an embedded vendor. That would be a dependency the Linux kernel folks wouldn't allow for building a kernel.

Kirby

Be Informed When Purchasing Hardware

[Anonymous Coward]

These sorts of issues illustrate why it's critical to consult the (rather incomplete) supported hardware lists. In addition, because the hardware lists can be vague and the vendors don't always provide complete information on the product, talk to someone who runs a similar setup. In addition, figure out what driver is used for your device and read the source code; oftentimes there will be comments about specific cards. Finally, see how long development has been taking place for a particular driver; the driver that's been around the longest may be the safest bet.

With regards to wireless cards, you shouldn't have any problems with Cisco or Motorola gear on OpenBSD.

Re:100% Free?

[Anonymous Coward]

Dude, that ISO's copyrighted by Theo. If you want an ISO, then quit complaining and make one; or download one from the 'net.

Re:100% Free?

[hackwrench]

Being encumbered with the legal status of the official ISOs and artwork is not 100% free. It's like saying that the governments of thw world are 100% free because some distributions of it are free.

Wow.

[pclminion]

Of all the things I did NOT expect in this life, I think the least likely would be to ever see the term "firmware activism."

Man, that's like Suffrage for Spoons, or something.

I've thought about switching back to OpenBSD

[TheHonestTruth]

Ever since my mandrake update borked my Netgear MA311v2 wireless nic. I have exactly the chicken egg problem: can't get the next update w/out the connectivity. I even tried 10.1 Community with no luck. OpenBSD always had great NIC support, but I got tired of dealing with chroot on my webserver (different machine) when I wanted to add gallery and other dynamic page processing stuff, so I switched both to Mandrake. Now I'm considering putting OpenBSD on my non-critical desktop to get the wirelss NIC support and keeping Mandrake on my webserver. How's that for backwards??

-truth

Activism is compliance with proprietors?

[jbn-o]

That's "activism"?

It's catering to what the proprietors want you to do--become more dependent on their products and work under their licenses. Ultimately, the copyright holder holds the power. In the Qlogic example, it was not any "threat" that gave Qlogic incentive to do what it did. de Raadt is trying to make this seem like OpenBSD held the power in this relationship. Qlogic realized that it had the opportunity to avoid a minor PR hassle and simultaneously not forgo sales of its hardware when it allowed others (including those ostensibly dedicated to "open"ness) become their buttress by distributing their copyrighted proprietary code.

What the OpenBSD team is doing is in line with the idea of gaining favorable cachet of "open source" to chase after technical convenience at the expense of a user's freedom to inspect, modify, and alter computer software to suit their needs. Proprietors know that users will give up their freedom if they never learn to value software freedom, and if they are provided with a frictionless path to doing what you want them to do. OpenBSD is working hard to build that frictionless path to becoming yet another proprietary software distributor.

Re:Activism is compliance with proprietors?

[Flower]

No. It's activism. A lot of people communicated their concerns to those companies and a majority of them have had their *decision makers* open a dialog with Theo and we are now getting results. Does it go as far as you personally want? Obviously not. But at least it's progress instead of the "suck it up and accept being inconvenienced" that you espouse. Even RMS compromises. Look at how parts of Ogg Vorbis got relicensed to promote adoption of the standard.

And just out of curiosity what totally free hardware are you using to post to /.? Video Card? Probably not. That's still talk. NIC? I ain't seeing much maybe I should make that an Ask /. question. Soundcard? What about your printer? Just how much are you sacrificing to be truely free? For the OBSD community it was some time to write a thoughtful letter to get the ball rolling.
Today we now have some easily distributable firmware for desired hardware. We'll see what tomorrow brings.

[NT] and firmware loader moots this point anyway..

[Luyseyal]

Re:100% Free?

[IBeatUpNerds]

This comment is 100% stupid.

Creating an ISO image from an OpenBSD FTP download is pretty simple. With one simple FTP session and a subsequent mkisofs command you have a bootable ISO image with a full operating system that installs in minutes.

I have even written a small python script that will download a package and all of its dependencies so you can include it in your ISO image for easy access. Pretty simple to do if you know how the packages are laid out internally.

Sure, there is no full ISO to download, but all the pieces are there. Nobody forces you to buy anything. Feel free to create an ISO image and share it yourself on a private mirror or bittorrent. I'm sure those of us who want OpenBSD on a CD will either buy the official set to support the project or create their own.

Re:100% Free?

[Anonymous Coward]

Nothing legal is stopping you, but doing so will put you on the enemy list in the small OBSD community.

suggested firmware activism: dvd burning PS2 games

[ClioCJS]

Can someone please hack the firmware of a dvd burner so that it can create DVDs that will play in a PS2 (homebrew games, copied games, whatever) ??

I mean, most of us have mod chips, but will someobody please think of the children?!?!?

Re:100% Free?

[Anonymous Coward]

Being encumbered with the legal status of the official ISOs and artwork is not 100% free. It's like saying that the governments of thw world are 100% free because some distributions of it are free.

You can get an official OpenBSD CD, mount it, copy everything off it and then burn a bootable CD from those files. No problem. Staying within the stipulated copyright placed on the official CD's.

What you can't legally do, is make a bit perfect image copy of the official CD's to pass around. The layout of the ISO filesystem on the CD has a copyright stipulated on them which prevents copying. Not the files within that ISO filesystem (which constitute the OS install files).

I don't see anything wrong with this. The files are what make up the OS and they are very free.

Hell, why download a ~600MB ISO image, when you can just download about 150MB for i386, then burn it to a CDR with cdrom36.fs as the bootable image? Or network install from floppy36.fs?

The official CD's are pretty neat btw, you can typically boot off any given official CD, onto more than one different architectures. You get cool stickers and insert, cool CDDA clean song and all up what I would consider a collectors item. Plus, you have contributed to the next version of this FANTASTIC OS.

Re:100% Free?

[Anonymous Coward]

Nothing stops you from downloading the source, building it on your own, stamping the binaries on CDROM and selling it.

Nothing stops you from downloading the binaries, stamping the binaries on CDROM and selling it.

It is the ISO filesystem image that has copyright stipulated, not the files within that ISO filesystem.

The OS is as free as the BSD licence that applies to it.

Re:100% Free?

[Anonymous Coward]

Nothing legal is stopping you, but doing so will put you on the enemy list in the small OBSD community.

Right. That's why Daniel Hartmeier (OpenBSD developer of pf fame), posted torrents to undeadly and slashdot of i386 binaries.

The fact is, they DO want the OS to be free and as such, make BINARIES available for free via ftp and there are tons of mirrors.

They know that people will pummel the ftp server once it is made available, so torrent seems like a great dissemination method to me. I clicked on Daniels torrent link and within about 5 seconds I was downloading at 150 kilobytes/sec (my ADSL caps at about 170) for the entire duration of the very short download.

I, like others I know of, buy CD's and also happily download the files before the CD arrives to my house. I'm okay with that. I'd just donate the money to them, but the CD's and shirts are just too cool to miss out on. Now that I am making good money for the foreseeable future, I think I will also be donating good money too.

Re:100% Free?

[Anonymous Coward]

a full operating system that installs in minutes.

This is no exageration either. I've been using OpenBSD for about 6 years. It typically takes me about 5 minutes to install OpenBSD, with seperate /, swap, /usr, /var, /tmp and /home.

And, after it reboots... I type startx and X comes up. No problems.

Then I just go into /etc/rc.conf and change xdm=NO to xmd="" to start X at startup. Add users, set softupdates in /etc/fstab, etc, and enjoy. Occasionally cvs up'ing the source to patch some obscure bug that is unlikely to become more than an annoying little short lived DoS, then in 6 months, tar my ~ and some /etc up, re-install from scratch and enjoy.

I love OpenBSD. Seriously, if I had a shitload of money, enough to secure my future with a home of my own, etc, with lots over, I'd be giving them lots of money. The future is going to be dominated by computers (it already pretty much is), so I think a free, secure OS is really important for everyone fortunate enough to own a computer.

Re:100% Free?

[hackwrench]

Nobody said there was anything wring with it, just that it makes OpenBSD conceptually less than 100% free.

Re:100% Free?

[Anonymous Coward]

Nobody said there was anything wring with it, just that it makes OpenBSD conceptually less than 100% free.

No it does NOT! Specifically, only the official ISO CD image is not free to copy. The files within that image (OpenBSD itself) are free to copy.

OpenBSD is not the official CD ISO image, but the official ISO CD image CONTAINS OpenBSD.

OpenBSD is free to copy, the image is not. Just mount the image and extract the files legally! Sheesh.

How hard is this to understand? OpenBSD is as free as the BSD licence against it, but a very specific image of a filesystem, which just happens to also have OpenBSD within it, is not free to copy. OpenBSD and the official image are mutually exclusive things.

The OpenBSD concept

[hackwrench]

The OpenBSD concept is much larger than the source code. Conceptually OpenBSD is encumbered.

Re:Theo? Activism?

[Anonymous Coward]

Whoda thunk it?

Could you ask in English?

Mythbuster: FCC and 802.11 channels

[aphor]

In 802.11 and other Direct Sequence Spread Spectrum radio equipment, ALL frequencies in a broad spectrum are used simultaneously at lower power. The "Channel" is really a virtual channel, and not like the narrow band of Citizens Band spectrum that CB each radio channel uses. Instead, DSSS channels are different Spreading codes governing how the broad spectrum is encoded by the transmitter and then decoded by the reciever to distinguish signal from noise. The signal profile at different frequencies reinforce each other after decoding. Thus, baseband interference has a low statistical impact on the total S/N ratio.

When one 802.11 radio transmits, it uses ALL of the analog bandwidth that other radios on different spreading "channels" are listening to. When a listening radio is on a different channel from the DSSS transmitter, the resulting decoded signal is part of the background static called the "noise floor". Thus, the only thing you gain from operating an 802.11 radio outside of FCC defined channels is obscurity. 2.4GHz digital spread-spectrum phones do this.

More importantly, the 802.11 radio standard operates in a band of frequencies called the ISM band. It is UNREGULATED below certain transmission power levels. Screw with your spreading channels to your heart's content. However, if your firmware allows you to increase the output gain of the radio transmitter, then you may be able to run afoul of the ISM unlicensed operation specs, especially if ytou built your own yagi antenna...