Tuesday, July 28, 2009

Microsoft, the GPL, and Nonsense from Folks Who Know Better

The big news this week, causing much contortion and hand-wringing among the Phoney Fundamentalists, is Microsoft's submission of its Hyper-V drivers to the kernel driver tree. Those pre-disposed to view Microsoft as Satan Incarnate are sure that there's a trick, a devious plot, in there, since Microsoft is, of coure, Satan Incarnate.

What I'm finding disturbing this morning is that none other than Bradley Kuhn of the Software Freedom Law Center has apparently added his voice to the chorus of those claiming that "Microsoft violated the GPL!" What's disturbing is that Bradley certainly knows better.

Stephen Hemminger, the Vyatta engineer who initially turned up the issue (and no one is claiming that there was no potential compliance issue with these drivers) denies that there was a violation. Microsoft has likewise denied that there was a violation (but, of, course, Microsoft is Satan Incarnate...)

Just to remind the folks who should know better (and enlighten the folks who have not actually read the GPL), a violation will occur under the following set of circumstances:
  1. Microsoft distributes a binary copy of a "program" (in this case, the Hyper-V drivers), which includes GPL-licensed code, to a third-party, sources not included.
  2. The third party makes a request to Microsoft for the sources corresponding to the binary they received.
  3. Microsoft fails to provide those sources, in a timely fashion, for no more than the cost of creating the media and shipping it to the requester.
Absent that specific chain of events, there's simply no violation. If Bradley doesn't use hasn't received a copy of the Hyper-V drivers, from Microsoft, through legitimate channels, he's got no standing to be requesting sources, and he can't assert a violation. However, that doesn't inhibit the untruthful propagandists from using Bradley's (unfounded) claims to support their own (inaccurate) assertions of perfidy on Microsoft's part.

The SFLC has absolutely no business making such a claim unless they're prepared to support it, and the only way they can do so is by producing the third-party mentioned above, something which has so far not happened. The SFLC should issue a clarification immediately: they've called Microsoft liars, which is about par for the course, but they've also called Stephen Hemminger a liar, which is completely unreasonable.

A FURTHER CLARIFICATION FOR THOSE WHO DON'T
READ COMMENTS BEFORE POSTING THEIR OWN

Microsoft could have failed to comply by either the sequence of events noted above, or by failing to include an appropriate notice of how to get sources. There are assertions (and I don't personally know for a fact that there isn't an appropriate offer buried in the wad of stuff one gets with a Windows Server 200 license, I don't have one) that no such offer was tendered by Microsoft.

Assuming, for the sake of the discussion, that this was indeed the case, then the copyright holder of the code being infringed (so far unidentified) could indeed have cause to complain about a violation—since the copyright holder is the only one in a position to actually do anything about it—but it would be difficult for him-or-her to start complaining after the sources had been published and submitted, and the code was brought into compliance with the license's terms.

Which seems to be exactly what Bradley's doing, albeit at second hand. Usually this sort of rhetoric gets saved for when the SFLC files an actual suit, but it's being dragged out here when such action was avoided, thanks to Microsoft's deciding in this instance to play by the rules. That the SFLC is doing this strikes me as either grandstanding or ambulance-chasing after the case has already been settled out of court. My comments about "sore winners" still stand.

AND SOME NEW INFORMATION,
CASTING FURTHER DOUBT ON THE WHOLE THING


directhex has pointed out that what Microsoft evidently did in the Hyper-V drivers is to have a GPL-licensed driver shim, which pulled in a hitherto-proprietary binary blob at boot time, no different than what Nvidia and ATI do with their drivers. About which no one is screaming "Violation! Violation!", even though neither Nvidia nor ATI appear to have corrected those situations...

Which of these things is not like the others...? This seems to be some sort of a trick question...

41 comments:

Alex Converse said...

"If Bradley doesn't use the Hyper-V drivers, he's got no standing to be requesting sources, and he can't assert a violation."

That's completely ridiculous and (should) you know it. The GPL isn't a use license, it's a distribution license. Bradley simply has to receive the drivers from Microsoft. He never needs to use them.

Unknown said...

I went to the Wikipedia to learn more about Bradley Kuhn, and it turns out that he pretty much wrote an ode to himself on his wiki page, in violation of the Wikipedia guidelines:

http://en.wikipedia.org/w/index.php?title=Bradley_M._Kuhn&action=history

It went from being an empty article to trumpeting every minor detail f his work: heavy on the opinions, and light on the facts.

Very much like this new statement.

Lefty said...

There's neither any evidence, nor any claim, that Bradley received anything at all, nor—if he had in fact received them—that he contacted Microsoft to obtain the sources and failed to get the sources as requested.

So, what's the basis for his claim?

Alex Converse said...

I'm not saying he has standing. I'm not saying it was distributed to him. What I am saying is that you are confusing the issue by bringing use into the equation at all. It's not about use; it's all about distribution.

Lefty said...

What I am saying is that you are confusing the issue by bringing use into the equation at all. It's not about use; it's all about distribution.

Considering that I speak explicitly about "distribution", I might be led to suspect that I'm simply confusing you. However, to spare you the trouble of feeling the need to comment further, I've amended the text.

Lefty said...

Well, writing one's own page on WIkipedia may be a bit unusual, and it would seem to be against "the rules", as I understand them, anyway; it doesn't have anything in particular to do with this article, however.

ocrete said...

Lefty:

Actually, the violation happens at item number 1 of your supposed case. If you read the GPLv2 carefully, you'll see that you either have to provide the corresponding source code with the binary or a written offer, valid for 3 years.. If you fail to do one of these, you're in violation. There is no need for a 3rd party to request anything for the violation to happen.

Anonymous said...

Assuming you are right (and we have to dimostrate this...), it is quite inopportune to justify Microsoft to play the "working to rule" game: they do just what is strictly required by law, ignoring what is required by common-sense.
They recently did so for ODF implementation, and by your words they did so for Hyper-V code (without mentioniong all other past episodes).
May this be considered symptom of opensource acceptance, or just provocation?

Lefty said...

And, in the absence of a specific complaint from someone involved, that constitutes a compliance issue, rather than a "violation".

Certainly, the folks at gpl-violations.org give companies that have neglected to include a notice—and that's a pretty common error—an opportunity to correct before they claim a "violation". Ask 'em.

Harald Welte recently posted on his blog that, having gotten hold of a Palm Pre, he found that the notice was inadequate, and the sources were not available for the better part of a week.

Harald didn't start shrieking about Palm being in violation of the GPL, nor did the SFLC send out Bradley to start making similar claims (in the absence of his even having a Pre) to the media.

Microsoft is being singled out to be the recipient of (unfair) "special treatment" here, it seems.

Why would companies be motivated to "do the right thing" if they get raked over the coals for it anyway? Matt Asay's suggested that it's this sort of thing that discourages many companies from getting involved in open source at all.

Lefty said...

It is quite inopportune to justify Microsoft to play the "working to rule" game: they do just what is strictly required by law, ignoring what is required by common-sense.

"Common-sense" has nothing to do with anything. If Microsoft has complied with the license's terms, they're in compliance, and they've "done the correct thing", period.

They don't have to sing the "Free Software Song" (thank ghods), and they don't have to do the little dance.

(And while my pointing this out is probably "inopportune" for the folks who want to be screaming "violation", I don't think that word means what you seem to think it does.)

May this be considered symptom of opensource acceptance, or just provocation?

I've never claimed it was the former, and I'd certainly never claim it was the latter.

What I have said is that it's concrete evidence that Microsoft has been convinced that if they want to play in our sandbox, they're going to have to play by our rules.

Joe said...

Sorry, Lefty, but your understanding of the GPL (v2) is incorrect. A violation of the GPL occurs if someone distributes a binary copy of a program and, at the time that the binary copy is distributed, there is no source provided and there is no written offer to provide source.

You appear to believe that the violation doesn't occur until someone asks for source and source is not provided in a timely fashion. This is incorrect. A distribution of a binary either complies with the GPL or it does not.

People mess this up all the time. For example, if you provide an apt or RPM archive that doesn't contain source packages, or instructions on how to get source, congrats, you're a GPL violator. Don't believe me? Go read the GPL again.

In practice, the copyright holders usually forgive these innocent violations. But the point is that they don't have to.

Joe Buck

pridkett said...

This is kinda typical of the SFLC. However, the problem is that it tends to work. It gets people riled up, and those people in turn continue to support the SFLC. I sometimes think of the SFLC as the Fox News of the Open Source world -- sometimes making outlandish statements to get their base to react.

Also, I find it funny how the GPL relies on the acceptance of laws and software licenses in the United States. However, Bradley Kuhn openly talks about how he used to make money playing poker online and how he rips DVDs and removes other DRM from media (both mentioned at Ohio LinuxFest 2007, I believe).

The fact is that if we want Open Source to be taken seriously, we need to make sure that the people that are out there making noise about Open Source are willing to do more than just make noise -- they need to be willing to show that they're respecting the law even if it is inconvenient for them. I highly doubt that the GPL is convenient for Microsoft.

Lefty said...

In practice, the copyright holders usually forgive these innocent violations. But the point is that they don't have to.

My point precisely: in the absence of the chain of events I described, and in the absence of a claim to the contrary having been made by the copyright holders, the most that can be claimed is that "there was a compliance issue which has been corrected".

I don't see Bradley going out and unilaterally and publicly claiming "violations" on the part of the numerous organizations which are "messing up all the time" on compliance issues. They get corrected, as this one did.

Why are we engaged in "special treatment" and "extraordinary hand-wringing" here? Only because it's Microsoft, evidently. It seems as though it's simply an effort to exclude and marginalize, just as Linus said.

I expect that from Roy and the trolls, but not from the SFLC.

Anonymous said...

Lefty,

I started to write a quick note here, but it turned into a longer post on my blog.

Best regards,
James

Lefty said...

@James: A good set of comments, but as I noted there, Bradley (not the copyright holder of the allegedly-infringed-upon code, nor apparently someone who received a binary, requested sources, and failed to get them) still doesn't have standing to claim a "violation". At best, he can say that there was apparently a compliance issue that has been fully (and "eagerly", if we're to take Sam Ramji's word on things) corrected.

We shouldn't reacting to an instance of actual good behavior on Microsoft's part by going out of our way to retroactively punish them for "bad behavior" that arguably didn't actually take place.

Harald Welte of gpl-violations.org doesn't do that, and for good reason: it's counter-productive, and certainly helps foster the impression among companies that might be contemplating involvement in FLOSS development that it might be more trouble than it's worth to them.

As I've said, we didn't have a public outcry over Palm's failure to provide GPL-ed sources for the Pre right away; why are we singling out Microsoft for such special attention, if not for the very reasons that Linus called out as being "exclusionary"?

Microsoft did the Right Thing here: they recognized and admitted that, if they want to play in our sandbox (and they do, they like selling those Windows Server 2008 licenses), they have to to it according to our rules.

Stuff like Bradley's support for the unfounded—in my view—claims of "violation" make us look like a bunch of "sore winners".

Tom said...

You are just defining the word violation on your terms. That is just subjective.
IMHO if you distribute something that does not comply with the license you are violating it and I think most people feel the same way.

The fact remains that MS _did_not_ comply with the license. No way you can twist that.

Verofakto said...

I don't presume to know enough about the legal angles to this, but I will note that it's desirable/convenient for the SLFC and/or the FSF and everyone else to coat Microsoft's GPL release in a veneer of dark brown muck.

There's a significant difference between "they released it and it was accepted" and "they had no choice, they were in violation of X".

Of course it's not appropriate to bang on the awesome drum and proclaim this to be the best thing that has ever happened to Linux, as some of the tech rags did.

I think this goes back to my theory that these people need Microsoft to be a bumbling, incompetent entity (except when they're conveniently involved in mind-boggling draconian conspiracies) that can do no good whatsoever. Things like Mono and releasing code that is incorporated into the Linux kernel undermines that carefully cultivated projection.

Lefty said...

@Tom, and you (and Bradley and others) are calling "violation!" when you have neither clear cause nor the standing to do so. And you've only started doing it after the violation was corrected.

The fact remains that MS _did_not_ comply with the license. No way you can twist that.

I've never said otherwise, there were clearly compliance issues, and Microsoft fixed them. That's a Good Thing, and complaining about how they "violated" the license at this juncture (especially when you weren't the one who suffered a "violation") isn't reasonable.

It seems like just more pathological "Microsoft hatred", since I can't recall any other companies which corrected their own compliance issues in a timely and responsible way being subjected to such treatment. Especially not from the SFLC.

Feel free to correct me if you're aware of counter-examples.

KimTjik said...

My point precisely: in the absence of the chain of events I described, and in the absence of a claim to the contrary having been made by the copyright holders, the most that can be claimed is that "there was a compliance issue which has been corrected".

No big deal really, but judicially such a justification would never stand ground if put to the test.

Licenses don't take into account what the author of the code claim. The only way to bend a license to mean something else would be to write a new license. I don't see however any real basis for attacking Microsoft in this matter. Red Hat's response seems more mature: we're happy to receive this contribution of code, and will keep on the watch (my free wording of some statements).

Pragmatically I would agree that the proper way of dealing with these issues is to always give a contributor the chance to immediately correct violations, or a lack of full compliance, of a license, and only if he/she/it refuses make much fuss about it.

Licenses tend to be a complex matter, why companies are advised to get competent help to make sure it does comply.

Lefty said...

Licenses tend to be a complex matter, why companies are advised to get competent help to make sure it does comply.

I'd add that irresponsible hysteria in the face of an act of actual compliance can only raise fear, uncertainty and doubt in the thinking of organizations that were contemplating involvement in this "free software thing", but don't understand what's involved in full.

If they thought they understood it, this circus would not increase their confidence.

KimTjik said...

If they thought they understood it, this circus would not increase their confidence.

Could be. There's probably much that could be done to make it easier for those interested in contributing software. Fortunately there's not just one license available.

On the other hand the chance of sorting out possible issues without going bankrupt or being nailed to an unfair settlement is better on this side of the fence.

I doubt there's any safe ground in this business.

Lefty said...

Fortunately there's not just one license available.

That would seem to depend, to a degree at least, on whom you ask...

Unknown said...

This statement means there was a problem: "blkvcs.i386 and its seven peers are derived from GPL code, or GPL licensed themselves, but not so source has been offered"

This statement means it's all a load of bollocks, and is EXACTLY what companies like NVIDIA, ATI, Highpoint, Promise etc do when they provide .o files to be statically linked against source to produce a kernel module: "blkvcs.i686 and its seven peers are proprietary and independent chunks of code"

Pick your statement. Before choosing, consider that blkvcs.i686 lives in a subfolder of a folder named "closed"

The big big question here, which NOBODY supposedly linked to the story has revealed, is "what is the difference between nv-kernel.o, libfglrx_ip.a.GCC4, him_rr2310pm.o, and blkvsc.i686?"

Joe said...

Lefty, I don't need standing to point out that the terms of the GPL were violated. The terms of the GPL are clear. You are spreading bad information.

Only a copyright holder can sue the infringer for copyright violation. You are confusing these two issues. The SFLC has considerable expertise and a hell of a lot of experience in GPL enforcement, and they win every time they take a violator on. They have a track record on these matters; you do not.

GPLv2 says that the license is automatically terminated when it is violated. The FSF and the SFLC have used this in the past to force concessions out of companies, as have the BusyBox developers. If you mess this up, the copyright holder can legally demand that you take your product off the market, and this was the threat that forced the manufacturers of router boxes to pay actual cash to the BusyBox folks.

Just releasing source, after you've been forced to, does not suffice.

Interestingly, GPLv3 is less strict than GPLv2 on this point: GPLv2 says a violation instantly terminates the license, while GPLv3 allows time to correct the violation.

Joe Buck

Lefty said...

Lefty, I don't need standing to point out that the terms of the GPL were violated. The terms of the GPL are clear. You are spreading bad information.

No, but if you're planning on doing anything other than whine about it, you do.

Also, you may want to read the addendum at the very end of the post. There's increasing reason to doubt that there was ever a violation at all, or at least that's a lot less clear, based on what Jo's turned up. It makes claims of certainty look even less responsible, I think it's safe to say.

Unknown said...

Lefty,

Think about what you are saying. The GPL v.2:
"4. You may not copy, modify, sublicense, or distribute the Program
except as expressly provided under this License. Any attempt
otherwise to copy, modify, sublicense or distribute the Program is
void, and will automatically terminate your rights under this License.
However, parties who have received copies, or rights, from you under
this License will not have their licenses terminated so long as such
parties remain in full compliance.

5. You are not required to accept this License, since you have not
signed it. However, nothing else grants you permission to modify or
distribute the Program or its derivative works. These actions are
prohibited by law if you do not accept this License. Therefore, by
modifying or distributing the Program (or any work based on the
Program), you indicate your acceptance of this License to do so, and
all its terms and conditions for copying, distributing or modifying
the Program or works based on it.
"

Microsoft distributes code linking to GPLed code (thus making the MS code a derivative) and doesn't offer source code. "Any attempt
otherwise to copy, modify, sublicense or distribute the Program is
void, and will automatically terminate your rights under this License." It doesn't say someone with standing needs to complain before it's terminated. It's just terminated. MS is then in violation by continuing to distribute code for which they don't have a license.

I don't need to have legal standing to make a statement such as this. If I want to sue MS, then sure -- I need legal standing. I can state that Hans Reiser is guilty of second-degree murder even though I don't have legal standing to punish him for it.

I don't think MS tried to get away with anything. I don't think the people there were malicious. I also don't see a huge stink from the community over the code release. There was a lot of "WTF?" going on when the code was first released, and I see the "MS released code to comply with GPL" stories as correction to the original release.

Stop being so emotional on this one.

Dan

Lefty said...

Dan, I'll stop "being so emotional" when you read blog postings all the way through before you comment on them.

You missed a couple of important points: the "special treatment" being afforded to Microsoft by the SFLC: we haven't seen them giving press interviews about how a company was "in violation" (an assertion about which there is some doubt, as you'd know if you'd read the whole article) after they've corrected the "violation".

Moreover, as Jo Shields has pointed out, Microsoft did nothing, apparently, that Nvidia, ATI, Promise and others aren't still doing today: they had a GPL shim linked to a proprietary binary blob, and there's no evidence that the binary blob contained even a single line of GPL code.

So why isn't Bradley going after Nvidia, ATI and others who are, if you subscribe to his view, still in violation?

Anonymous said...

Hi Lefty,

> Harald didn't start shrieking about Palm being in violation of the GPL

I don't get it. The first paragraph of Harald's blog post on his discovery was "As it has been reported at many places online, the Palm Pre has started to ship as a CDMA model in the United States. However, as it seems, at this time it is not GPL compliant and thus a copyright infringement!"

I don't know what you mean by shrieking, so I can't tell whether this was it, but you'd expect that making a post on your very public blog that announced that Palm was violating the GPL and ending in an exclamation mark would suffice. Why is the statement about Microsoft very different?

Lefty said...

Why is the statement about Microsoft very different?

Several reasons. First, it comes from the SFLC, in a statement made to the media, rather than in a personal blog entry.

Second, Harald sensibly doesn't use the loaded term "violation", given that Palm—now aware of the the issue—hadn't had time to respond.

Third, Harald statements were made while Palm was out of compliance. Bradley's were made after Microsoft had complied.

Like I said, this smacks of grandstanding.

Given Jo's discoveries about the structure of the drivers (not apparently different than the ATI drivers, or the full-function Nvidia drivers for Linux, or many other examples), it begins to look like irresponsible grandstanding on the SFLC's part.

In short, Bradley seems to be playing directly to the "exclusionary" Microsoft-loathing "extremists" that Linus talked about, and without Bradley's even being in full possession of the facts.

Anonymous said...

Ah, okay.

> Given Jo's discoveries about the structure of the drivers (not apparently different than the ATI drivers, or the full-function Nvidia drivers for Linux, or many other examples), it begins to look like irresponsible grandstanding on the SFLC's part.

Just FYI, many kernel developers do consider those drivers illegal, not least Greg K-H, who was involved in this case and often speaks publicly about his belief that any binary drivers are illegal:

http://broadcast.oreilly.com/2008/10/how-linux-supports-more-device.html

So, I think the answer to "why haven't ATI and nVidia been sued yet?" is more about "Linus doesn't believe in suing people, and his opinion holds enough sway to stop the other important contributors from doing so as well" than "Because binary-only drivers are not illegal".

> there's no evidence that the binary blob contained even a single line of GPL code.

If a copy of the binary blob were made available, I'm pretty sure we could find it. It would be fairly impossible to write a Linux driver like this one that doesn't derive from Linux. For comparison, we can look at the source drop they've made now, and decide whether we think it's derived from Linux.

Anonymous said...

Lefty,

Please note that Steve Hemminger said publicly on his blog that he discovered Microsoft to be in violation. I have worked with Steve and know him to be honorable and understand the basics of the GPL. If he believes he discovered a violation, I believe him. Also Greg K-H says he also "pointed out the obviousness" of the requirement that Microsoft needed to GPLv2 the code.

I have a full blog post on the topic as well. The original article you link to made slightly exaggerated conclusions about my comments. He quotes me directly later, and those quotes are accurate.

-- bkuhn (using the only OpenID I had handy that Google will accept)

Anonymous said...

Also, I don't think it's true at all that the driver they've submitted is a shim. It doesn't link to anything else! The hypervisor runs the kernel containing their open-source driver, and that guest kernel's address space is unmodified. (This is not at all similar to ATI/nVidia.)

Lefty said...

So, I think the answer to "why haven't ATI and nVidia been sued yet?" is more about "Linus doesn't believe in suing people, and his opinion holds enough sway to stop the other important contributors from doing so as well" than "Because binary-only drivers are not illegal".

Which brings us directly back to the core of my original concern: why is Microsoft being singled out by the SFLC as a "violator" after they've complied, when we apparently have other major organizations in ongoing, and apparently identical, situations?

For extra credit, does this sort of thing encourage or discourage companies shipping the sort of "illegal" drivers you mention to get into compliance?

Lefty said...

The original article you link to made slightly exaggerated conclusions about my comments.

Uh huh.

Again, Bradley, the questions I'm really curious about—and they're direct and specific and I'd enjoy equally direct and specific responses here—are:

First, why are we dragging out this rhetoric—more typically used when a suit is actually brought against a recalcitrant company which has steadfastly refused to come into compliance after friendly notes, warnings, shots across the bow, etc.—in response to a company's compliance?

Second, if the issue was—as it seems to be—a case of a GPL-licensed shim and a "binary blob", why aren't you talking about all of the other companies that are equally "violators" and aren't doing anything to comply, thus remaining in a state on ongoing "violation"?

Third, do you think your statements encourage or discourage other companies in similar "violation" right now as far as their future compliance efforts might be concerned?

KimTjik said...

I'll just add a piece of information that everyone running Nvidia proprietary drivers would see while booting without a splash screen, or since it goes to fast the output of dmesg | grep nvidia:

nvidia: module license 'NVIDIA' taints kernel.

You'll see that every time you boot a computer having that module installed. If the Microsoft driver is clean, as I understand it is, it won't be "harassed" by anything like it.

From this perspective Nvidia has always got a lot of criticism and the community continues to remind Nvidia about their pretty lame contribution to free drivers. Hence I don't think there's any valid argument for saying that Microsoft is worse treated.

Lefty said...

Hence I don't think there's any valid argument for saying that Microsoft is worse treated.

No?

Has the SFLC been making statements to the press about Nvidia's "violations" and I missed them...?

Unknown said...

If a copy of the binary blob were made available, I'm pretty sure we could find it.

http://www.microsoft.com/downloads/details.aspx?FamilyID=AB7F4983-93C5-4A70-8C79-0642F0D59EC2&displaylang=en

Go nuts. .exe can be extracted by file-roller.

Anonymous said...

I think Lefty's right -- Microsoft's being treated differently than we treat nVidia and AMD, and it's not obvious why. One potential answer is that the nVidia/AMD situation with binary modules has been unchanged since before most of us started using Linux in the first place, and people aren't banging on our doors asking us to explain what's going on; it's already well-understood.

Conversely, this Microsoft situation has been full of surprises, and it doesn't seem unlikely to me that the SFLC's release is in response to a barrage of press inquiries of a kind they haven't received before about other violations. There was a strong "hell freezes over" element to this submission, as stated by Greg K-H.

I think it would have been good for the SFLC to mention that Microsoft isn't the only company to have a proprietary Linux module, to avoid the appearance of jumping in at this story -- of an already-fixed violation -- just because Microsoft is involved, as Lefty is complaining about.

Ed Ropple said...

I think Lefty's right -- Microsoft's being treated differently than we treat nVidia and AMD, and it's not obvious why.

I disagree. I think it's pretty obvious why: it's Microsoft (dum dum dummmmm). As has been noted on multiple open source blogs as of late, this behavior is not new, and certain people are prone to overreaction based on the parties involved (such as Microsoft).

I think it would have been good for the SFLC to mention that Microsoft isn't the only company to have a proprietary Linux module, to avoid the appearance of jumping in at this story -- of an already-fixed violation -- just because Microsoft is involved, as Lefty is complaining about.

This is a good idea, and makes perfect sense if you don't think that they're jumping in just because Microsoft is involved. Personally I'm not so sure about that (especially the more I read about Mr. Kuhn).

TGM said...

Well let's take that last question and break it down. nVidia/ATI creates a driver and decides they don't want their trade secrets leaked. Not a good siituation for us, but at lease they've gave us a driver.

Then we have Microsoft. Code full of back doors, inherently insecure (look at IE4,5,6,7 - I'll give them a break with IE3) and they want to put code into our kernel we can't debug?? I'm glad they've released the source, and we can all get along together :)

Hmm... Actually I think I will moan about nVidia seeing that (apparently) 29% of bugs are down to their drivers...

http://www.downloadsquad.com/2008/03/28/29-of-windows-vista-crashes-caused-by-nvidia-drivers/2

Lefty said...

Then we have Microsoft. Code full of back doors, inherently insecure (look at IE4,5,6,7 - I'll give them a break with IE3) and they want to put code into our kernel we can't debug??

Well, to be accurate, they put code into their customers' kernels that their customers couldn't debug. Presumably no one was holding a gun to the customers' heads.

I'm glad they've released the source, and we can all get along together :)

I think that's a healthy reaction. No one, least of all me, is suggesting that Microsoft is suddenly everyone's best friend, but we should also recognize that they recognized that they need to be playing by our rules here. That's a significant change from the past, and to focus on the degree that they used to be in technical "violation" (i.e. compliance issues), as opposed to the apparently-equal degree that Nvidia and ATI, among others, still are, seems unreasonable and unhelpful to me.

Hmm... Actually I think I will moan about nVidia seeing that (apparently) 29% of bugs are down to their drivers...

That's fine, but I'm wondering where the folks at the SFLC have been all this time. Will they wait until some future day when Nvidia releases their full drivers and then tell us about how they were violating all this time...?

As Bill Cosby said, "That's like putting the 'Dip' sign six feet after you hit the dip."