Submitted by Now this is interesting. We see what is at its core a very valid concern, in practice not a problem to anyone, and, thanks to the tone of the press release, close to trolling. The Free Software Foundation Latin America is complaining about something that has been known for a while – there is some non-Free code stuck in the Linux kernel (mostly firmware). A valid issue of concern from an idealogical viewpoint, but sadly, the tone of the press release turns this valid concern into something close to trolling.
So, what exactly is the problem here? As you all know, the Linux kernel is licensed under version 2 of the GNU GPL; however, there are parts of the kernel that are not Free and are closed-source. This is mostly driver firmware code that lives in a separate folder, with warnings and all that, and this code needs to be loaded specifically. It’s basically code for which no open alternative exists.
Linus (for some reason, addressing him by his last name seems odd) has made it quite clear over the years that he chose the GPL not because of ideological reasons, but because of practical reasons. His stance towards accepting small bits of proprietary firmware code into a walled-off, non-default folder fits this mentality rather well; it increases the usefulness of the Linux kernel for those that just want something that works, while allowing Free software purists to properly ignore the code.
How much code are we talking about? Well, Linux 1.3.0, released in 1995, was entirely Free software – no binary blobs whatsoever. In 1996, version 2.0 contained 143kb of binary blobs in its 4.7MB compressed tarball, growing in size to 1.3MB in the 26MB tarball of version 2.4.19 in 2002. Linux 2.6.33 contains a total of about 2.5MB of binary blobs.
This is, at its core, a valid concern, however, most people simply won’t care. Linux has become a pretty big market, and people simply need it to work, and most of them don’t care that some parts of their machine need some binary stuff to work properly. If the choice is “optional binary blob and working hardware” or “hardware that doesn’t work”, the choice is easy for most.
Raising awareness of this concern is a noble goal at heart, but some form of understanding from the FSF-LA would’ve been prudent. The Linux kernel plays a very big role in the success of Free software, and without it, Free software still wouldn’t have a kernel. In other words, the tone of the press release isn’t particularly helpful, and surely won’t do anything to convince people – in fact, it’ll only piss people off. Addressing Linus Torvalds directly, blaming the entire situation on him and him alone, calling it a bait and switch – that’s not constructive. It’s trolling.
Purists want one thing, practicalists want another thing. But the truth is, it’s easier to just opt out from using those binary blobs than it would be for random people to find them when they happen to need their devices.
Still, I don’t quite think having binary blobs in the kernel source tree is even an issue in the first place, what is an issue is this: do they have permission to distribute those binaries? If they didn’t have I’d understand all this ruckus, but however if they indeed do have all the needed permissions then there is absolutely nothing to complain over!
For purists, it is: suppose I’m a purist and I write some code for the Linux Kernel (in the pre-binaryblobs era). Releasing the code under the GPL, I can rest assured my code will only be used in accordance with the GPL, and not be distributed bundled with evil proprietary code.
Now Linus decides to ignore the GPL and distribute *my* code, along with evil proprietary code! I didn’t want that – that’s why I released my code under the GPL – and Linus didn’t ask for my permission.
(i’m not really a purist, and haven’t contributed anything to the kernel apart from some bits of documentation, but this makes it easier to get the point across
).
Now Linus decides to ignore the GPL and distribute *my* code, along with evil proprietary code! I didn’t want that – that’s why I released my code under the GPL – and Linus didn’t ask for my permission.
First of all, the binary blobs are not part of the kernel itself in any way or form: they do not run in the kernel, they do not run on the CPU itself, and they are not needed for any part of the kernel itself to function.
The only time the kernel uses the firmware blobs in any way is when it just uploads them to the hardware device, that’s when the device executes it.
As such I just can’t help but feel like someone is going for pure strawman arguments here.
When I download linux-2.6.37-rc1.tar.bz2 , I find the firmware blobs inside. I’d say that makes them part of the kernel in at least *some* form.
You could argue what happens at run time is irrelevant: the GPL governs distribution, not execution.
I honestly believe both sides of the story have some merit to it, and deserve to be mentioned…
No. That makes them part of the archive containing the kernel source.
Differences, oh my.
“When I open my fridge I find a beer inside. I’d say that makes it a part of my fridge in at least *some* form.”
Makes about as much sense.
Edited 2010-11-10 10:32 UTC
“When I open my fridge I find a beer inside. I’d say that makes it a part of my fridge in at least *some* form.”
Makes about as much sense. [/q]
The beer is not required to manufacture nor operate the fridge, it’s not part of the plans for the fridge and in fact, is not a part of the fridge at all, only a part of its contents.
“When I open my fridge I find a beer inside. I’d say that makes it a part of my fridge in at least *some* form.”
Makes about as much sense. [/q]
The beer is not required to manufacture nor operate the fridge, it’s not part of the plans for the fridge and in fact, is not a part of the fridge at all, only a part of its contents. [/q]
Same with those binary blobs. They are not required to run the Linux Kernel at all. They do, however, aid parts of the Linux Kernel in working with the underlying hardware on specific systems, but the Linux Kernel could just as easily operate on those systems without the binary blobs being used.
For example, the QLA2000 SCSI controller requires a binary blob to operate. However, one could just as easily boot off of a CD to run the Linux Kernel on the same system. The hard disks behind the QLA2000 SCSI controller and the controller itself won’t be accessible to the user, but the system will operate. Adding the binary blob for the controller will allow the controller and its associated disks to be operational for the user; but is not required for the Linux Kernel to operate on said system, even if said systems is otherwise useless – e.g. it has no other drives for the user to run the system from.
When I download a driver.zip, I get a readme.txt along side the actual filename.drv. I can choose to read that .txt or not. It is in the driver package but the actual driver works with or without readme.txt
When I download kernel.zip, I get firmware along side the actual linuz.bin kernel binary. I can choose copy those fimware blobs along side the kernel binary or not. They are in the kernel package but the actual kernel works with or without them in place. You may get the firmware when you download kernel.tar.gz but you don’t have to pay any further attention to them then you pay to random text file under /docs/.
Perhaps they should package a kernel.tar.gz and kernel-withfirmware.tar.gz but otherwise, it seems like a non-issue.
Personally, I think firmware should be on the hardware board (it’s easy to re-flash through an “update wizard”) and drivers should be open source by default practice. I have never heard justification for making hardware drivers “competitive advantage” that balanced out the end user grief currently caused by such thinking. Until an open source comparable alternative becomes available, one has to use the closed alternative that works though. I’ll happily use Nouveau when it’s performance and function support matches or exceeds Nvidia-GLX for example.
You could argue what happens at run time is irrelevant: the GPL governs distribution, not execution.
The GPL is such a harsh DRM that it not only directs what have to be in the code or not, yet also what is bundled in an archive file, being later of any use or not ? Haven’t I heard that RMS doesn’t like restrictive licencing scheme ? Where is freedom ? Everyone is free to use or not these binary blobs (as you call them) Remove them and you skip out the users of Linux that do not care about this silliness…
Kochise
These binary blobs are primarily firmware for various hardware devices… The alternative is that the firmware is stored on a flash chip on the device itself, however this would increase the price of said devices.
Either way you end up running the binary code, although at least with it being loaded by software you have the potential to modify it more easily.
Meh real purists use OPENBSD everyone else just be trollin’
Tomorrow’s likely headline.
Why not a Celebrity Death Match between Stallman and Linus? :p
I think katana would likely beat nunchucks…
http://xkcd.com/225/
But no match for Steve Job’s with his ninja stars
But in the end Richard Stallman will catch them all off guard with the poison darts from his musical flute.
Does he dodge via high-speed soulja boy?
On the one hand, it pisses me off royally when a piece of hardware refuses to work due to some stupid missing driver. Wireless cards, for example–which have pissed me off to extreme levels to where I just don’t care any more, I’ll take the blobs.
On the other hand, it seems that some FSF-approved distros do surprisingly well when it comes to compatibility with some hardware, and in some cases (somehow) manage to work with some hardware that refuses to work on just about any FSF non-compliant distro out there (including even Ubuntu and openSUSE).
I think my stance is that I’d rather be using a 100% “free” OS, but will tolerate some blobs if absolutely necessary. So far, those tend to be display drivers (nVidia) and wireless cards (Broadcom). To be fair, nVidia’s proprietary drivers actually are very good and tend to be well-supported, but it can still be a pain in the ass to get them installed on some distros, or at least an inconvenience having to do so manually.
Repeat after me: firmware blobs are not binary drivers. Firmware blobs are not part of the kernel. Firmware blobs are not part of the OS. Firmware blobs are not run on the CPU.
You’re confusing “binary drivers” with “binary firmware”. They are not the same thing. You are arguing about the wrong thing.
And they still use binary firmware, without any “OMG, i haz not the open-sources!”
Again, completely beside the point. Firmware is not a driver.
Really, what needs to happen is for more end-user education into the differences between “device firmware” running on the physical device, and “device driver” running in the kernel as part of the OS. They are *very* different things, and the OSSness of one does not affect the OSSness of the other.
Correct – the distinction is that the firmware runs on the device’s built-in processor utilizing its own “private” memory space. It should have no direct access to the host computer’s processor/memory space without the kernel that is running on the host computer explicitly providing access to it via the driver.
It is obvious this distinction exists.
The GPL talks about a ‘work’ that is ‘distributed’.
From the license:
It’s obvious the Linux Kernel is a work that in part contains another GPL’ed work (many, actually). You could certainly argue that because the firmware blobs, even though they’re not drivers, are distributed with the kernel, they are part of ‘the work’ and thus should be licensed under the terms of the GPL.
Now I’m not saying that’s the only interpretation – but it’s certainly not an entirely unreasonable one.
I’m not taking sides – but if the question whether it’s acceptable to have binary firmware blobs in the kernel were a trivial clear-cut one, this debate wouldn’t keep popping up like it does.
A distro like ubuntu contains GPL and non GPL (LGPL, BSD, MIT, etc. etc.). Your argument is that since this is all distributed together with GPL software, it must all be GPL (or nothing).
Besides being absurd on its own, there is nothing in the GPL that mandates this. The Kernel is a software bundle, as is a distro.
Not correct. The distribution clause only kicks in at a certain set of circumstances. The deciding part of the verbiage is the bold part below.
You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License
Since the binary firmware is not derived from (part of) the Linux kernel, nor does it contain (part of) the Linux kernel, it is not part of the work. This is not derivation, but mere aggregation and that is perfectly allowed by the GPL.
The stance of the FSF against binary firmware is that it forms a non-free, non-modifiable piece of proprietary software for the end user, but it is modifiable by the copyrightholder, so it could in theory also be free software. The only thing that could be said about the aggregation of binary firmware with the Linux kernel, is that Linux, in doing so, promotes non-free software.
I’m not too worried about it, as RMS has indicated that if that same loadable firmware was to be burned into ROM (Read Only Memory) onto the device, he would be fine with it, as the modifiabilty would be precluded from everyone. I prefer the fixable, loadable firmware, over the kind that is burned unfixable into ROM. The software currently is unmodifiable by me (or other end users) in both cases and burning it into ROM would just make our peripheral devices more prone to carry annoying bugs.
Also, the firmware is very device specific code. While the possibility of malfeasance is present (spying and the like), I don’t think there is a lot of firmware out there working against the owners of the devices. The primary function is to control the peripheral device.
> I prefer the fixable, loadable firmware, over the kind that is burned unfixable into ROM.
So do I, with the caveat that it doesn’t come with an over-restrictive license.
Firmware in ROM is, well, already in ROM, it was purchased together with the device and no copying or processing is involved in making the device work. OTOH, firmware distributed with drivers is subject to licensing terms, which may prevent people from distributing it along the opensource drivers. To me, as long as the blob is freely redistributable, it is as good as firmware in ROM.
I assume that Linus Torvalds is savvy enough to keep the blobs that are shipped with Linux above board. The last thing the kernel project needs is a copyright claim with merit.
/ot
Why not wish for something at least theoretically doable, like, I don’t know, world peace or something. Here in my dimension, I cannot recall any end-users who give a fraction of a fcuk about such things.
Repeat after me: Copyright restricts distribution. Unauthorized distribution of firmware is not different from drivers or music or books.
Not true. I believe the Mercurial definitive guide at http://hgbook.red-bean.com/ is a counter-example.
Freely distributable, modifiable, commentable, etc. You may even send feedback to the author. I haven’t looked for a copyright notice (or read the book yet) but I presume he retains the copyright, no matter how free he makes his work for others.
Being a copyright owner doesn’t make you an evil person all of a sudden. Copyright was legitimate when created. It’s what people have been doing on its behalf that has become ugly in some cases.
*Palmface* over misunderstood oversimplification.
Of course, copyright gives the copyright holder the right to control distribution. Of course, he may give it up, retaining other rights (like attribution) or imposing conditions on distribution (like GPL). That’s what free software & co. are about.
What we’re talking about here is distribution of binary blobs that aren’t free software and their copyright holder retains the right to distribute them. IOW, distribution of those binary blobs which Linux distributors don’t have the right to distribute.
I don’t understand how your post relates to mine, in which I specifically quoted the specific piece of text that seemed arguable to me. I was not replying to the news item or making my position known about it even though it was my first post on this item.
And what is the “misunderstood oversimplification” about? He/she said “Copyright restricts distribution” without any moderating adverb and I said “not always, see this example”. I don’t think there is any oversimplification there.
The firmware is often bundled into the driver so when Windows folk think of “driver” they are thinking “driver + firmware injected when driver loads”. The driver is no use without the firmware and the firmware may make the hardware work but without a driver…
To make matters worse, when you ask a manufacturer why they don’t provide an open source driver “we do not own all the relevant patents” or “the driver contains trade secrets” which boils down to “the firmware is bundled into the driver and we won’t separate the two”.
It’s not such a surprise that people consider firmware part of the driver.
It’s firmware, it not part of the Linux kernel and it does not run in the OS. Firmware != driver. Heck, even OpenBSD ships with firmware blobs. It would have been nice if the devices in question did not require firmware upload to work but in the end it doesn’t really matter and it doesnt make Linux non-free.
To say that this makes Linux not OSS is incredibly clueless.
Exactly. It’s a lot of hullabaloo over nothing.
Well, the FSF runs computers with a free BIOS (read firmware) at their offices as well. They want a completely free stack. If their is a bug in the firmware, you want to fix, you need the source. Simple as that. That is how the FSF got started.
A lot of the basic functionality is written in microcode/firmware.. and rewriting this manually for every device in existence would be unbelievably tedious, and for smaller microcontrollers the resulting “free” replacements would probably be identical given the constraints of the hardware.
Documentation for writing drivers and loading firmware is what is needed from companies.
As has been repeated here, firmware is not the same as a binary blob kernel driver.
Edited 2010-11-10 00:17 UTC
Not looking to provide flame bate (though this will probably be interpreted as such by some) but doesn’t that pretty much describe the FSF these days? There was a time when the FSF, and those involved with it, contributed a lot of valuable things to the F/OSS landscape. Is that time past? All I see from them now is new licenses and complaints when people don’t follow what the FSF thinks they should be doing. Quite frankly, I find them annoying. Do they do anything productive these days? That’s a genuine question, I’d really like to know since all I see from them lately is bitching and moaning about one thing or another.
The guy that wrote this announcement, Alexandre Oliva, is maintainer of Linux-libre, GCC developer, and Red Hat employee. He does lot of things.
As for FSF in general, they do lot of things too. Reason why you don’t know about it is the fact that they don’t brag like some. They don’t rename operating systems to their firs names, they don’t name VCS after themselves, they don’t re-brand other people’s work. Usually their work get re-branded by others, and that is reason you don’t hear about them. Also, they don’t claim they are Gods and they don’t hold trademarks on names of programs they write. Guess who does all of the above?
FSF only raises issue when there is something critical for all users. They try to protect your freedom. You don’t know how free firmware is important until you have a device that doesn’t work and mfr tells you that they don’t care. But no good deed goes unpunished, so those who defend your freedom get called whiners. But announcement was not trolling. Next release of Linux kernel will have firmware clearly separated, and I think this has to do with this announcement. It would make job easier for Linux-Libre folks, since non-free firmware will be easier to locate and rewrite. Somebody has to do that stuff, or some day mfrs might decide to enforce copyrights on firmware against GNU/Linux, and then Linux kernel will have no hardware support. Then you feel sorry that you bashed people who tried to help. But it will be too late. It is really easy to be “moderate” or “pragmatist” when you know there is someone else to take a hard line and do dirty job, no matter what you say. Without people like RMS and Alexandre Oliva, we wouldn’t discuss about free firmware now. We would be running everything proprietary, and maybe even lack right to read:
http://www.gnu.org/philosophy/right-to-read.html
Btw, I am not trolling. If I am, then I got trolled into trolling 🙂
But the difference is OpenBSD will only distribute firmware if it is under acceptable licensing terms.
Define acceptable licensing terms…
http://kerneltrap.org/node/4118
Seems any firmware which can be redistributed without any contractual obligations.
Perhaps but these folks are against all closed firmware regardless of license.
These people are complete idiots. Typical Linux users.
FSFLA who’s they?
They can always use GNU HURD. That’s totally free.
Not sure that will suffice to make and keep them happy… What is life worth when you have nothing to complain about?
Oh, maybe the #1 position in the vaporware hall-of-fame is a good subject?
If the GNU/Hurd will take over the world, this will be only the first step accomplished.
The next step will be forcing the use of Open Hardware.
Here is a picture of RMS and his laptop with a FSF sticker over the IBM logo:
http://stallman.org/photos/rms-working/img_0902.jpg
I would really prefer if the source to this firmware was available. This is for two practical reasons:
1) If the hardware company folds, we have a plan B.
2) I can see if they are doing secret things with my data in there.
That said, I understand that it may not be possible, so I’ll just buy hardware that doesn’t need them whenever possible.
Note to hardware vendors: If you want me to buy your stuff, release proper open drivers and firmware source.
Just because the device does not need firmware uploaded it doesn’t mean it doesn’t contain closed-source firmware. In fact, your computer is already full of closed–source firmware.
Uhm, every single device in your computer right now, from the harddrive, to the NIC, to the graphics card, to the optical drive, to the mouse, to the keyboard, everything, is already running a binary firmware blob.
The only difference is that most firmware is “burnt” into the chip on the hardware at the factory. Some devices support loadable firmware, and that’s the firmware that gets shipped on floppy, or as part of the driver, or with the OS, or with the kernel. Regardless of how its shipped, it’s all the same.
No such thing as computer hardware that doesn’t need and have firmware. It being not being flashable or changeable in any way doesn’t mean it’s not there.
This is an old problem, that’s already been mostly solved. I believe it’s possible to disable binary blobs when you (or your distributor) build the kernel, and there already exist distributions that do this. This is a problem that already has a compromise solution in place.
I’m curious what the FSF-LA’s goal is, here, as they’re bringing up some fairly old news.
This is an old problem, that’s already been mostly solved. I believe it’s possible to disable binary blobs when you (or your distributor) build the kernel, and there already exist distributions that do this. This is a problem that already has a compromise solution in place.
As far as I know, most distros do offer several versions of kernels and there likely exists also one which doesn’t ship with the binary blobs. And as you said, there exist whole distros dedicated to shipping only F/OSS software, no proprietary anything. And you know, you don’t even need to disable the blobs anywhere: just delete them.
It’s more than just “some distributors pull them out.” I do believe that, when you check out the kernel’s source and run the configure script, it has a high-level option to not include any binary firmware blobs. An effective solution to this problem is built right into the kernel’s make system.
Solutions to this problem don’t just exist, they are well-supported and incredibly easy to use.
It would seem, based on the reading of comments on that article, that the guy who wrote this was one of the people who helped clean up the mess in the first place (although I can’t confirm that). I found this blurb for example:
“When lxoliva started working on linux-libre, it *wasn’t* marked as such. It was scattered all about the kernel tree sometimes in separate files, sometimes mixed in with the main driver, etc.”
I do know it was now all moved to a seperate directory.
But this is a good thing about Linux, people can do with it what they want. Even if they do not agree with Linus.
Yes. AFAIK Debian distributes its kernels without binary blobs, but has the blobs available in separate, easily-installable packages in the ‘non-free’ section.
Seems like a fine solution if you ask me. Perhaps it’d be nice if kernel.org also started distributing the firmware as a separate package, wouldn’t that end this discussion once and for all?
Some firwmares have version dependency against specific drivers. There are cases where firmware for newish devices evolve quickly changing programming interface along the way. The drivers and firmwares need to be updated in lockstep in those cases. Distributing them together is just much less painful for everyone involved.
I’m all for idealist – they are the driving force of change. That said, this press release reads like propagandist drivel. Like most propaganda, the facts are all right, but the conclusions are all wrong.
Users do not need to be “warned” about this. Users are the reason firmware blobs exists: primarily because they scream bloody murder when some piece of hardware doesn’t work. The vast majority of users do not care one bit about the fact that there is an opaque blob of bits in the kernel to enable their wireless card to function – they just want it to work.
Yes, it would be great if the hardware guys would supply the source code for this stuff. But they don’t. Why don’t these guys target their angst in the proper direction – at the hardware vendors?
Its not like this is a new issue, it has existed for years. The kernel guys have decided since this stuff doesn’t “run” in any conventional sense, and since users want it, and since it doesn’t affect the kernel’s operation in any meaningful way, they just drop in into the source tree and move on. Its either that or refuse to include it, which would punish users for no good reason.
It is beyond ridiculous to say that Linus, or anyone connected to kernel development, are trying to “bait and switch” users.
I see in their statement where they talk a lot about freedom, but it seems to me that these people are more interested in forcing people to use open source software than they are about freedom.
IMHO, having REAL freedom also includes the ability for one to have the freedom to use proprietary software if they so choose, so I don’t really see these guys as being much different than those companies who want to completely lock down your PC so they have absolute control over what you run.
That being said though, I think it’s a good thing to have extremists on both sides… whether it be politics, religion, or whatever. It probably helps keep things in balance.
and a little bit more interesting too. How dull would the tv news be without those agitators?
Without wanting to make any ads. I currently don’t use OpenBSD and don’t have it installed it anywhere and I am not a big fan of it, because I simply prefer other systems.
BUT if one wants a stable, usable and secure system that DOES NOT come with blobs it seems to be the best choice.
Just in case someone is looking for something.
For me they are too conservative and IMO hyping their security too much, even if it is very secure.
Just wanted to say it, because the “really free/open source” Linux distributions don’t seem to get a lot of support. I hope something like Haiku or Syllable could become a real FOSS system.
Oh and what about the “Only Free Software” option (which you can enable at the boot loader of the install CD) in Ubuntu? Does it come without Firmware?
OpenBSD does include binary firmware, look in /etc/firmware on a post 4.8 snapshot.
In this directory is both the binary firmware being loaded into devices and a “license” file that explains the terms of redistribution.
Firmware is code executed by the microcontroller on the device, originally this code was located on embedded ROM’s and only upgradeable if shipped back to the manufacture, given the complexity of newer devices (802.11) it became easier to have this code be uploaded and executed when the driver was loaded, as a bonus an updated driver can contain firmware which fixes functionality of the device.
This can be extracted from the drivers and with permission from the manufacture, redistributed along with the operating system.
Every device in your system has firmware/microcode, as mentioned by another here.. even your hard drive includes a firmware, it’s a necessary component of modern devices, it defines how they function.
I use OpenBSD myself, but, I backup my claims before spreading misinformation.
Firmwares are not binary-blob drivers, which exist in kernel/supervisor space and can adversely affect the function of the system.
This loadable microcode is now new.
I was writing loadable microcode for VAX & PDP-11 Computers in the early 1980’s.
In my simplistic view (waiting for incoming missiles) if you have the hardware and it requires loadable firmware/microcode to run then how you load it is immaterial, i.e. it don’t matter if it loaded from Windows or Linux or unix or Fred’s Toy O/S.
Again in my simplistic view, if you have legally purchased said H/W then it is not complete until you load the F/W.
What does matter to me is that said firmware is binary identical to that shipped by the H/W manufacturer. If you start messing with their original then sorry, you have messed up.
IN my view if you write your own code and release it under the GPL then that’s fine as well. Just don’t break into the manufacturers blob. Leave it alone and IMHO (IANAL etc) you should be fine legally.
If H/W makers decide to insist that the F/W can only be loaded into the device from a certain O/S then I think they are open to lawsuits but given the totally broken & messed up morass that is the US legals system I wouldn’t be surprised if such suits failed.
The linked article specifically mentions firmware blobs and (incorrectly IMO) groups them in with what is generally described as binary blobs. OpenBSD does not accept binary blobs, but they do accept firmware blobs. The main difference is that firmware blobs are not executed by the CPU and do not contain any code that interacts with the kernel.
Someone also pointed the JSON license is not free (I’m not quite sure which provision, perhaps not using it for evil).
Just because there is a chunk of binary does not mean it came from ANY source originally, nor does it follow that if there was a “compiler” that it could be modified into anything useful. Some of that info is data not code – would changing the whole thing to unsigned char firmwareimage[] = { 0x00, 0x01… }; make the happy – they would have source!
Consider FPGAs – the source isn’t code as such, it’s logic.
The stupid part is that if the manufacturers used ROMs that you could NEVER fix so that if specs or regs change you would have to buy a completely new part it would be GPL compliant. Because it is in volatile memory, or flashable memory, it becomes a matter for controversy.
Stallman first stated it, but in effect he and FSFLA would prefer a Masked ROM chip soldered to the board that could never be upgraded to one that there would be no useful way to change the binary but theoretically could or has a ram that downloads the code.
I like opensource hardware and would love to have all the schematics and PCB layouts of the cards, but 30 years ago no one thought that it was important for freedom. The same freedoms apply though – examine and learn, use for any purpose, modify, distribute mods (or instructions). Hardware being completely non-free doesn’t bother them at all?
Firmware sits in the middle of this. Again, if it was a masked ROM or etched board it would be hardware, but because it is the implementation for hardware?
But look at the four freedoms again…
To use it for any purpose – yes, and you can modify i thought it would be difficult.
To study how it works – You would need to be a master EE, and even then you might not know how it works or affects things. I doubt even 1% of the FSFLA members could explain to me how the device they are concerned about the firmware works or even how to upload. Have they even heard of JTAG?
To redistribute it and any mods – you have that.
Access to the source is NOT a precondition for distributing modifications – the firmware or the source for it is NOT generic so doesn’t apply to different chips. Talking with something over a standard bus, or between two devices it is important since it lets me do things like code once and run it in a PPC Mac and an intel PC if they both have PCI. But these are microcode for some very idiosyncratic chip.
If they want to play with such things, for example wifi, there are GNU radio – software defined radios which program in standard languages. If they can’t program that, and wifi, gps, bluetooth aren’t simple – why is it important to have access to source you can’t understand so won’t be able to modify?
Updating a firmware through a wizard is no more complex than installing an updated driver or app. Flash memory is also cheap; they should put the firmware in flash memory. Firmware can still be updated and does not have to exist within the driver bundle. Driver’s return to being a basic bridge between kernel and hardware interface.
Drivers stopped being separate from the firmware with the release of the first winmodem back around 1994. Note that is *16* years ago. I don’t think you can make the case that it should be changed when it has been a functional, and accepted method of delivery for 16 years.
Yes, the Winmodem has been the posterchild for shit developed hardware. I’d suggest it as evidence that the driver+firmware is a consumer hostile and limiting design. Granted, my first introduction was about 16 years ago when helping a friend setup a BBs; “WTF.. where are the jumpers? How do I tell this where Dos+Renegade is going to talk to it?”
Mind you, my introduction to Winmodems was through there limitations rather than any benefit the design offered. Maybe a poor first impression but 16 years of experience since seems to confirm my sentiments.
But, even with 16 years of Winmodem.. there is no reason they can’t go back to separate firmware and driver. Especially with flash chips available and dropping in price. If your harware uses firmware, give me a damned update utility that flashes it to the hardware and later flashes updated firmware images as needed. Stop artificially limiting my hardware choices with BS like driver bundled firmware and support for only one OS.
Actually, I don’t really see a reason why more hardware can’t function like hard drives. Somehow hard drive manufacturers manage to compete though they’ve all agreed on a standard hardware interface that any OS can implement.
Linus has stated this many, many times. The only reason the GPL is used is to keep people playing fair, not so Free Software can take over the world. I find it funny that people so out of touch that they don’t realize the project is borderline hostile towards their organization decided to make press releases pretending this isn’t the case.
Maybe they could spend their time better by actually accomplishing something and getting HURD (which _is_ an ideological project) out the door, so that innocents like them won’t be caught up in the non-free software scam that is linux.
In closing, here is the leader of the FSF eating something off of his foot. http://www.youtube.com/watch?v=I25UeVXrEHQ (I can’t help it, the fsf brings out the troll in me)
Edited 2010-11-10 00:47 UTC
FSFLA is not FSF, it’s an independant organisation which speaks for themselves.
No, I think it’s just because you are a typical douchebag who relies on ad hominem attacks because you have no real arguments.
Were you being serious? I was sort of joking with that last line, but you are essentially calling yourself a douchebag too, since you are resorting to an ad hominem attack
My argument (I wasn’t arguing by the way, I was pointing out something funny) was that linus has said he doesn’t care about the ideology of the FSF, especially in regards to the linux project, several times and quite publicly. If the leader of project explicitly says he doesn’t care about what you stand for, it is sort of funny when you turn around and make a big deal over something that indicates that the linux project doesn’t really care about what you stand for.
My argument against the FSF in a more general way is that they are the George Bush of the software world. The like to use highly emotionally charged language that isn’t really appropriate in order to pretend what they are doing is more important then it actually is. Nelson Mandela stands for freedom, Mahatma Gandhi stood for freedom, the FSF stands for source code being available to everyone, and calling that freedom is inappropriate and insulting to say the least.
I did feel a little bad about that comment, thank you for providing me the opportunity to elaborate.
I figured you would say that, but I figured you should take what you dish once in a while, and I will continue to do so in this post.
You have systematically posted that (admittedly gross) link whenever FSF has come up, which has absolutely no bearing on what he says or any relevance to the subjects whatsoever.
It’s like trying to question the german football coach’s knowledge of the game because he ate a booger infront of millions on live tv during the world cup.
Instead of arguing you are trying to discredit him by saying ‘look! he ate some shit off his foot, don’t listen to him!’. It’s really spineless and I think you should try and grow a pair instead.
I totally disagree with RMS claims that proprietary code is unethical, I also very much disagree with the use of the word ‘freedom’ when what it describes is ‘rights’. And I will ARGUE against those things, I will not however, make fun of his person in an attempt to diminish his arguments because I don’t like them. That’s pure cowardice.
Dude, I literally posted that _one_ other time on this forum, and it was in the context of talking with werecatf about how much we dislike the man.
On the other hand, that second paragraph you will find in various shapes and forms in my posts here over the last few years. I have yet to have anyone challenge it in any way other then saying “I disagree” with nothing else to back it up.
I’ll admit though, this time it was totally out of context (hell, right after I admitted to trolling). But posting something once out of context is not systematic.
Well, chances are someone else has posted it here aswell and since I’ve seen you do it a couple of times I assumed that it was also you. But I still think it’s crap thing to do even twice.
I don’t know why you hate the guy, I know that somethings he says I agree with, somethings he says I disagree with. But even if I did have some reason to dislike him in person, then I would bring that up instead of as you do go at him through something entirely else since I seriously doubt that you hate him because he ate some crap off his foot.
If RMS was my best friend, I would post a link to that video. Since I’ve never even come close to meeting the man, the video is less funny. (I love embarrassing my friends.)
It’s still funny, though. And, it was just an aside to google_ninja’s post.
Google_ninja is right in how the FSF has appropriated Linux for it’s own ideological goals that do not meet with Linus Torvald’s stated goals.
However, you chose to completely ignore the meat of his comment. You have not addressed his true claim in the least bit, and instead attacked an admittedly frivolous part of his comment as if it was the heart of his statement.
Well, given that Google_Ninja’s states that he really dislikes RMS this was posted as a character assassination and nothing else.
If so then Linus has likewise ‘appropriated’ both the GPL, GCC and big chunk of linux userland for his own goals that do not meet with FSF’s stated goals. What’s the difference?
I personally have disagreements with alot of FSF’s ideology. And while I can definately relate to the practical problems that started the whole movement (printer driver) I think that the whole ‘freedom’ thing is propaganda drivel and should be called for what it is, which is ‘rights’ covering distribution of open source code.
But looking past that I find alot of things they say that I agree with, I think the GPL is a good open source licence for cooperative development (and given it’s vast popularity I’m not alone in this), I also think their fears of software patents are especially well founded, and they were quick in warning about this.
And while their views are often extreme when it concerns proprietary code (too extreme by far for me), it’s not as if there’s not the same extremes on the other side of the spectrum (does anyone think Ballmer suddenly thinks open source is ok?).
Now I see many things that people may disagree with Stallman on, but I don’t see why someone would dislike him the way Google-Ninja does. It seems to me that often when people don’t like GPL and/or FSF, but feel they lack arguments that would make sense/carry weight they will go for attacking the person. And I find that shitty and weak.
The difference is FSF (and, especially FSF-LA) are using Linux as the poster child to push their ideology.
Linus is NOT using GCC to push his ideology. He is using it to compile software, which is the intended purpose of GCC.
I, too, am a fan of free software. I am bothered by the FSF’s “our way or the highway” attitude, and the extremism. However, you still haven’t responded to google_ninja’s original statement:
Every post in response to his has been about the link of RMS eating something from his foot. None have been about his suggestion that FSF should actually finish the HURD, which is aligned with their ideology, instead of pretending Linux was born from their goals.
However, you very quickly resorted to name calling:
And ignored the real argument that he did make.
FSF and FSFLA are totally separate organisations.
Of course FSF realises (though not initially) that Linus isn’t using GPL for any of the ethical reasons FSF was hoping he would use it for. I haven’t personally seen these press releases where they pretend this is not the case though, can you point me there (and again, FSFLA is not FSF).
FSF obviously knows the Hurd train left ages ago. They themselves acknowledges that when Linux started to gain traction and having adopted the GPL as licence they focused efforts on GCC and userland to assist that project and left Hurd by the wayside in terms of development focus. This is nothing new, but the abandonment of the Hurd is something FSF detractors like to point at as a ‘look, this is what they contribute with’, which is a pure lie. Meanwhile GCC is still the compiler on which pretty much the entire open source movement and alot of the proprietary software outside of Redmond depends upon. So pretending that FSF isn’t offering anything but GPL and their ethical software propaganda is just ridicoulus. GCC is most likely the most widely used compiler software in the world, GPL likewise is the most widely used open source licence, again I don’t like their freedom propaganda but their impact on the open source landscape and computing in general is undeniable.
[/q] [/q] [/q]
This has been discussed over and over again, no the Hurd is dead, FSF knows it and in hindsight it would probably been better for all involved if they had not pinned their hopes on Linux as the Free Software OS flagship but they did. On the other hand, by doing so Linux got a strong compiler and likewise strong userland alot faster than it would have done otherwise, particularly having a free compiler was hugely important for Linux in order to attract a large number of developers who could assist in the project.
Well, there’s http://www.debian.org/ports/hurd/ it doesn’t seem like dead and I think FSF people stiil believes that Hurd will come someday and free our computers from evil.
If they did then they would pour their resources into it, but they’re not. People keep working on Hurd because they think it’s fun, just like people working on ton of other open source os projects that realistically has zero chance of making it big.
So… are any of those Pepsi cans Stallman’s? Surely he knows that Pepsi’s recipe isn’t open or available to Pepsi drinkers.
Why doesn’t he support free and open soda recipes?
Alexandre has done this sort of grandstanding about Linux before.
I first saw it on on the LKML back when the GPLv3 was first posited and Linux chose to stay with GPLv2.
He engaged in a debate (and I use the term loosely) with Linus. You can probably find all the posts in the LKML archives but they get pretty tedious.
And to be fair to the wider FSF, I don’t think his personal opinions even represent how they feel.
I think he reared his head on the fedora-devel list a time or two as well. Very tedious, very fast.
What’s interesting is that there doesn’t seem to be real connection between FSFLA and the FSF, other than the names being similar.
I guess it’s easier to complain about how the people taking the risks (investing their money into developing products) provide access to their hardware than it is to get a completely “free” working kernel like GNU Hurd. Maybe they should quit whining and start writing some functional code.
Real people need to make a living. When you have folks paying for your living through their tuition in acedemia, you have the luxery of idealism. When it’s your money that is at risk, then in my opinion, it gives you a lot of say as to how “free” and open you choose to make it. Others can choose not to use it. But standing up and critisizing Linus for including these folks’ binary blobs that they can easily choose not to compile into their kernels seems like a complete waste of time, effort, and news headlines.
These people at the FSF really need to get a life.
“The Linux kernel plays a very big role in the success of Free software, and without it, Free software still wouldn’t have a kernel.”
Something tells me that any number of BSD folks would tell you otherwise… unless of course this is just a license thing where “free” only means GPL.
This link has an interesting take on the difference between binary firmware and drives and how some companies are now blurring the lines between the two.
http://onlamp.com/pub/a/bsd/2006/04/27/openbsd-3_9.html
Personally I think we should reject binary drivers where possible and it would be great if people like Linus took positive stand on the matter.
And again: the issue here is binary firmware blobs, not drivers.
Differences, oh my.
“They hate yer freedom!”
And this story / discussion thread is demonstration of why I don’t read or write GPL code.
And this story / discussion thread is demonstration of why I don’t read or write GPL code.
I too prefer the BSD license.
Thelema Software Foundation…
“1 – Do what thou Wilt shall be the whole of the law”.
Quite fair 🙂
Crowley Public License
Uh, not the most liberal license around:
What license do you use then?
cc-by
Anyone who takes ANY of them seriously are complete morons, the FSF and co stopped being an asset to OSS years ago, and are now reduced to nothing but trolls using the same tactics that they decry Microsoft as being evil for doing.
They will ultimately destroy the OSS community, just wait and see.
It’s GNU/Linux, you heathen scums.
Linux is Linus’s kernel. He can do with it what he pleases. FSF doesn’t have a working, widely used, & widely successful kernel. If they don’t like how he manages HIS kernel project, then they can find another kernel to piggy-back off of. Or maybe they can use their own kernel. No, I guess they can’t, since their kernel seems to be quite a bit stuck in the “All talk, no action” phase. Then I guess that means that they should shut up & sit down. If they made Linus mad enough, it’s possible that he could remove future versions of the kernel from FSF space by changing the kernel’s license. I doubt it would hurt him as much as it would hurt them. And before anybody gets butt-hurt over me calling it FSF instead of FSFLA, you should keep in mind that FSFLA is just a branch FSF.
Edited 2010-11-12 20:07 UTC
So you basically agree that Linux is proprietary and property of one person? But you still want to attack FSFLA even thought you agree with them? And you attack wrong FSF (no, it is not branch, it is separate organization).
EPIC FAIL.
I think he wanted to say that Linus is Linux’s original developer and maintainer, at first his OS was to be run on the most large range of PC hardware, whatever the drivers… correction, the firmwares were open-source or not. He released his own work openly, that’s not Linus’ problem to release third-parties’ sources as well.
When FSF decided Linux is no more Linus’ property because 1) Hurd is incomplete, needs a working kernel 2) Linux uses GNU tools 3) thus Linux is renamed GNU/Linux 4) Linus adopted a viral DRM that should have forced him to release these third-parties sources OR at least remove the binaries, Linus persisted to maintain this established compatibility, yet removed them from the main tree and moved them aside for user-awarness usage.
What bugs many people is how FSF vampirize Linus’ efforts and kernel and dictate him their narrow-minded view on how the world should bend before them and offer them everything for free on a silver plate, begging for pardon. Linux-free is the attempt to get rid from the proprietary binaries, and well, that’s a good move for the fanatics out there, those that cannot stand freedom to use these binary firmware files or not.
Since it is obviously no use trying to speak, or just dealing with these FOSS-zealots, let’s get move to another more important topic. This ain’t going nowhere, script kiddies are just opening their mouth wide to be fed with open source code, making them believe they owns, checks, rules, controls and can changes how the world spins. Yet this is a fallacy, they just have to grow a little more.
Some said this is easy for zealots requesting to be given free source code, while they are working on academic projects and are thus fed by the nation, not some private corporations or just little companies. They cannot cope with the capitalisms and those companies trying to protect themselves from IT leechers by not providing their secret recipes.
If having such a grunt over closed-sourced firmwares, drivers, softwares, etc, why all those talented moaners don’t try to figure out, organize themselves, create dedicated hardware from scratch (CPU, GPU, audio, network, …) and the corresponding drivers, and offer everything for free to the said community ? Let’s do a real move, if you’re real mens…
Kochise
Edited 2010-11-14 10:33 UTC
FSF doesn’t vampirize Torvalds, it is other way around.
You say FSF is lame because Hurd doesnt work? Where is Torvald’s Photoshop replacement? Where is Torvald’s desktop environment? Where is Torvald’s Emacs? Where is Torvald’s compiler? He must be pretty lame when he didn’t wrote any of those things, instead he co-opted FSF’s and named them all Linux.
How you like them apples? What I said above is may not be politically correct, but it is lot more correct then what you said about FSF.
No, Linux is not Torvald’s property, even it is obvious that some Torvaldists think so. Reality is that FSF started writing free operating system, did the most work and when Linux came along they decided to embrace it and help it instead of competing with it. A nice thing to do. They advocated for people to go and help Linux kernel, they even de-emphasized Hurd and stopped all investment in it. All they wanted is to they get some credit for the OS they wrote, not for kernel named Linux (even though lots of FSF followers worked on it) but on whole OS. They deserve it.
I really don’t get why all this hate against FSF. They sometimes overdo it, but so does Torvalds. He hates C++ fans. He is abusive against peole who like kernel debuggers. He likes to flame people on LKML. He said OpenBSD folks are masturbating monkeys. List goes on and on… So how come everyone from FSF is an asshole while Holly Torvalds never does wrong? All Torvalds did is start one pet project, which was then improved by people who thought they are doing it in the name of Freedom as defined by FSF. So in a way, Linux kernel is Torvald’s bait and switch. I don’t think it open core (at least not intentionally) but Torvalds in a way hijacked credits over other people’s work to satisfy his enormous ego. I can name more than 10 kernel devs who did lots more work on Linux kernel, and yet they get lot less credits than Torvalds.
Lets face it, if there was no FSF, Linux kernel would be alternate kernel for MINIX operating system, it wouldn’t have free userland, and it would probably be under Torvald’s shitty “no commercial but give me code back” license. Than means no Android, no Red Hat, no Canonical. And probably we wouldn’t have this conversation.
If someone’s religious beliefs prohibit working on Sunday, eating pork, uncovering hair or using Linux then it is pointless to argue because those beliefs on based on faith not reason.