LWN: Comments on "Hall: IBM, Red Hat and Free Software: An old maddog’s view"

Hall: IBM, Red Hat and Free Software: An old maddog’s view

IanKelling — Wed, 01 Nov 2023 04:56:42 +0000

> But it's a punch in my face, as an upstream developer. I released the software such as that anybody (including Red Hat and the end customer) has the written right to redistribute without further restrictions.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

pizza — Sat, 12 Aug 2023 23:37:28 +0000

> It does make me wonder if the GPL will have further clarification from the FSF in the future now that Red Hat’s subscription agreement problems are more publicized.

RH's subscription terms have been essentially unchanged for over twenty years, going all the way back to the RHEL 2 release in 2002, a full five years before the GPLv3 was released. Furthermore, this approach is the same as what Cygnus did with GCC (ie an actual GNU project!) starting in 1989, about two years before the GPLv2 was released!

If anything, the FSF has _strengthened_ the language permitting what Red Hat is doing -- They make it clear that GPL rights only apply to the specific release the user has in their hands and there is no guarantee of support or any future updates [1]. Additionally, it explicitly states that support and future updates can be withdrawn if the user exercises their rights to source and "installation instructions" under the GPL.

So I have a very hard time seeing how the FSF could address what RH is doing without also effectively requiring everyone to provide a warranty (and thus updates/support) for free. Which simply isn't going to happen.

[1] although support/updates may be provided for a fee [2]
[2] which necessitates an additional agreement/contract stating what limits/terms of support, the actual fees, and so forth.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

sinuhe — Sat, 12 Aug 2023 21:27:10 +0000

I think this has merit in that the subscription agreement seems disingenuous of Red Hat. It is also skirting the legality of copyleft licenses like the GPL, and threatening (and sometimes actually) auditing the subscription compliance, including internal use of the source code, not mere redistribution publicly with trademarks removed. The publishing of public sources seems to me as a way of more openly complying with copyleft, especially the GPL. See this article which I think summarizes the problems well that maddog doesn’t hit on: https://sfconservancy.org/blog/2023/jun/23/rhel-gpl-analy...

Company lawyers want to keep their company in compliance to avoid the cost of legal issues, audits, and so on, and in general are used to proprietary licenses, not open source EULAs, including trademark, so they typically won’t distinguish the nuance. It does make me wonder if the GPL will have further clarification from the FSF in the future now that Red Hat’s subscription agreement problems are more publicized. Remember, not all licenses are copyleft in RHEL, but there is a substantial base with the GPL.

Finally, it’s curious there’s not been word from esr on this. He fell out with Red Hat awhile ago, yet has espoused abandoning copyleft. It presents an interesting conversation if he were to blog again.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

Wol — Fri, 04 Aug 2023 08:02:53 +0000

> If Red Hat refunds you 11/12 of $1000 for the unused part of the support contract

> I don't use Red Hat and have no idea what its policy is in this regard, but I'd be curious to find out.

Legally, I personally think that it's extremely clear. YOU broke the terms of the contract (that is, the support contract, not the implied licence contract). As such, YOU gave Red Hat the right to walk away. Which means they have no obligation to do anything, let alone refund the money.

As has been repeatedly said, this is all about 3rd parties reselling Red Hat's work, without paying Red Hat. Regardless of the legalities (which I think are pretty clear), the politest term for this behaviour is "ripping off".

True, I'm quite happy to take advantage of other peoples' generosity. But I also try and "pay forward". And I don't get all "high and mighty" when or if they decide it's not worth the candle and no longer want to be so generous.

Cheers,
Wol

Hall: IBM, Red Hat and Free Software: An old maddog’s view

eduperez — Fri, 04 Aug 2023 05:52:56 +0000

On one hand, the customer signs a support contract, and a series of conditions that terminate such contract. On the other hand, the customer receives the binaries and their sources, according to the GPL license, with no strings attached. There are no restrictions on the rights granted by the GPL, there are some conditions to terminate the support contract.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

bconoboy — Fri, 04 Aug 2023 05:27:26 +0000

I suspect you're misremembering what the deal with Cygwin was. It was licensed under the GPL, rather than the LGPL. That had consequences for people who wanted to write proprietary software that linked against the cygwin1.dll. Cygnus, then Red Hat, maintained complete copyright control of the Cygwin source base, allowing the sale of a second license that let people keep their source proprietary.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

NYKevin — Fri, 04 Aug 2023 01:33:11 +0000

> John Deere doesn't sell it's devices below cost, now, does it? Yet it locks them up. And the same thing is done by printer manufacturers (although in that case some models are sold below cost) and makers of anything complex enough to be locked up.

John Deere has a business model where they charge you an arm and a leg for first-party repairs. Printer manufacturers have a business model where they pretend that their ink is more valuable than unicorn blood. Both of those business models are defeated by user tampering.

> Even if there are no actual, full-blown, crypto-key-enabled lock down there are no incentive to make things compatible with one, single, Linux image (look on these endless attempt to develop a way to create one, single, Linux kernel image for ARM devices).

Do you mean at the time, or now? Because now, if you try to sell a computer that doesn't run Linux, you'll find that the datacenter side of the equation is wholly uninterested in dealing with you, so you'll have to retail them to individual consumers, which is much more of a PITA than selling them in bulk directly to FAANG or whoever (not to mention, you probably get lower margins on retail than B2B).

At the time, of course, datacenters were less of a "commodity hardware" thing and more of a "nobody gets fired for buying DEC/IBM/what-have-you" thing. Nevertheless, I tend to imagine that, in a B2B context, you're inevitably going to have startups eyeing the cheap end of the market, and asking questions about exactly how much you get for paying the IBM tax. Under a balkanized hypothetical, you'll have companies picking the cheap arch, or even companies trying to use multiple arches and figuring "hey, they all speak TCP/IP, right?" Inevitably, the inefficiencies here will drive the more expensive arches to become more niche and specialized, and the cheap arches towards the mainstream. That's just the invisible hand of capitalism.

> (witness development of Androd or smart TVs).

The difference is, nobody is trying to build a datacenter out of phones or TV sets. Some people are buying Android in bulk for their employees, but they save money with BYOD rather than by buying cheap devices. So there's much less economic pressure on the inefficiency, and it persists.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

farnz — Fri, 04 Aug 2023 00:51:43 +0000

If you exercise your rights under the GPL, Red Hat may well still let you keep the support contract - the relevant behaviour is not "exercising your rights under the GPL", but "giving a third party the benefit of the Red Hat services". So, for example, if you give me a copy of the Red Hat version of the kernel, so that I can debug a kernel issue for you, no breach has taken place, and Red Hat won't terminate. Or, for another example, if I'm paying for the original author to support rp-pppoe for me, and I send them a copy of Red Hat's source and binaries for rp-pppoe so that they can apply the fix to the RH version for me, that's also going to be fine with Red Hat - I'm not giving away the benefit of my RH contract to a third party, I'm paying a third party to give me more benefits than RH does (presumably because the original author of rp-pppoe is better at supporting my use case than RH are).

If you do give a third party the benefit of the RH services, then you are in breach of contract, and you lose your support contract. In such a circumstance, Red Hat's terms section 3.1 says:

All Fees, expenses and other amounts paid under the Agreement are non-refundable. The Software Subscription Fees are for Services; there are no Fees associated with the Software licenses

Hall: IBM, Red Hat and Free Software: An old maddog’s view

anselm — Fri, 04 Aug 2023 00:37:57 +0000

Suppose you pay $1000 for a one-year support contract, and then a month later exercise your rights under the GPL.

You can “exercise your rights under the GPL” with the GPL code in your copy of RHEL until the cows come home – you can modify it, distribute it in original or modified form, etc. as specified in the GPL –, and this applies equally no matter whether you have an active paid-for RHEL subscription, whether Red Hat has cancelled it, or whether you just let it expire. After all, that's what the GPL says, and we can reasonably assume that Red Hat's lawyers know the GPL very well, and that Red Hat doesn't want to be sued by the copyright owners of GPL code in RHEL.

The question as far as Red Hat is concerned, however, isn't whether you “exercise your rights under the GPL”, it's whether you abuse your support contract with Red Hat by giving someone else the benefit of a RHEL subscription (e.g., by feeding them the RHEL updates you get from Red Hat so they don't need to buy their own subscription). In that case Red Hat may decide that they do not want to give you further updates under that support contract, which is fair enough – presumably they like being ripped off as little as any other company. But this is in no way, shape, or form a GPL issue because your “rights under the GPL” are not touched by this at all; getting updates to code is expressly not a GPL right.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

dskoll — Thu, 03 Aug 2023 23:46:46 +0000

IMO, a lot hinges on what exactly Red Hat does. Suppose you pay $1000 for a one-year support contract, and then a month later exercise your rights under the GPL. If Red Hat refunds you 11/12 of $1000 for the unused part of the support contract, then I agree it's not really an additional restriction on the GPL. However, if they cancel your contract and don't refund the unused amount, then I don't think it's clear that it isn't an additional restriction.

I don't use Red Hat and have no idea what its policy is in this regard, but I'd be curious to find out.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

rgmoore — Thu, 03 Aug 2023 21:03:39 +0000

Wonder what made them change gears, they tried to avoid distributing anything GPL for years.

Money. They could see the big revenue growth was going to come from cloud computing, and a lot of potential customers thought "cloud" meant Linux. More generally, Microsoft's desire to avoid distributing anything GPL was more about trying to FUD the concept of FOSS than any real legal difficulty. Once it was obvious the FUD hadn't worked and there was real money to be made in the FOSS space, their reluctance went away.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

farnz — Thu, 03 Aug 2023 18:24:35 +0000

How is it in exchange? You do not lose your GPL rights by taking up RH's support contract; you retain them, even if you later lose your RH support contract for breaching the terms of that contract.

Further, exercising your GPL rights does not automatically terminate your RH support contract; the specific clause for termination involves you using your support contract to give a third party the benefit of an RH subscription. Thus, you can distribute code using your GPL rights without losing your RH support subscription. If it was an exchange, you would not be able to do that at all - and yet you can, as long as (in RH's view) you're not giving a third party the benefits of your support subscription.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

mb — Thu, 03 Aug 2023 17:48:30 +0000

>it's something Red Hat offer as an addition on top of your rights.

Nope. It's an offering in exchange for the customer's GPL rights. You can't have both at the same time. That's why it is a restriction.

(This is not legal advise)

Hall: IBM, Red Hat and Free Software: An old maddog’s view

farnz — Thu, 03 Aug 2023 17:13:06 +0000

Right, but the customer has no right to support. They have a right to redistribute the sources, granted by the GPL, but their support is contingent on Red Hat's goodwill. And even if Red Hat stop support, they still have the sources, and can still distribute them.

The whole reason it's not an additional restriction is that what you lose is not something you have a right to under either copyright law or the GPL terms - it's something Red Hat offer as an addition on top of your rights.

To give you a fair analogy, if I said I'd run naked through the city centre every time you redistribute, that's an additional right granted to you. If I then say "but I don't like the way you're redistributing, I'm no longer going to run naked", that's not an additional restriction - that's removing something that I added on top of the GPL.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

farnz — Thu, 03 Aug 2023 17:09:35 +0000

You're describing a different situation.

In the Red Hat situation, you can distribute the software as much as you like; the worst that happens is that Red Hat stops doing something you want them to do.

In your situation, you're asking the third party to do something additional to the licence if they redistribute. That's what makes it an additional restriction - there is something else that you must do to exercise the rights you have under the GPL.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

farnz — Thu, 03 Aug 2023 17:04:22 +0000

The customer can continue their business exactly as it was before the support contract ended. They don't have to have a support contract from Red Hat to continue their business - they can run the software unsupported, or provide in-house support, or pay another company (not Red Hat) to support their systems.

If the customer wants support for the OS, then they have to find a support provider who provides what they want. But that is entirely separate to the GPL - the GPL does not obligate anyone to distribute future versions to you (only matching sources for versions they have distributed to you already), and the GPL does not obligate anyone to provide support for software.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

paulj — Thu, 03 Aug 2023 16:07:02 +0000

Again, we're just applying our non-legal views here.

A legal professional would be able to say "The case of X in the FOO court/jurisdiction, with GIVEN_FACTS, the court concluded ABC". And if there are no cases where GIVEN_FACTS are similar enough, they'd say that instead and point out we need a case.

If I receive binaries from RedHat, and if - as the GPL allows - I publish those, and RedHat consequently terminate my support contract and access under terms in that contract that explicitly equate "using Subscription Services in connection with any redistribution of Software" with "material breach of the Agreement", then it would seem to me that I am indeed restricted from exercising my GPL rights if I wish to maintain my support agreement with RedHat.

I have no idea what factors matter here, and how they inter-play and balance out. It seems to me that no-one else in these comments section knows either - at least not to the extent that they can present a scholarly argument, with citations to relevant case law (??).

So it seems to me that the best thing to do here is to acknowledge that uncertainty, rather than continuing claiming we know for sure. FWIW, without some references to actual law, judgements or other detailed legal analysis, I'm going to stick to my view that no one here knows either way, to any significant certainty.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

mb — Thu, 03 Aug 2023 15:06:32 +0000

>It can't be a restriction if it doesn't restrict anything.

Ok. I will distribute your GPLed software with an added letter to a third party. I will only give the SW away, if the third party signs that letter. The words of the letter are: "You may only exercise your GPL rights, if you run naked over the city market place."
What if the person doesn't want to run naked and want to redistribute the SW?
Not a restriction?
Am I violating the license that you gave me?

If this is not a restriction, than I could add anything to the GPL as an amendment.
The "further restrictions" paragraph would be useless.

Maybe it is useless. I don't rule that possibility out.

(This is not legal advise)

Hall: IBM, Red Hat and Free Software: An old maddog’s view

anselm — Thu, 03 Aug 2023 14:16:42 +0000

Still it is a restriction.

It can't be a restriction if it doesn't restrict anything. Your rights with respect to the GPL code in the version of RHEL on your computer are exactly the same whether you have a current RHEL subscription or not.

The only right that is at issue here is the right to receive future updates from Red Hat as part of the subscription, and that is obviously Red Hat's to grant or withdraw, according to their separately-negotiated support contract with you (not the GPL).

Hall: IBM, Red Hat and Free Software: An old maddog’s view

Karellen — Thu, 03 Aug 2023 13:52:56 +0000

It's just riffing on an insult from Good Morning, Vietnam.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

paulj — Thu, 03 Aug 2023 12:49:38 +0000

I don't think it matters whether the restriction is in a separate agreement. Though, you could argue the customer freely chose to agree to restrict themselves from (some of) their GPL rights granted to them by (generally) non-RH copyright holders, in order to avail of RH's support services. Still it is a restriction.

How it balances out with the other factors is something only a court could determine (assuming we don't have clearly applicable precedent, that a lawyer could already point to).

Discussing back and forth here ("it's a restriction" -> "but it's a support contract, and GPL doesn't require support!" -> "but..") isn't really going to settle how a court would balance these different factors against each other. ;)

Hall: IBM, Red Hat and Free Software: An old maddog’s view

anselm — Thu, 03 Aug 2023 12:38:51 +0000

The issue is how that plays out with the "no further restrictions on rights" aspect of the GPL - and how that balances out with the other, counter-veiling, factors - that have been pointed out many times in this thread.

Many people don't seem to get that “no further restrictions” thing. If I give you the sources to GNU Emacs under the GPL but attach a clause saying “no distribution of this on Sundays”, that's adding a “further restriction” to the GPL (which is not allowed). OTOH, if Red Hat includes a clause in their separately-negotiated support contract saying that they reserve the right to cancel that support contract for people who they notice are ripping off RHEL, that is not a “further restriction” on the distribution of GPL code in RHEL, it's a support issue – and the GPL stipulates explicitly that receiving code under the GPL does not automatically entitle one to support or future updates.

In particular, there is nothing in the GPL that keeps you from buying a subscription to RHEL and using that to jump-start your own Linux distribution. Sure, Red Hat will be likely to cancel your (separate) RHEL subscription for the future, but even then you're still completely free to use and distribute the GPL code from RHEL that you already have, pursuant to the GPL – it's just that you won't receive any new RHEL updates from Red Hat. If the Red Hat support contract cancellation clause did, in fact, constitute a “further restriction” on your rights to distribute the GPL code in RHEL, then this would not be the case.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

paulj — Thu, 03 Aug 2023 11:12:59 +0000

I totally understand why RedHat would want to cut off customers who redistribute RH packages.

The issue is how that plays out with the "no further restrictions on rights" aspect of the GPL - and how that balances out with the other, counter-veiling, factors - that have been pointed out many times in this thread.

It'd be nice to get an authoritative view on that. It will require someone risking their money and suing one of the companies using these kinds of clauses. RedHat are generally good, and always offer sources. So, can be kind of sympathetic to them.

There are /other/ companies (in open source networking, packets and routing, say) who I am 99.9% sure are using this trick, but who do /not/ make sources routinely available, and who I strongly suspect penalise customers who exercise right to redistribute *source* (cause, the project I worked on, we /never/ got their modifications fed back to us quietly by customers). They would be /morally/ the better companies to go after. But the additional issues could muddy any opinion obtained - in so far as it might shed light on RedHat's practice.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

anselm — Thu, 03 Aug 2023 10:57:41 +0000

RH explicitly states they will react punitively - terminating the customer relationship - if a customer exercises GPL granted rights. It's not a random act of ceasing a relationship with a customer, it is _explicitly_ predicated on exercise of GPL rights.

It stands to reason that if you're a RHEL subscriber, Red Hat would rather keep you as a paying customer than not. Hence no one at Red Hat will care if you “exercise your GPL rights” with respect to one or five or two dozen GPL packages from RHEL. They will presumably notice if you break your support contract (not the GPL), e.g., by making a complete almost-RHEL distribution of your own based on RHEL and offering that to third parties, and particularly if you're planning to do so going forward, in direct competition with Red Hat itself, based on the support that you yourself are getting from Red Hat. That is what is going to cause Red Hat to void your support contract, not your giving GNU Emacs or Bash sources from RHEL to your buddy under the GPL. This is no more or less than what any other company would do under the circumstances, and it has nothing whatsoever to do with the GPL.

IOW, if you want charity, run Debian.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

paulj — Thu, 03 Aug 2023 10:49:44 +0000

Ah, it was a slightly later Alpha that /. started on, a DEC Multia:

http://www.obsolyte.com/dec/multia/
https://news.slashdot.org/story/07/10/10/1445216/a-brief-...

Note the many OSes it could run. A bit later, but the path towards this was inevitable.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

paulj — Thu, 03 Aug 2023 10:25:49 +0000

RH explicitly states they will react punitively - terminating the customer relationship - if a customer exercises GPL granted rights. It's not a random act of ceasing a relationship with a customer, it is _explicitly_ predicated on exercise of GPL rights.

The GPL doesn't require anyone to give support for distributed software, but the GPL also disallows the distributor to impose restrictions on the rights it grants.

Be interesting to see this point litigated and get an authoritative legal view on this.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

Wol — Thu, 03 Aug 2023 09:50:26 +0000

Of which SuSE is the perfect example.

YaST forbade distribution for money. It was easy (and perfectly okay) to copy a SuSE CD. You could give them away, you just couldn't sell them ...

Cheers,
Wol

Hall: IBM, Red Hat and Free Software: An old maddog’s view

paulj — Thu, 03 Aug 2023 09:45:37 +0000

You could run multiple OSes on the Amiga. Multiple ones were available, and you could write your own if you wished. Nothing would have stopped Linus hacking on any of those other machines. i386 just was the most available then. If it hadn't been an i386 PC it'd have been something else.

I mean, "Well, maybe community-made OS made specifically for this one, specific, device. If you are lucky." - you're describing Linux.

The luck here was Linus and the other hackers he attracted around Linux.

The i386 PC is an unimportant detail. *MANY* companies were racing to build affordable, commodity, capable, 32bit, MMU with paging, machines and get them out into the hands of the mass market. It was inevitable one or more of the *MANY* companies who _had long_ seen this opportunity and were _rushing_ to fulfil it would in fact _do what they were already trying to do_.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

paulj — Thu, 03 Aug 2023 09:40:35 +0000

It may not have picked one, I'll agree. And I'll agree with the other comment that we can't know what alternative histories there would have been.

One thing is certain though, when *many* companies _have already_ been working on building ever more capable, ever cheaper, ever more commodity computers; as most of society (even outside the actual people in the industry) are agreed computers are going to revolutionise things; it was _inevitable_ we were going to get *exactly the thing those many companies were already competing to provide*: A capable, affordable computer suitable for hobbyists to tinker on and write OSes that were never going to amount to anything.

tl;dr: Had Intel, IBM, Compaq - and the following army of cloners - not won dominance, Linus would have been hacking on an Amiga 3000, or an Atari ST, or an Alpha, or... etc. And we would still today have computers everywhere (from your pocket to your car to servers, to the moon) running Linux. Regardless of architecture.

And gosh, Linus *was* hacking on an Alpha by the mid-90s. ;) Maddog didn't mention he arranged that. (And the little Alpha 150 workstation DEC made was fairly cheap - almost affordable; I think Slashdot began running on one of those!).

Hall: IBM, Red Hat and Free Software: An old maddog’s view

kleptog — Thu, 03 Aug 2023 09:32:14 +0000

> Except that the GPL explicitly forbids such an additional restriction.

We've been here before. Whether what RH is doing would be considered an "additional restriction" or "a customer of RH using their freedom of contract to negotiate away some rights in exchange for something else" is entirely context dependant.

RH saying that the customer must pay $1 million would probably be considered an additional restriction.
RH saying that they may terminate the contract is almost certainly fine. After all, RH doesn't even need to give a reason to terminate a contract, they could do it because they don't like the name of your company, or because it's Tuesday.

Ain't the law fun? What counts as "additional restriction" is determined by (case) law in your jurisdiction, not the GPL. Otherwise you could argue that the fact that RH isn't paying your internet connection to upload the sources to other people is an "additional restriction".

Hall: IBM, Red Hat and Free Software: An old maddog’s view

mb — Thu, 03 Aug 2023 09:19:51 +0000

>without further restrictions on what they can do or not with those binaries and sources.

Yes, they do restrict exactly that. The customer can choose either to not distribute the sources or to distribute them and loose support. Which makes it a further restriction. It's a direct coupling.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

mb — Thu, 03 Aug 2023 09:16:41 +0000

>RH is not forcing the customer to do a damn thing.
>They are simply refusing to further interact with the customer.

Which forces the customer do react.
Or how do you think they are going to continue their business?
This is not a free decision. It is enforced by RH.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

eduperez — Thu, 03 Aug 2023 08:56:48 +0000

> The problem is that the company puts a clause into the business agreement that adds an option to terminate the agreement as a punishment for exercising GPL rights. This relation is why I think this is a further restriction of the GPL license.

The key point here is that RH is not adding any restriction over the GPL, *on the binaries or sources already distributed*. The customer receives the binaries and their sources, as stipulated by the GPL, without further restrictions on what they can do or not with those binaries and sources. On the other hand, there is a support contract, that RH can terminate at will, but that termination has no effect *on the binaries or sources already distributed*.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

khim — Thu, 03 Aug 2023 08:29:33 +0000

> It sounds like your claim is that a similar business model (sell lots of software to subsidize your underpriced hardware) would have developed for commodity computers

Nope. I'm just saying that attempt to close thing up is natural for the business. It's almost instinct. John Deere doesn't sell it's devices below cost, now, does it? Yet it locks them up. And the same thing is done by printer manufacturers (although in that case some models are sold below cost) and makers of anything complex enough to be locked up.

Even if there are no actual, full-blown, crypto-key-enabled lock down there are no incentive to make things compatible with one, single, Linux image (look on these endless attempt to develop a way to create one, single, Linux kernel image for ARM devices).

If PC market is an exception (and it is an exception) then there needs to be reason for it to behave differently from all other markets.

Part of the reason is the use of software supplied buy third party developers. But even bigger reason is OS supplied by third-party. PC is unique not because it has lots of programs (smartphones or IBM servers have lots of software, too), but because OS and hardware come from different source (OS from Microsoft, PCs from hundreds of hardware vendors). This is what's unique in a PC world and this is result of lucky accident which happened when IBM lost control over IBM PC-compatibles market.

But without Microsoft being big enough to force a single standard there would have been balkanisation and lock downs, anyway (witness development of Androd or smart TVs).

Hall: IBM, Red Hat and Free Software: An old maddog’s view

NYKevin — Thu, 03 Aug 2023 07:58:32 +0000

> 2) forces the customer to migrate to another OS and loose support for old machines -> Not a GPL violation? We disagree.

Huh? RH is not forcing the customer to do a damn thing. They are simply refusing to further interact with the customer. If the customer's old machines need support, and RH no longer wants to provide it, that's the customer's problem. It does not magically become an instance of coercion just because it *happens* to be convenient for RH to continue providing support.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

NYKevin — Thu, 03 Aug 2023 07:49:40 +0000

> Or you can call that tortious interference since there are no estoppel in copyright law...

That's not how any of this works. Estoppel is an equitable right that attaches under a wide variety of circumstances. I am skeptical that it attaches under *these* circumstances in particular (at a minimum, you would need to produce a written record of the FSF and/or GNU explicitly and unambiguously endorsing a particular interpretation of the GPL, and even then it would likely only bind those parties and not e.g. Linus), but to assert that it never applies to copyright at all goes way too far. For example, a promissory estoppel might be styled as an "implied verbal license" or something like that, but the effect is the same: The plaintiff loses their case.

Tortious interference, on the other hand, is a matter of contract law. Since most copyright licenses are contracts, including the GPL in most jurisdictions,[1] it follows that tortious interference with a copyright license is possible, but again, I'm not seeing how it applies here. Vaguetweeting about how you don't like Red Hat's interpretation of the GPL can hardly rise to the level of tortious interference. At a minimum, you'd have to explicitly solicit RH's customers to breach their support agreements, and I am not aware of the FSF (or anyone else) actually doing that.

[1]: https://lwn.net/Articles/747563/

Hall: IBM, Red Hat and Free Software: An old maddog’s view

NYKevin — Thu, 03 Aug 2023 07:33:27 +0000

> Look on what happened with game consoles. Once upon time these, too, were “an affordable, capable, available, well documented” devices. Yet… what happened to them now?

The entire business model of game consoles has (at least historically) been to sell moderately powerful computers at slightly below cost, and make up the losses in game sales. If you can't sell (enough) games because you can't (sufficiently) control the OS, then you have to raise the price of the console above the price of a comparable gaming PC (or some competing console), and then consumers notice the price is too high and jump ship. That does not apply to generic commodity computers, which are normally sold above cost, the manufacturer does not care what happens after the consumer buys one, and there is (for the most part) nowhere else for consumers to go if the price is too high (aside from other manufacturers of the same product).

It sounds like your claim is that a similar business model (sell lots of software to subsidize your underpriced hardware) would have developed for commodity computers, but I find that difficult to believe. The whole point of games is that they are, to some extent, consumable. You play a game, and then, eventually, you stop playing it and go play something else instead. You may, eventually, return to an old game, but you're still going to want to play something new every now and then, so you keep buying new games. This is not how generic productivity software works - users buy the software they want, and then it may be many years between paid updates, particularly in the time period we're talking about (i.e. long before the subscription model became standardized). You simply don't have enough of a steady income stream to pay for all the hardware.

Hall: IBM, Red Hat and Free Software: An old maddog’s view

mathstuf — Thu, 03 Aug 2023 02:13:40 +0000

> >You are not a lawyer.

> How do you know?

A lawyer would not be sitting here making broad-based statements without knowing more details about their client's situation or at least disclaiming "this is not legal advice and even if it was, I am not *your* lawyer".

Hall: IBM, Red Hat and Free Software: An old maddog’s view

khim — Wed, 02 Aug 2023 22:29:14 +0000

It's not that CD-ROM haven't included any retaliation clauses. It was commercial product and, as such, included proprietary installer and some proprietary programs. And if you copied that you were obvious copyright violator, plain and simple.

It haven't included retaliation clause for the case where you may copy only free software from it, but without installer it wasn't much useful.

IOW: it haven't needed any retaliation clauses because it just made the whole story with attempts to achieve bug-for-bug compatibility impossible.

Whether court would consider that “the same thing” as what RedHat is doing now is open question, but practically it was much easier and simpler way to achieve what RedHat is trying to achieve.

And then RedHat bought them and tried to eat their cake and have it too (pretend that they don't sell proprietary software and yet, someone, stop RHEL freeloaders).

Hall: IBM, Red Hat and Free Software: An old maddog’s view

ballombe — Wed, 02 Aug 2023 22:02:41 +0000

> Why would there be such a clause if there were no violators?

But if there were no such clause, this is not a valid legal precedent, contrary to what you claim.