|
|
Log in / Subscribe / Register

Meaningful vs. meaningless support from businesses

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 5:54 UTC (Thu) by FlorianMueller (guest, #32048)
In reply to: Meaningful vs. meaningless support from businesses by bojan
Parent article: Red Hat Responds to U.S. Patent and Trademark Office Request for Guidance on Bilski

Your mobile phone theory is demonstrably wrong because mobile phones became pretty affordable despite strong IP protection. It didn't require parasitic business models to bring prices down from the $20K level.

I don't say Red Hat shouldn't compete in the market. The question is on which terms they should compete, and to what extent politicians can consider Red Hat's business model deserving of support.

My point is that innovation has to be paid for in the end, and an entire economy can't follow the model of taking other people's developments and monetizing them because then nobody would create a product in the first place. I know and my article recognizes that Red Hat now does some R&D (not huge in absolute amounts but significant); still, there's no other company of that size who neither created nor acquired the core of its product. It would be illusionary to think that politicians are going to change the intellectual property rights regime in order to accomodate monetizers instead of original innovators.

to post comments

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 6:29 UTC (Thu) by bojan (subscriber, #14302) [Link] (19 responses)

> Your mobile phone theory is demonstrably wrong because mobile phones became pretty affordable despite strong IP protection. It didn't require parasitic business models to bring prices down from the $20K level.

That is actually your theory. You are claiming that it is better that public spend more (10 times more, no less) on the same thing. And it has nothing to do with IP protection. You are seeing the current situation as is and saying that nothing else can work, based on nothing but assumptions.

There is a reason we get cheap stuff from China these days. They can make the same thing for less. We cannot. How many are willing to pay 10 times more for essentially the same thing? Not many. And yet, you are saying we should continue doing it, because nobody wants to write open source software (demonstrably false).

> My point is that innovation has to be paid for in the end, and an entire economy can't follow the model of taking other people's developments and monetizing them because then nobody would create a product in the first place.

Red Hat contribute _all_ of their software back to open source and they wrote quite a bit. They also pay salaries of a large number of open source developers. What would you have them do? Just say "Oh, we won't use PostgreSQL, we didn't write it from scratch. Oh, we won't use Apache, we didn't write it from scratch." Please be serious. Red Hat are exploiting the fact that _other_ people found open source attractive. So can others. If everyone in the market contributed as much as Red Hat, open source would continue to be innovated just fine. Red Hat are well and truly paying their rent.

This thing where you describe Red Hat as some sort of parasite is offensive to anyone that knows that they put every last line of code back into open source.

You obviously have a problem with Red Hat because they are successful. You should really get over it. They didn't get there by stealing other people's stuff, like you are trying to portray. That stuff was there for picking, whether Red Hat wanted that or not. And once they got on it, they started doing the right thing - contributing back.

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 11:23 UTC (Thu) by FlorianMueller (guest, #32048) [Link] (17 responses)

You are claiming that it is better that public spend more (10 times more, no less) on the same thing

No. The market should result in competitive prices but on the basis of healthy, R&D-centric business models (including healthy, R&D-centric open source models). I took Red Hat's own numbers and assumptions, plus publicly available information on R&D budgets. The 10:1 ratio is their claim. My reasoning was that their case isn't a model for the innovative economy at large and you don't seem to contest that point -- which is the key conclusion it was all about.

There is a reason we get cheap stuff from China these days. They can make the same thing for less.

That's a question of a competitive global market. My angle was which approach to innovation -- parasitic weak-IP vs. R&D-centric strong-IP -- would work for an economy at large. If you let China copy any medications researched in the West, it will be cheaper but at some point there will be no more research. That's what intellectual property rights are for, and China has software patents by the way.

How many are willing to pay 10 times more for essentially the same thing? Not many. And yet, you are saying we should continue doing it, because nobody wants to write open source software (demonstrably false).

The thing you say I said is wrong. I never said that.

On the first part, I've been in this industry for a long time and I can see that open source development has certain unique strengths, but I can't see that full-time programmers on a payroll are 10 times more productive because of open source. So if you want the same quantity of the same quality of programmers, you need more or less the same R&D budget.

Red Hat contribute _all_ of their software back to open source and they wrote quite a bit.

They didn't create their original product. That's different from MySQL and JBoss (yes, I kno Red Hat acquired the latter, but it never acquired Linux).

What would you have them do? Just say "Oh, we won't use PostgreSQL, we didn't write it from scratch. Oh, we won't use Apache, we didn't write it from scratch." Please be serious.

Instead of you telling me to be serious, I have to ask you to stop claiming things I said or implied that I never said or implied. This is already the second such item I have to address in reply to just one posting.

I don't complain about them bundling other software on open source terms. You bring PostgreSQL or Apache, which aren't the issue. I said Red Hat didn't create or acquire the core of its offering: Linux. In all other market segments the leader has either created or acquired the core of his offering. Therefore, no serious politician will dare bet (in terms of intellectual property rights) on the Red Hat model, which so far works in only one market (Linux -- and not even in any other open source segment on any noteworthy scale), to be replicable and to be a model that can replace the traditional R&D-centric strong-IP innovation model.

You obviously have a problem with Red Hat because they are successful. You should really get over it.

In addition to claiming things I said or implied, you now turn this into a speculation on my motivation. Actually it's already the 2nd time you allege envy of Red Hat. That's not a facts-based, issue-focused way to have a discussion.

I've repeatedly made it clear, on my blog and here, that the angle from which I look at it is whether they are an argument for or against software patents with their business model, and responsible politicians can only conclude "for", regrettably.

They didn't get there by stealing other people's stuff, like you are trying to portray. That stuff was there for picking, whether Red Hat wanted that or not.

I'm not trying to "portray" it as stealing. Yet another thing you falsely claim or misguidedly conclude. It's not about a moral dimension, it's all about whether it's an economically sustainable approach to innovation. Even a parasite in a biological sense is just part of nature and can't be condemned morally, but an ecosystem can't live out of just parasites because eventually they all need a host.

My whole point, to state it one last time in this reply, is that politicians look at how innovation must be incentivized and (in terms of IPRs) protected, and Red Hat is a counterproductive "success story" to show to politicians when arguing against software patents. It's almost as bad as if you used the success of manufacturers of generic drugs as an argument against pharmaceutical patents.

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 12:20 UTC (Thu) by RobSeace (subscriber, #4435) [Link] (2 responses)

> I said Red Hat didn't create or acquire the core of its offering: Linux.

What are you talking about? They're creating it right now! Have you seen the
various LWN analyses of kernel changes? Red Hat is consistently the top company
making changes to the kernel... They (along with various others) ARE the creators
of Linux!!

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 12:24 UTC (Thu) by FlorianMueller (guest, #32048) [Link] (1 responses)

My blog posting doesn't deny their contributions at this stage and more recently. But if politicians talk about how to incentivize and protect innovation, the assumption is that the original innovators as well as those (often the same ones) taking the initial risk of market adoption have to be incentivized with strong intellectual property rights.

Meaningful vs. meaningless support from businesses

Posted Oct 1, 2010 7:20 UTC (Fri) by bojan (subscriber, #14302) [Link]

> But if politicians talk about how to incentivize and protect innovation, the assumption is that the original innovators as well as those (often the same ones) taking the initial risk of market adoption have to be incentivized with strong intellectual property rights.

You are definitely right about that. It is an assumption.

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 12:58 UTC (Thu) by pboddie (guest, #50784) [Link] (5 responses)

If you let China copy any medications researched in the West, it will be cheaper but at some point there will be no more research. That's what intellectual property rights are for, and China has software patents by the way.

I can't let this one go. What you're saying is that no-one would be bothered any more to develop medicines to prevent/treat/cure diseases that in many situations would take people's lives and in many others would noticeably diminish the quality of people's lives. I think you need to stop reducing everything to a single dimension with the label "revenue" attached.

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 13:07 UTC (Thu) by FlorianMueller (guest, #32048) [Link] (4 responses)

I firmly believe that patents are needed for the pharmaceutical industry. One can be against software patents and still in favor of pharma patents because those industries have very different characteristics. Most of the arguments used against software patents wouldn't work against pharma patents. In the event you are against pharma patents, you're entitled to your opinion but you won't convince any political decision-making body anywhere in the civilized free world that pharma patents should go away. The alternative would be state-run research aka the most radical form of communism.

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 15:04 UTC (Thu) by jbh (guest, #494) [Link] (1 responses)

> state-run research aka the most radical form of communism.

I don't know what to say. This is just silly.

Meaningful vs. meaningless support from businesses

Posted Oct 3, 2010 23:12 UTC (Sun) by Wol (subscriber, #4433) [Link]

aol!

Florian needs to take a look at reality. Most NEW treatments (as opposed to reruns of old drugs) ARE state-funded. They have a habit of coming out of University research.

The other big problem is that most big-pharma research is aimed at the self-inflicted illnesses of affluence. It's widely known that most of the killer diseases of the poor are simply ignored. Unless, of course, they have a habit of causing the poor old western tourist some discomfort ... malaria for example.

Cheers,
Wol

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 15:57 UTC (Thu) by pboddie (guest, #50784) [Link] (1 responses)

In the event you are against pharma patents, you're entitled to your opinion but you won't convince any political decision-making body anywhere in the civilized free world that pharma patents should go away.

You haven't heard my argument, so how would you know that it wouldn't convince anybody unless you believe that either no argument is sufficient for the abolition or curtailment of pharmaceutical and life science patents or that everyone who needs to be convinced is already thoroughly and irreversibly corrupted by the idea that the availability of medicines must be driven on a purely profit-oriented basis?

The alternative would be state-run research aka the most radical form of communism.

Because, after all, states have absolutely no interest in funding research into keeping their citizens healthy/alive? Even the US government spends considerable amounts of money on medical research (from which corporations have benefited hugely), so maybe a reanimation of Senator McCarthy is necessary to combat this "form of communism". If nothing else, it would at least provide this discussion with a participant who might share your enthusiasm for such alarmist and unfounded rhetoric.

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 16:06 UTC (Thu) by FlorianMueller (guest, #32048) [Link]

The thing I said about "state-run research" didn't relate to a certain amount of government-funded research, which obviously takes place everywhere in the capitalist world, but to replacing the entire private research sector with state-run research.

In terms of your argument that I haven't heard, I was cautious enough to say "[i]n the event" up front. I still don't see an argument that pharma doesn't need patents, and if I saw one some other time, we might slip off topic at some point.

I used generic drugs as an example. The companies providing them do very limited R&D. It's nice for consumers if their perspective is shortsighted; but they also want the next generation of drugs to be researched and those investing in that research (plus those taking the risk of original market introduction) must be incentivized and protected. That was the perspective. I'd be very interested in analogies from other industries. That would be more useful than a debate on the validity of pharma patents (I don't see any parliament in the industrialized world even talking about the possibility of abolishing those patents).

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 14:03 UTC (Thu) by bojan (subscriber, #14302) [Link] (7 responses)

You know, when you continue with this nonsense about hard working folks at Red Hat being parasites, I just don't have any patience to reply to with anything but an assertion that you must hate their guts.

You obviously don't understand the first thing about open source. Without community, there isn't any (examples: OOo, OpenSolaris). With community, there is (examples: Linux, Apache).

For anyone that understands what Red Hat does, it is quite clear they do not need to "acquire" Linux. After all, their core product is a distro, not the kernel (which I will bet they can swap to say BSD if they wanted to within 5 to 6 Fedora cycles).

When 90% of your distro is coming from elsewhere, this is not called parasitic, it is called open source community. The waste of effort in $50G is largely unnecesary duplication and profiteering, which FOSS mostly eliminates.

So, no, open source is not ten times more efficient because programmers are ten times better. If is more efficient because there are no artificial IP barriers that would cause unnecessary duplication and profiteering. What Red Hat writes today, you can improve tomorrow.

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 14:13 UTC (Thu) by FlorianMueller (guest, #32048) [Link] (6 responses)

about hard working folks at Red Hat being parasites

A business model can be parasitic no matter how hard people work. Those are unrelated issues.

If is more efficient because there are no artificial IP barriers that would cause unnecessary duplication and profiteering. What Red Hat writes today, you can improve tomorrow.

That isn't an argument for hugely lower total R&D costs. You can argue that it's desirable to have that freedom to modify the code, or "to enough eyeballs, all bugs are shallow", etc. But at the end of the day, a given amount of development requires a given amount of total input. If from one vendor, it's tha vendor's R&D budget; if from multiple vendors, it's their collective R&D budget. But ultimately a programmer will be on someone's payroll, and that someone paying the programmer must make money somewhere. You just attribute the R&D budget of proprietary software companies to "largely unneces[s]ary duplication and profiteering". I can't see "duplication" being such a big issue, especially since commercial licensing deals can always be worked where they're more efficient than duplicate efforts; and "profiteering" happens with all business models.

I presume many people here program for a living and want to be sure that there's a lot of demand on the job market for software developers. I'd like to know your theory for how the realization of Red Hat's vision (them making $5B in exchange for destroying $50B others make) would be good for programmers seeking (or seeking to retain) a job. Should tens of thousands of developers currently developing proprietary software be laid off and enter the job market? If not, what's your theory for how they will all be better off if Red Hat does what its CEO says?

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 22:04 UTC (Thu) by bojan (subscriber, #14302) [Link]

> I presume many people here program for a living and want to be sure that there's a lot of demand on the job market for software developers. I'd like to know your theory for how the realization of Red Hat's vision (them making $5B in exchange for destroying $50B others make) would be good for programmers seeking (or seeking to retain) a job. Should tens of thousands of developers currently developing proprietary software be laid off and enter the job market? If not, what's your theory for how they will all be better off if Red Hat does what its CEO says?

Look who's whining about job losses now. It's always the same thing - we can't change the economy because people will lose jobs - an argument used by monopolies worldwide. Well, yeah, that's how transformation generally works. A less efficient model is replaced by a more efficient one.

It is not about who's going to keep their jobs. It is about how much money the general public need to spend on a particular type of good.

So, the theory has nothing to do with programmers being better off. It has everything to do with general public being better off.

BTW, if Red Hat need to grow ten times (example) to get to $5B, that means that 10 times more _open_ _source_ programmers will get a job there. In terms of others, like myself, who get paid by _other_ companies (not strictly in the business of open source) to use and maintain this software, we'll continue being paid just fine. And we'll contribute back just fine (in my case, admittedly, not as much as I would like to, but somewhat). And you may also find that should more proprietary software be displaced by open source, more folks like myself will be around in these "other" companies, creating an even more powerful community. You are ignoring this effect entirely.

Your world view involves replacing one thing with another, without any change whatsoever occurring anywhere. This is not the nature of progress. Disruptive changes, like this one, change many different aspects of society and new equilibrium is then established. This is then shattered by a different model, when the weight of the next big change is sufficient to collapse the new, now old model. At which point, more whining will be heard about job losses and the like.

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 22:24 UTC (Thu) by bojan (subscriber, #14302) [Link] (3 responses)

> I can't see "duplication" being such a big issue, especially since commercial licensing deals can always be worked where they're more efficient than duplicate efforts;

Of course. I'm sure Oracle, IBM and SAP share most of their code. Give me a break.

> and "profiteering" happens with all business models.

Yeah, and Bill Gates and Larry Ellison are not worth $80G combined. Where were you in the last two decades?

Meaningful vs. meaningless support from businesses

Posted Oct 1, 2010 4:02 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (2 responses)

I'm sure Oracle, IBM and SAP share most of their code. Give me a break.

Sharing code across companies with a different focus only makes sense for larger pieces. Cross-licensing patents is another thing.

There are plenty of problems when trying to integrate open source code available under one license into code under another. GPLv2 isn't even compatible with GPLv3...

Yeah, and Bill Gates and Larry Ellison are not worth $80G combined. Where were you in the last two decades?

Aggressive questions like the last one don't strengthen any of your arguments. Of course I'm aware of the money that's been made in that sector. That's part of capitalism, and as long as competition rules are enforced properly, wealth creation is a good thing. The biggest achievement of those entrepreneurs is not that they made themselves mega-multi-billionaires but that they made many of their employees (typically through stock options) millionaires over the years.

Red Hat has reached a market cap of $8 billion, so it has also had that effect, but the ratio between the value it created in terms of stock and the employment it created for the economy appears less favorable.

Meaningful vs. meaningless support from businesses

Posted Oct 1, 2010 6:06 UTC (Fri) by bojan (subscriber, #14302) [Link] (1 responses)

The difference in this "wealth creation", is that proprietary software does this by overcharging for their monopolies (i.e. what happened in the last two decades), while open source does it by charging for service and support. There is a lot less chance the second model will run away into profiteering, because competiotors have a much easier way of getting in.

Unfortunately for your theories, Red Hat are, as I said before, very good at what they do. So, they take the lion's share (for now).

I have no idea when was the last time you installed anything Red Hat made, but if you did, you would know that there is plenty of innovation there. So, although you are trying to present a situation in which they supposedly don't have the "core" of their product (whatever that's supposed to mean), I can assure you that their "core" is just fine and being rewritten daily. How do you think one gets from Linux From Scratch to Fedora, if not by innovating.

In the end, the main test is which model produces the same goods cheaper. You seem to think that we have to keep spending $50G instead of $5G, because someone may lose their job.

Meaningful vs. meaningless support from businesses

Posted Nov 14, 2010 19:30 UTC (Sun) by promotion-account (guest, #70778) [Link]

And also remember that RedHat has engineers on most of the relevant FOSS layers. From the kernel, to the plumbing layer (NetworkManager, *kit packages, etc), to GCC, to X, to glibc, to GTK, to the foundational GNOME libraries (libxml2, etc), to the user-facing GNOME applications, to the RPM packagers themselves.

So they are not really 'taking away' anything. Our stack wouldn't be the way it is without RedHat.

And speaking of jobs, there are lots and lots of FOSS developers who are having projects that they love thanks to these developers original contributions. Where are they? They are allover the place in the usual silicon valley companies.

Meaningful vs. meaningless support from businesses

Posted Oct 3, 2010 23:15 UTC (Sun) by Wol (subscriber, #4433) [Link]

Sounds like you haven't heard of the word "symbiosis", actually.

Cheers,
Wol

Meaningful vs. meaningless support from businesses

Posted Oct 1, 2010 3:05 UTC (Fri) by vonbrand (guest, #4458) [Link]

Not only that: Red Hat has bought companies/software packages with the express purpose of open sourceing the software. Look at 389 Directory Server, the KVM stuff, JBoss.

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 9:45 UTC (Thu) by klbrun (subscriber, #45083) [Link]

>My point is that innovation has to be paid for in the end,

As open source software development has become more complex, software configuration management tools became increasingly outdated, and now we have two CM tools using the distributed CM approach: git and Mercurial, both developed as open source products. Proprietary CM tools have not been able to keep up, and these tools will transform the way software is developed in the 21st Century.

Sure, somebody paid the people who developed these CM tools, but the innovation was not directed by the payer. That is why Google has a policy of engineers devoting 20% of their paid time to developing their own innovations.

Meaningful vs. meaningless support from businesses

Posted Sep 30, 2010 12:37 UTC (Thu) by gnufreex (guest, #70396) [Link] (51 responses)

How's that you condemn Red Hat's "parasitic" business model, and at same time you support other "parasite", TurboHercules? By your logic, they are trying to destroy IBM's mainframe revenue, hence destroying innovation ('cause in your world revenue equals innovation - that shows you don't know squat about free software, or for that matter, software development in general).

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Sep 30, 2010 13:30 UTC (Thu) by FlorianMueller (guest, #32048) [Link] (50 responses)

I don't "condemn" Red Hat's business model. What I say is that if their economic model were to be presented to politicians as an argument against software patents, they would look at this as an unsustainable model for the economy at large. Red Hat is free to do what open source licenses and the law allow. But to the extent that Red Hat wants intellectual property laws to be weakened only to accomodate its approach, it doesn't have a compelling case. On the contrary, a majority of politicians would likely conclude that such destruction as outlined by Red Hat's CEO isn't desirable.

TurboHercules isn't a matter of patent policy. It's an antitrust case that's now being investigated by the European Commission, and it has implications for IBM's credibility vis-à-vis open source when comparing IBM's overall claims to support open source and interoperability to its practices in its core business.

TurboHercules didn't go to the EU and ask them to abolish patents, or to force IBM to make patent licenses available free of charge.

Instead, TurboHercules lodged a complaint, saying that IBM acts in violation of antitrust law by tying its mainframe hardware to its proprietary z/OS operating system. That same kind of tying complaint was made by IBM itself in the EU against Microsoft, even twice: in connection with the Media Player, and in connection with Internet Explorer.

TurboHercules never said IBM should give away z/OS for free. Of course IBM has the right to market it as proprietary software. But a company in a dominant market position (in this case, a monopolist, which is the ultimate form of dominance) can be required under antitrust rules to untie components. TurboHercules' original request to IBM (the entire correspondence is documented here) said that IBM should make an offer of commercial terms that IBM could choose as long as it's reasonable. In other words, if IBM did licensing on so-called RAND or FRAND terms, everything would be fine. So IBM wouldn't be expropriated at all. They would be paid for their intellectual property.

IBM not only refused to make an offer but also brought up patent infringement assertions. Instead of acting constructively and cooperatively like, for instance, Microsoft makes its patents available, IBM uses some patents for exclusionary strategic purposes to foreclose competition. If a monopolist uses IP for that purpose, there's also an antitrust dimension. But the key thing here is again that the problem would be solved by IBM making its intellectual property available on FRAND terms. That, again, means they get paid, not expropriated.

Since TurboHercules asked for an offer of reasonable licensing terms, that's what IBM should have offered, and the patent part could have been resolved as part of the overall deal. But IBM wants to foreclose competition. I never said that markets should be monopolized. There should be and must be undistorted competition. The Red Hat issue is not a competition issue but one of business models in connection with patent policy. The TurboHercules case is a competition issue, not a matter of patent policy.

There is also the open source/open standards dimension. IBM claims to support and advance the cause of open source and open standards. It furthermore claims that only royalty-free (RF) access to patents is compatible with FOSS. RF is the most liberal extreme; strategic exclusionary use ("no way, José") as practiced by IBM itself is the other extreme. FRAND is a constructive solution in the middle. If IBM says in the open standards context that FRAND isn't good enough but doesn't even offer FRAND when its own IP is concerned, that doesn't make sense. They should either offer FRAND and demand it from others; or offer RF and demand it from others. They should be consistent.

Finally, it's a fact that TurboHercules made the suggestion that IBM should add to its open source patent pledge any patents it deems relevant. That was a way of pointing out that IBM's open source patent pledge isn't sincere if it doesn't contain the patents that really seem to matter to open source. However, since TurboHercules had stated clearly that IBM should offer reasonable terms, everything would be resolved with IBM granting a license on such terms. So what's needed to solve this problem is a license deal; the abolition of software patents wouldn't be needed.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Sep 30, 2010 14:50 UTC (Thu) by ovitters (guest, #27950) [Link] (2 responses)

I find it interesting/funny that in your view, politicians completely agree with your assertations.

Meaning: Even with your viewpoints, something else could be valid as well.

Politicians' positions

Posted Sep 30, 2010 15:55 UTC (Thu) by FlorianMueller (guest, #32048) [Link] (1 responses)

I find it interesting/funny that in your view, politicians completely agree with your assertations.

I spent a lot of time discussing patent policy with politicians; I also discussed other IP issues with them (in connection with the commercial exploitation of professional sports); I read the kinds of studies and papers they read; I see what they write and hear what they say.

So when I talk about where they stand and what shapes their thinking, I'm describing rather than presupposing agreement.

In some ways, it's the reverse causation: because I see where they stand in terms of views and objectives, I believe it makes sense to adjust one's demands to reality in order to pursue ambitious but achievable goals.

I only take the firm position that something is a position of a vast majority of politicians when that's really my observation. I separate that from issues I consider worth pursuing but where majority support is far from certain.

Politicians' positions

Posted Sep 30, 2010 22:13 UTC (Thu) by bojan (subscriber, #14302) [Link]

> what shapes their thinking

Primarily a desire to get re-elected. So, they will generally support whichever position can get them over the line with more of their electorate.

Five years on

Posted Sep 30, 2010 16:13 UTC (Thu) by pboddie (guest, #50784) [Link] (12 responses)

Instead, TurboHercules lodged a complaint, saying that IBM acts in violation of antitrust law by tying its mainframe hardware to its proprietary z/OS operating system.

Please explain how this is different from a game vendor tying its proprietary game client to its proprietary game services - a situation you regarded as being completely acceptable five years ago.

I'm not arguing against you here, but I'd like to know how you square this particular circle. Or, to adopt your own analogy, to know how IBM isn't entitled to ask that you build your own stadium.

Five years on

Posted Sep 30, 2010 16:31 UTC (Thu) by FlorianMueller (guest, #32048) [Link] (9 responses)

Please explain how this is different from a game vendor tying its proprietary game client to its proprietary game services - a situation you regarded as being completely acceptable five years ago. I'm not arguing against you here, but I'd like to know how you square this particular circle. Or, to adopt your own analogy, to know how IBM isn't entitled to ask that you build your own stadium.

Thanks for asking that question. Even five years later you recall that "bnetd" conversation over a third-party client for Blizzard games, but there are good reasons for why those cases are fundamentally different.

Blizzard isn't dominant in the computer games market; and it wasn't when we had that debate. So there's no antitrust case that can be made, unlike in IBM's case.

The "bnetd" developers didn't argue on the basis of antitrust law anyway. They argued with fair use and copyright, and didn't want to be bound to Blizzard's end user licensing terms the way Blizzard interpreted them.

In terms of building one's own stadium, I mentioned that in connection with an architect's right to withhold authorization of modifications of his work. That's where we're in the field of copyright and the integrity of what an artist creates. I believe that should be a consideration in connection with computer games, which are artistic creations. I don't see that point in connection with whether operating systems should be made available separately from hardware.

Also, always bear in mind that TurboHercules just wanted IBM to offer reasonable licensing terms. They didn't want IBM to be deprived of anything. That's another difference.

If free software were a constitutionally guaranteed cconcept and as such put above anyone else's intellectual property rights, TurboHercules wouldn't have a problem with IBM and bnetd wouldn't have one with Blizzard. But under the existing legal circumstances, the difference comes down to the applicability of antitrust rules. Antitrust can't be used to impose "freebies". However, it can require a dominant player to make something available on fair, reasonable and non-discriminatory terms.

Apart from legal considerations, there's also the community dimension (IBM's patent pledge etc.) and the political perspective (if IBM asks for royalty-free interoperability when other companies' rights are concerned, why don't they offer any licenses to their interoperability patents at all?). That's also different from Blizzard, which never advocated open source and royalty-free standards and therefore doesn't have IBM's hypocrisy problem.

Five years on

Posted Sep 30, 2010 17:29 UTC (Thu) by pboddie (guest, #50784) [Link] (8 responses)

Thanks for asking that question. Even five years later you recall that "bnetd" conversation over a third-party client for Blizzard games, but there are good reasons for why those cases are fundamentally different.

Actually, someone else brought the matter up fairly recently - I didn't read LWN in 2005 - and so it seemed like a reasonable moment to ask for a clarification. In fact, in that very article, someone brought up IBM mainframes and operating systems as an example of the tying instincts of certain elements of the computing industry, and so it might have seemed that since you had no problem with one vendor tying its software to its services, that IBM's behaviour might have been just as acceptable to you at that point in time.

Blizzard isn't dominant in the computer games market; and it wasn't when we had that debate. So there's no antitrust case that can be made, unlike in IBM's case.

But Blizzard dominated the market for services supporting its own games. Why is it different that IBM dominates the market for hardware that can run its own operating system?

And does caring about people who have chosen to be locked into IBM's architecture help the rest of us who have to unbundle Microsoft Windows every time we want to buy hardware from mainstream retail vendors? That's a far more important battle that never seems to end.

Five years on

Posted Sep 30, 2010 17:56 UTC (Thu) by FlorianMueller (guest, #32048) [Link] (7 responses)

IBM's behaviour might have been just as acceptable to you at that point in time.

If I had been asked to comment specifically on mainframe at that time, I wouldn't have had enough knowledge about that market to be able to understand the implications of IBM's conduct.

What we're talking about are two principles that must be weighed against each other: the concept of intellectual property, and the need for undistorted, effective competition. By default, it's all about intellectual property, and the bnetd case was just about intellectual property. In some cases, competition considerations (antitrust etc.) come into play, and then the result can be a different one than under IP law itself. However, if competition law sort of overrules IP, there must still be a reasonable compensation of the right holder, so IP is still respected.

But Blizzard dominated the market for services supporting its own games. Why is it different that IBM dominates the market for hardware that can run its own operating system?

One can define all sorts of markets, but that doesn't necessarily make them "relevant product markets" under competition law.

The mainframe market has been considered a relevant product market on prior occasions. It's definitely not just part of a larger server market. IBM's price discrimination makes the execution of legacy workloads about 10 times as expensive as new workloads, by selling the same hardware but with minor microcode modifications that prevent the execution of legacy workloads on those lower-priced components. Just looking at that behavior, it's clear that mainframe legacy workloads belong into a market that's separate from the overall server market. In Blizzard's case, there isn't anything similar.

I personally would like all operating system providers to have an obligation to provide reasonably priced licenses to their software regardless of hardware. But so far that's not a legal requirement for all of them to meet. Only antitrust law can currently impose this on a vendor.

help the rest of us who have to unbundle Microsoft Windows every time we want to buy hardware from mainstream retail vendors? That's a far more important battle that never seems to end.

If a PC vendor sells his machine only with Windows and doesn't offer it with Linux (or simply without any operating system), then that means you are required to buy Windows if you want to buy that particular machine. That's the opposite of the IBM mainframe situation. IBM requires you to buy its extremely expensive hardware if you want to run its operating system. The hypothetical equivalent for Windows would be if Microsoft stopped supplying hardware companies with OEM licenses and instead became a hardware vendor itself, forcing you to buy their hardware if you want to run Windows at all. You can be sure there would be an outcry, and antitrust intervention, if that happened.

The PC hardware market is competitive enough and there are vendors who offer you devices with Linux pre-installed. If there's choice, then bundling isn't the kind of problem that it is if there's a dominant vendor, let alone a monopolist.

The idea of illegal tying is that you do business (such as buying a product) with someone who's in a dominant (at worst, monopolistic) position and requires you to accept unfair conditions or he won't do any business with you at all. No single PC manufacturer is in such a position that you couldn't just go to another vendor if you prefer Windows-less PCs. But with IBM, customers currently have no choice if they want to run those legacy applications.

Five years on

Posted Oct 1, 2010 13:01 UTC (Fri) by pboddie (guest, #50784) [Link] (6 responses)

If a PC vendor sells his machine only with Windows and doesn't offer it with Linux (or simply without any operating system), then that means you are required to buy Windows if you want to buy that particular machine. That's the opposite of the IBM mainframe situation.

How is it "the opposite"? I want one thing but if every separate product I inspect involves me having to take another thing with it, then effectively I have to take both of those things.

IBM requires you to buy its extremely expensive hardware if you want to run its operating system. The hypothetical equivalent for Windows would be if Microsoft stopped supplying hardware companies with OEM licenses and instead became a hardware vendor itself, forcing you to buy their hardware if you want to run Windows at all. You can be sure there would be an outcry, and antitrust intervention, if that happened.

Well, I grant you that at this point you've reproduced the IBM situation, but then there would be no illusion about who is forcing customers to take whose products, although there's always that argument that mainframes and operating systems written for them are all part of a single product. Meanwhile, PC vendors and Microsoft can point the finger at each other, while in practice almost everybody has to acquire Microsoft products when they buy a computer.

The PC hardware market is competitive enough and there are vendors who offer you devices with Linux pre-installed. If there's choice, then bundling isn't the kind of problem that it is if there's a dominant vendor, let alone a monopolist.

Yes, but we're talking about effective choice here, not whether alternatives exist. It's a bit like large retail chains getting into trouble when they secretly agree on prices whereas a bunch of "man plus dog" vendors probably won't see anything like those levels of scrutiny. Why is that? Because the large chains' collusion will affect the vast majority of consumers.

No single PC manufacturer is in such a position that you couldn't just go to another vendor if you prefer Windows-less PCs.

True, but the matter of effective choice arises again. If 99% of retailers force purchasers to acquire Windows and those purchasers don't know about the other 1%, they effectively have to acquire Windows with every purchase.

But with IBM, customers currently have no choice if they want to run those legacy applications.

I guess that from a bundling or tying perspective we have all been told often enough by now that the IBM mainframe case is extreme, if you regard the market for hardware solutions running IBM's proprietary software as a genuinely open market, but my point was that there are indisputably genuinely open markets that need not exhibit such extremes (and by not doing so actually demonstrate obvious open market credentials) that could more usefully do with some proper regulation that would benefit many more people.

Five years on

Posted Oct 1, 2010 17:49 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (5 responses)

There isn't a competition issue with what you describe.

You can buy Linux computers from major vendors like Dell and HP. Someone who wants Linux can get machines without Windows. The reason they're less aggressively promoted is that there's less demand on the desktop for Linux than for Windows, but antitrust law is meant to ensure that demand and supply can drive a market as opposed to abuse distorting it.

The mainframe operating system was actually also available independently from the machine for about three decades -- the time when plug compatible mainframes were available from other manufacturers. Due to the so-called Consent Decree by the US DoJ, IBM was forced to make components available separately.

In the PC market, there's no need to force anyone because you can buy Windows separately, and you can buy every hardware component separately.

You say that the two situations (mainframe and PCs shipped with Windows) are the same. They are not the same because a case of illegal tying is characterized by customers being forced to take something they don't want in order to buy something they need from a dominant vendor. No single PC vendor is dominant, and I mentioned above that Dell, HP and others sell you what you want, without Windows if you want. By contrast, mainframe customers need z/OS to run those legacy applications worth $5 trillion (yes, trillion) and are forced to buy IBM hardware.

You complain that maybe some customers buy desktops with Windows simply because they aren't aware of the possibility of buying Linux desktops, especially since in many retail stores those have less presence. Again, that's a supply-and-demand thing. By contrast, mainframe customers know very well what they want and can't get it, due to an abuse of a dominant position in the form of illegal tying.

Regulators can't and won't intervene if it's just an awareness problem that can be solved with marketing. If the likes of Red Hat (which is the actual context of this discussion) or Ubuntu believe that end users should be made aware of the possibility of buying Linux-powered rather than Windows-powered PCs, it's their job to generate awareness. They could also do so in partnership with retail chains. If there's an unexploited or, more realistically, perhaps a bit underexploited market potential, market dynamics will result in some retailers cooperating with some vendors and seizing that opportunity. There's nothing preventing anyone from doing that in connection with Linux. Hence, there's no antitrust problem.

You ask for regulation but you have no antitrust issue. You'd like the market to be different, and I can assure you that I, too, think it's essential that customers are fully aware of Linux as an alternative and that they have easy access to it -- but the market can take care of itself in that regard, while the mainframe market shows clear symptoms of total market failure due to abuse.

Five years on

Posted Oct 2, 2010 23:40 UTC (Sat) by pboddie (guest, #50784) [Link] (4 responses)

You can buy Linux computers from major vendors like Dell and HP.

If you can find them without playing "beware of the leopard". Likewise with machines without any operating system whatsoever.

Someone who wants Linux can get machines without Windows.

Yes, that's true: I'm using exactly such a machine. However, ignoring most of the machines built from components, virtually all of the machines sold on the Web site where I bought mine are bundled with Windows, including the "build your own" machines where you don't even get to do that without a copy of Windows magically appearing together with that mainboard, CPU, RAM, and so on. Now observe the same phenomenon throughout the breadth of the retail channel directed at "the consumer".

The reason they're less aggressively promoted is that there's less demand on the desktop for Linux than for Windows, but antitrust law is meant to ensure that demand and supply can drive a market as opposed to abuse distorting it.

Which brings us to the interesting case of the Web browser, does it not? After all, can't people who want Firefox or Opera install it themselves?

In the PC market, there's no need to force anyone because you can buy Windows separately, and you can buy every hardware component separately.

It's not just about being able to buy Windows separately; it's about being able to buy the hardware separately as well. (Besides, the only people needing to buy Windows separately are those who want to upgrade the version thrust upon them when they bought their existing hardware.)

You say that the two situations (mainframe and PCs shipped with Windows) are the same. They are not the same because a case of illegal tying is characterized by customers being forced to take something they don't want in order to buy something they need from a dominant vendor. No single PC vendor is dominant

Yes, but the operating system vendor is dominant. And doesn't competition regulation have anything to say about a bunch of companies getting together and stifling competition? If it were all about a single company doing something, all the cases involving collusion between, say, oil companies would fall apart when someone points out that it's not one single gigantic oil vendor involved.

Anyway, my point was that there's a genuinely open market for personal computers and operating systems that gets undermined by the vast majority of computers somehow needing to have a particular vendor's product installed on them. (Talk about supply and demand all you like, and I don't entirely disagree that such things are an influence, but why it appears almost impossible to drop a supposedly separate product from a purchase remains a mystery.) You assert that there's a genuinely open market for solutions that can run a particular proprietary mainframe operating system.

Since I'm a believer in interoperability, I support the idea that you should be able to run whatever you want on anything you can - that you shouldn't have to take the hardware if you just want the software - but not only do I not think that it has to be a completely extreme situation before the regulators need to get involved, I also think that such an extreme situation risks being perceived as something other than a market needing regulation. And then, if that judgement doesn't go your way, by your own criteria you've just put the regulators out of work.

So I hope the regulators feel that they have a role to play, even though history suggests that in the EU, at least, they would rather mess around with things like media players and Web browsers than get to grips with the underlying causes.

Five years on

Posted Oct 3, 2010 0:32 UTC (Sun) by dlang (guest, #313) [Link] (1 responses)

back when IBM was investigated for their mainframe monopoly, the mainframe basically _was_ the computer market (there was a smaller market in mini computers, but microcomputers/PCs were not a noticable factor)

today the entire mainframe market is happily ignored by almost all businesses using computers.

At this point, saying that IBM has a monopoly on the mainframe market and should be forced to separate their hardware from their software is very similar to saying that Apple has a monopoly on their market (they are the only ones who make the hardware, and they forbid anyone from using their software with anyone else's hardware)

It would be nice to have both open, but the law doesn't require it, and in the Apple vs Pystar lawsuit recently, Apple got the court to rule in their favor in a big way (namely that copyright forbade pystar from installing OSX on their computers, even though they had purchased the appropriate number of legal copies)

In many ways, apple's control of their market is much more significant to consumers than IBMs control of the mainframe market.

Five years on

Posted Oct 3, 2010 4:54 UTC (Sun) by FlorianMueller (guest, #32048) [Link]

today the entire mainframe market is happily ignored by almost all businesses using computers.

This statement is true only if you base it on the number of companies but disregard size. In terms of economic weight, we talk about legacy program code worth $5 trillion ($5,000 billion) still in use, and about 80% of the world's business data still being processed by mainframes. IBM itself said a couple of months ago that Western civilization runs on the mainframe. You can read more about it here.

At this point, saying that IBM has a monopoly on the mainframe market and should be forced to separate their hardware from their software is very similar to saying that Apple has a monopoly on their market

From a competition point of view, those cases are different. In order to counter the claim that they dominate what competition law calls a relevant product market, both companies would have to argue for a broader market definition including all sorts of competitors. Apple would have a much better basis to do so than IBM. Apple's products are generally more expensive than those of their competitors, but the difference is limited enough that one can argue they are exposed to competitive pressure from others operating in the same relevant market. By contrast, if you look at IBM's price discrimination in the mainframe market, you can easily see that mainframes and other servers aren't the same market. Otherwise IBM wouldn't charge you about ten times as much -- for identical hardware -- for the execution of mainframe legacy workloads as for new applications operating in a more competitive market. Also, mainframe memory costs about 60 times as much as memory for other servers. Sure, mean time between failures is very high, but still that's the kind of price difference that shows it's not the same market.

It would be nice to have both open, but the law doesn't require it,

In IBM's case it can certainly be imposed, like it was in the past, and I'm confident that this will be the outcome of the ongoing investigations in the EU.

Five years on

Posted Oct 3, 2010 4:45 UTC (Sun) by FlorianMueller (guest, #32048) [Link] (1 responses)

Which brings us to the interesting case of the Web browser, does it not? After all, can't people who want Firefox or Opera install it themselves?

At first sight one may believe there's an important parallel, but there isn't under competition law.

If bundling is abused by a dominant vendor in one market to gain an unfair unadvantage in another market (which was the argument in the Media Player context, used by complainants as a precedent for the Internet Explorer discussion), then there's a competition problem. But note the "dominant vendor" requirement. There's no dominant PC company, and instead major players like Dell and HP do offer Windowsless Linux configurations.

Yes, but the operating system vendor is dominant.

It was found dominant. But being dominant isn't illegal. It's the combination of dominance and abuse that's illegal.

And doesn't competition regulation have anything to say about a bunch of companies getting together and stifling competition?

Sure, that would be called a cartel: a form of cooperation with the object of illegally restricting competition. There's no indication that Dell, HP and others have conspired against Linux. On the contrary, some vendors are known for a very pro-Linux attitude, but they all have to respond to what the market (in the context you care about: the consumer market) wants at this stage.

even though history suggests that in the EU, at least, they would rather mess around with things like media players and Web browsers than get to grips with the underlying causes.

It's interesting that you say this because five years ago I criticized the European Commission in a similar way. I said that on the one hand they were pursuing Microsoft with antitrust law but on the other hand supporting the very software patents Microsoft wants. I would have preferred a focus on software patents, especially since I was mostly concerned about patents on codecs in terms of the ability of others to compete with the Media Player.

But as far as interoperability is concerned, the European Commission is indeed looking at what might ultimately become a legislative solution that would affect all "significant market players" in order to establish general rules rather than only pursue individual cases,

Five years on

Posted Oct 4, 2010 11:37 UTC (Mon) by pboddie (guest, #50784) [Link]

Rewinding to bring in the context...

They are not the same because a case of illegal tying is characterized by customers being forced to take something they don't want in order to buy something they need from a dominant vendor. No single PC vendor is dominant
Yes, but the operating system vendor is dominant.
It was found dominant. But being dominant isn't illegal. It's the combination of dominance and abuse that's illegal.

Which is what you already said before you trimmed away the context. Yes, you can get alternatives to Windows on hardware purchased through mainstream retail channels, if you managed to navigate to the disused lavatory where those alternatives are "on display", and you might even be able to ask for no operating system at all, but most "consumers" get told that Windows is "part of the product" and that they have to take their complaints to the vendor (the classic lazy retailer excuse, contrary to consumer regulation in various countries), who in turn often tell the customer the same thing or that their deal with Microsoft precludes any kind of refund. At which point, most people determined enough to pursue the matter that far (and few people are) are likely to give up and write off their loss, which is the gain of the "dominant vendor", of course.

Now maybe this isn't a "big R" regulatory matter that involves finger-wagging from the European Commission - although it is baffling that they entertain the dog and pony show that is the Web browser parade with all its back-and-forth between the EU and Microsoft, but not something effective like this - but it certainly is a matter that requires some kind of "small R" regulatory intervention, at least from the perspective of anyone whose idea of buying stuff isn't having extra stuff thrown into one's shopping basket and being made to pay for it. But I accept that our perspectives may differ on such matters.

Five years on -- one addition

Posted Sep 30, 2010 16:36 UTC (Thu) by FlorianMueller (guest, #32048) [Link] (1 responses)

I just explained the key differences, but I'd also like to look at it from the customers' perspective. With Blizzard's games and that "bnetd" server, all that people needed was an Internet connection and a legitimately purchased copy of those Blizzard games (meaning a cost of $50 more or less) to play them with/against others. In the mainframe case, there's a massive lock-in problem and it's a huge market. There's no Blizzard-related lock-in, other than people perhaps being addicted to some of those games ;-)

Five years on -- one addition

Posted Sep 30, 2010 17:54 UTC (Thu) by dlang (guest, #313) [Link]

I see things the other way around.

the game software is perfectly usable on any PC, it can even be used to play against other users directly without the bdnet server (at least on a local LAN where you are not behind NAT devices). Saying that it's illegal to replicate the Internet accessable server that the local games can talk to isn't reasonable.

I've already said elsewhere that I don't have a big problem with IBM's stance.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Sep 30, 2010 16:17 UTC (Thu) by gnufreex (guest, #70396) [Link] (33 responses)

I was not talking about patents, and I know all about TH case that you brought against IBM. You don't need to link to your blog, I read all that already.

You blogged about Red Hat business model, and called it "parasitic" and "destructive". You said Red Hat does more harm than good. That is attack on open source. You are now weaseling out, saying that it is not but it is. Real word for Red Hat business model would be commoditizing. Not destroying, or parasitic.

So my question was how come you support TurboHercules' attempts to commoditize Mainframe, and at same time you are attacking Red Hat because it comoditizes UNIX and Windows? It was simple question.

Your convoluted post apparently boils down to: TH is willing to pay the racket and stop being open source (by OSD) and free (as in freedom). Red Hat wants to be open source company, so in your book, Red Hat must be stopped or else it will endanger Microsoft.

Is that right?

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Sep 30, 2010 16:56 UTC (Thu) by FlorianMueller (guest, #32048) [Link] (32 responses)

I was not talking about patents

I'm sorry if you'd have preferred me to dismiss your question as off-topic in this patent context. I tried to interpret it from an on-topic perspective.

Real word for Red Hat business model would be commoditizing. Not destroying, or parasitic.

We must keep those two angles separate. I'm all for commoditization per se; that is a question of market dynamics. The "parasitic" business model relates to a company having built its success on the basis of something it didn't originally create (or acquire its creators).

You can commoditize -- by providing customers with solution meeting their needs but at a greatly reduced price -- without being "parasitic". MySQL started to commoditize the database market, but it developed its product with its own resources. So you can commoditize and be open source, and still don't have to be "parasitic".

So my question was how come you support TurboHercules' attempts to commoditize Mainframe, and at same time you are attacking Red Hat because it comoditizes UNIX and Windows?

In the mainframe market, IBM is a monopolist and abuses that position, so we have a clear case of a market failure. That's why antitrust intervention is needed. If TurboHercules's antitrust complaint succeeds (the fact that it gave rise to a full-blown investigation is obviously great progress), the outcome will be a license deal on fair, reasonable and non-discriminatory (FRAND) terms. So we talk about commoditization that fully respects the commercial value of the intellectual property involved. IBM can take those license fees and invest them in further innovation, while customers get the benefit of more choice, which will reduce IBM's monopoly rent.

TH is willing to pay the racket and stop being open source (by OSD) and free (as in freedom).

They'd love to get a license from IBM on terms that are 100% in line with FOSS ideas. The problem is that if they can't get a license on that terms, they'll have to make the best out of what they can get. Antitrust can't force IBM to grant a royalty-free license. All that antitrust can achieve is a requirement to grant a FRAND license. So the problem isn't TurboHercules' support of FOSS -- it all depends on IBM. Contrary to supporting FOSS, IBM so far doesn't even offer a FRAND license. So IBM is right now on the other side of the spectrum.

You say TH is "willing to pay the racket" when all we're talking about is IBM. As far as the proprietary z/OS operating system is concerned, it would be wonderful if IBM open-sourced it, but since it doesn't happen, customers need a way forward. They have legacy code in which they invested (according to IBM's own estimate) $5 trillion (5,000 billion dollars). Being able to run z/OS on non-IBM hardware would be an important step.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 1:25 UTC (Fri) by bojan (subscriber, #14302) [Link] (31 responses)

> MySQL started to commoditize the database market, but it developed its product with its own resources.

Oh, would you please lay off it already. The success of MySQL is not just their own. LAMP is what succeeded, where L stands for a "Linux distribution" (the kernel on its own is rather useless), which for all it matters, could be FreeBSD (sorry guys, you are parasites too). MySQL have been "parasites" (the "wrong doing" implying term you keep using) on the back of that as much as anyone.

Or are you suggesting that all those systems running LAMP have been running Windows NT/2000 Server? I think not. Most of the LAMP system have been either Red Hat or Debian (oh, and you guys are parasites too).

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 4:04 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (30 responses)

Actually WAMP is also pretty widespread. And there are multiple P languages, or people use non-P languages like Ruby, C#, Java.

But you don't have a point there because MySQL created its product and took the risk of market acceptance. Red Hat didn't do that for Linux. I said before that I'd view it differently if Red Hat had created Linux and then threw in Linux-based software as part of a distribution.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 5:20 UTC (Fri) by bojan (subscriber, #14302) [Link] (1 responses)

> But you don't have a point there because MySQL created its product and took the risk of market acceptance.

What product? I assert that MySQL without the operating system is utterly useless. Where is the MySQL distribution of the OS, please? According to you, they took no risk, because they took for granted that someone else is going to do the OS for them.

Your cherry picking is wrong on so many levels, it's not even funny. Red Hat do _different_ things to MySQL. They took plenty of risk when they created RHEL or are we now rewriting history?

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 5, 2010 22:31 UTC (Tue) by nix (subscriber, #2304) [Link]

Also, RH pour a lot into PostgreSQL, and have for a long time. But apparently that doesn't save them from being 'parasites', because they contributed to a free software project that Florian doesn't approve of. Or something. (If there is logic here, I'm not seeing it.)

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 5:43 UTC (Fri) by gnufreex (guest, #70396) [Link] (27 responses)

>But you don't have a point there because MySQL created its product and took the risk of market acceptance. Red Hat didn't do that for Linux.

That is just bullocks. Red Hat is one of biggest reasons for success of FLOSS in general. They and Cygnus put it on the map. I would even argue that MySQL owes lots of it's success to Red Hat. If MySQL wasn't in RHEL, it would be stillborn. So MySQL was exploiting Red Hat's distribution channels. Are they parasites?

>I said before that I'd view it differently if Red Hat had created Linux and then threw in Linux-based software as part of a distribution.

So your idea is that only Linus Torvadls and Richard Stallman should make money of off GNU/Linux, and everybody else are parasites? Because RMS created GNU and Torvalds jumpstarted Linux development...

That is really, really crazy. And it is really convenient when you call on some unnamed politicians and say that you have no opinion of your own, you only reporting what they are saying to you. You should educate those politicians because they are obviously clueless about FLOSS. It turned other way around, you are echoing their ignorant opinions. You are basically advocating for converting open source to proprietary software because you only know to think by proprietary terms.

It is also very hypocritical that you praise MySQL's business model but you cry foul because Oracle bought it. It is all funny game when you are shareholder and when GPL-fear-mongering business model of MySQL AB makes money to you, but when someone else buy it, you want to un-GPL it.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 6:19 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (13 responses)

@gnufreex, this isn't the first website on which we have disagreements, but please note that this here isn't #boycottboy's website but instead a really good discussion community -- and adjust your behavior accordingly.

So MySQL was exploiting Red Hat's distribution channels. Are they parasites?

They benefited from it, but they developed MySQL and Red Hat distributed it because it was useful and at some point in strong demand. I can't see any particular action here that could be considered "parasitic". Note the word "action", which is different from just having permitted through an open source license that Red Hat redistributed it.

So your idea is that only Linus Torvadls and Richard Stallman should make money of off GNU/Linux, and everybody else are parasites?

Starting a project is usually just a part. The kind of innovative effort that politicians want to protect with intellectual property rights would include the entire product development and taking the overall risk of market acceptance.

It's great that Linux became a success without that protection. I repeat: it's great. But in connection with intellectual property rights, the question is whether this would work for the economy at large. Since IPRs affect the economy at large, this means that accomodating Red Hat is not an option.

You are basically advocating for converting open source to proprietary software because you only know to think by proprietary terms.

Wrong. I want both to compete with each other. But open source will have to compete under the framework of an intellectual property rights regime that must suit traditional innovation models, and open source must respect that system and find its place within it, such as by obtaining patent licenses where necessary (video codecs, unless VP8 is proven to really be unencumbered by third-party patents).

It is also very hypocritical that you praise MySQL's business model but you cry foul because Oracle bought it

I raised legitimate concerns, and if those concerns hadn't been worth investigating, the EU (plus "neutral" Switzerland) and China wouldn't have delayed their approval for five months beyond US DoJ approval nor would Russia have delayed its approval by almost another two months beyond EU/China and imposed formal remedies on Oracle. So I apparently had a point.

Those were competition concerns. They have nothing to do with the shareholders of a company making money. Competition/antitrust concerns are all about the public interest in effective, undistorted competition -- nothing else.

when GPL-fear-mongering business model of MySQL AB makes money to you, but when someone else buy it, you want to un-GPL it.

The lie that I wanted to "un-GPL" MySQL was originally propagated by GroklXX, a website that also said Oracle was going to be a great owner of Sun's patents until Oracle decided to use some of them against Google. I have debunked the outrageous un-GPL lie to the fullest extent. You can find my rebuttal here, including links to several documents I used during the merger control process and in which you can see I vigorously defended MySQL's GPL-based business model and advocated its continuation (but under a different owner than Oracle).

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 8:21 UTC (Fri) by bojan (subscriber, #14302) [Link] (9 responses)

> I can't see any particular action here that could be considered "parasitic".

Your problem is that you pick some arbitrary point along the software development cycle and say that anything done before is not parasitic, while anything after it is.

You obviously do not see value in innovation that Red Hat Linux was when it was originally released. You think that just because they used "parts" to assemble their product, there is no innovation. This is even ignoring the fact that they in fact wrote RPM and a whole bunch of graphical admin tools very early on.

At the same time, MySQL, who also found certain "parts" in the environment and then wrote their DB engine are pure innovation. Well, they sure didn't write the kernel, the sockets, the libraries they used, compilers, the SQL language etc. And because you decided (completely arbitrarily, mind you) that their point of development and onward is called "innovation", that's OK then.

I will bet that in total Red Hat contribution to open source is both more important, more original and more voluminous then that of MySQL.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 8:27 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (8 responses)

No, Red Hat wasn't nearly as important as MySQL AB in terms of innovation. MySQL was a unique product meeting certain needs. By contrast, I remember very well using SUSE Linux (then spelled "SuSE") in the late 90's when my online gaming startup ran its servers on Linux. We never needed anything from Red Hat but were able to use Linux, while you can't use MySQL without using something from MySQL AB.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 9:04 UTC (Fri) by bojan (subscriber, #14302) [Link] (7 responses)

> We never needed anything from Red Hat but were able to use Linux, while you can't use MySQL without using something from MySQL AB.

You are truly hilarious mate! Can't use MySQL without MySQL AB. ROFL! Who cares about them - it's SQL one wants.

You know, I was running my web business long time ago without any MySQL and I had SQL. It's called PostgreSQL. Or maybe I just dreamt it... Hilarious.

The saviours of the world: MySQL AB. Still laughing. Thanks!

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 9:08 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (6 responses)

You don't seem to recognize that MySQL built its code out of scratch. They sort of complied with the SQL standard, but SQL is a language, not a code base...

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 9:14 UTC (Fri) by bojan (subscriber, #14302) [Link] (5 responses)

> but SQL is a language, not a code base...

Wow! Thanks for the education - really didn't know that - wink, wink. Oh, wait - there is this thing called SQL92. What is that? A spec? Not written by MySQL AB? What? NO INNOVATION?

Seriously, you crack me up man!

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 11:30 UTC (Fri) by dlang (guest, #313) [Link] (4 responses)

this is also the first time I've heard that MySQL complies with the SQL standard :-)

hint: in many ways it doesn't

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 12:59 UTC (Fri) by bojan (subscriber, #14302) [Link] (3 responses)

Seriously for a second, many people chose MySQL in the olden days because it was fast and ACID was not an issue for the web. Fair enough, but claiming that an open source implementation of an SQL server was a revolution done by MySQL AB is a complete fantasy.

I learnt my open source SQL on Postgres, all before 2000. Never ran MySQL until I found myself in a job in 2004 where GBs had to be written into the DB fast and the product (not open source) supported MySQL.

So, any claims that MySQL was the only show in town is just nonsense. And that seems to be the main plank of his argument.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 17:58 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (2 responses)

So, any claims that MySQL was the only show in town is just nonsense. And that seems to be the main plank of his argument.

How much more absurd can those claims of things I said or implied get?

I never said MySQL was the only show in town. I never implied. On the contrary, I'm well aware of PostgreSQL. My own online gaming startup used it in the late 1990s. I also mentioned it in my position paper on MySQL's acquisition by Oracle and discussed it throughout the merger control process.

How can someone arrive at such an unbelievable conclusion? It's unfathomable. It calls into question someone's good-faith intention to discuss the issues on a reasonable basis.

What I said is that MySQL AB (the company, including the predecessor whose assets it acquired) developed MySQL, so someone wanting to use MySQL needed something developed by MySQL AB, while someone using Linux back at the time we used SuSE didn't need anything from Red Hat.

I repeat, for the few here who are reading-impaired and simply type before they read and think: MySQL AB was needed for MySQL. Not for SQL as a whole. That one predates MySQL AB, obviously.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 23:08 UTC (Fri) by bojan (subscriber, #14302) [Link]

Wow, you have outdone even yourself now. The big innovators at MySQL AB invented something that already existed!

PS. You know, I am actually being unfair to all those hard working folks that made MySQL an open source success. My hat off to them. Your arguments regarding this, however, are complete and utter nonsense. I am more then willing to agree with you when you present reasonable points of view. This doesn't qualify.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 2, 2010 6:06 UTC (Sat) by bojan (subscriber, #14302) [Link]

> What I said is that MySQL AB (the company, including the predecessor whose assets it acquired) developed MySQL, so someone wanting to use MySQL needed something developed by MySQL AB, while someone using Linux back at the time we used SuSE didn't need anything from Red Hat.

Yeah, you argument is without meaning. I can equally well argue this:

Someone wanting to run an open source SQL server didn't need anything from MySQL AB. Someone wanting to run Red Hat Linux had to get stuff from Red Hat.

These statements, just like yours, mean almost nothing. It's like saying that you need to go to Ford dealership if you want to buy a new Ford. Well, yeah, thanks Captain Obvious!

So, I can only conclude that you must think that somehow MySQL had some great innovation or invention that nobody else did. Well, it was just another implementation of an open source SQL server. And because web boomed and people needed some particular features of their SQL server, more of them chose MySQL than PostgreSQL. I already acknowledged that. It is a sign of success, of course, but not a sign of some revolutionary innovation that needs its own special monopoly to be successful.

Equally, Red Hat Linux was just another implementation of a Linux distribution (or should I say open source stack). Not Linux - distribution. It did particularly well because of particular features it had, as compared to other distros.

So, pretty similar stories at the beginning, really. After that, and because Red Hat decided to take the whole thing really seriously and to a whole new level, they started offering excellent support and expertise, provided world class certification program, which further strengthened their brand. And they succeeded beyond anything MySQL AB managed to do. Along the way, they innovated far more than MySQL AB ever did.

Without putting MySQL AB down for a second (as I said, hat off to hard working folks there), Red Hat are an innovative open source company. And completely dedicated to releasing everything they make as open source.

Your diatribe about some parasitic business model or some such sounds like a whole lot of sour grapes to me.

And this is even without addressing the nebulous claims that people should pay more for the same stuff, because some programmer somewhere may lose their job.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 8:32 UTC (Fri) by gnufreex (guest, #70396) [Link] (2 responses)

You are one who needs to adjust behavior because you are one who is trolling and bombarding people with decade-old Microsoftian arguments that have been refuted countless times over the years.

You conveniently disabled comments on your splog, and you are now spamming and CCing all over the internet. You link to your splog so people argue with you here. If you really want a debate, why don't you enable comments on your splog? You don't because you are a troll. And you use your splog as your cave. Allowing someone to challenge your reasoning right in front of your cave... that wouldn't fit in your spam tactics. You want your spam to be discussed on other sites. I guess you didn't got banned from Groklaw for nothing. Hope LWN does the same.

I'm not gonna feed you anymore. I've seen enough to be sure you're hired by Microsoft, and frankly, don't know why I doubted that in the first place. It is bloody obvious.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 1, 2010 8:49 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (1 responses)

I was looking really hard for some substance in your comment but couldn't find it. You replied to a comment of mine with which I provided lots of facts. Now you argue about my blog not having comments enabled and why that is the case in your opinion. As far as I link to my blog here, it's to facilitate further research by people who really want to find out. You also talk about another site (GroklXX) having supposedly "banned" me. I'm available for discussions, online and in real life when I attend industry events. Also, I've answered many questions on Twitter (@FOSSpatents). So if you wanted to discuss the issue, you'd have plenty of opportunities, especially here.

IBM/TurboHercules can be resolved with license; no need for abolition

Posted Oct 4, 2010 9:37 UTC (Mon) by Wol (subscriber, #4433) [Link]

It's more I think Florian has abandoned Groklaw. He gets called out too much by people who know what they're talking about.

I don't think PJ has banned him. But I don't think he's got quite the thick hide trolls need to withstand an army of troll-hunters. The regulars there have rather driven him off.

Cheers,
Wol

Red Hat invested venture capital for MySQL

Posted Oct 1, 2010 8:38 UTC (Fri) by mjw (subscriber, #16740) [Link] (12 responses)

> Red Hat is one of biggest reasons for success of FLOSS in general. They and Cygnus put it on the map. I would even argue that MySQL owes lots of it's success to Red Hat. If MySQL wasn't in RHEL, it would be stillborn.

And Red Hat was one of the venture capital investors providing seed funding for MySQL AB: http://www.mysql.com/news-and-events/generate-article.php...

No, Red Hat did not provide "seed funding for MySQL AB"

Posted Oct 1, 2010 8:44 UTC (Fri) by FlorianMueller (guest, #32048) [Link]

And Red Hat was one of the venture capital investors providing seed funding for MySQL AB

As a former MySQL AB shareholder, I know that this is wrong. If you look at the press release you linked to, it's from 2006. MySQL AB was founded and received its seed investments in 2001 (the year I became involved as well). It received some more funding that year as well as in 2002 from European investors. Then in 2003, Benchmark Capital (knowing for funding eBay, Red Hat, Twitter), Index Ventures (known for funding Skype), SAP's corporate venture capital group and others came in. Red Hat came in three years after those, five years after the seed round.

Red Hat invested venture capital for MySQL

Posted Oct 1, 2010 8:49 UTC (Fri) by gnufreex (guest, #70396) [Link] (10 responses)

Thank you. I knew that but forgot to mention.

But mentioning that wouldn't help anyways, Florian will invent (and maybe patent) some crazy explanation why Red Hat's money is parasitic and why MySQL AB would be better off without it. That happens when you feed the tr.. I mean lobbyist.

Red Hat invested venture capital for MySQL

Posted Oct 1, 2010 8:53 UTC (Fri) by FlorianMueller (guest, #32048) [Link]

Prior to your thanking him for this, I had already posted a correction. Red Hat came in very late (and owned a tiny share of the company), five years after the seed round. I doubt MySQL ever even spent one cent of Red Hat's money because it didn't have any significant burn rate at the time and was sold shortly thereafter.

Red Hat invested venture capital for MySQL

Posted Oct 1, 2010 9:01 UTC (Fri) by mjw (subscriber, #16740) [Link] (8 responses)

> But mentioning that wouldn't help anyways, Florian will invent (and maybe patent) some crazy explanation why Red Hat's money is parasitic and why MySQL AB would be better off without it. That happens when you feed the tr.. I mean lobbyist.

Ha ha. But yes, you are right as you can see above. The venture capital Red Hat injected into the MySQL AB company wasn't "seed capital", so obviously it is just more proof of the parasitic nature of Red Hat :) How dare they try to take a chance and invest in innovative companies and business methods! Well, it was worth a try to inject some facts into the discussion.

Red Hat invested venture capital for MySQL

Posted Oct 1, 2010 9:06 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (7 responses)

@mjw It's disappointing to see that you react so unreasonably to my clarification. It was *you* who said that Red Hat was a seed investor. You could have phrased it more broadly, then I wouldn't have had to correct it. It certainly does make a major difference whether a company places a seed investment in something when the risk is highest or comes in when the company is already in great shape and ends up being sold not much later.

Obviously it's not "parasitic" to invest money at any stage. It's just that coming in at that stage isn't when you can claim to be a co-creator of the product.

Red Hat invested venture capital for MySQL

Posted Oct 1, 2010 9:24 UTC (Fri) by mjw (subscriber, #16740) [Link] (6 responses)

Dude lighten up. We are just making fun of your impulsive desire to discredit Red Hat at every possibility you see. Don't focus on such small details like a different interpretation of the word "seed capital" to score some point. The point was just that they do invest in things like MySQL through various ways.

There are just many different ways Free Software companies like Red Hat invest and innovate in communities, projects, projects and their customers. They just like doing that by providing high value at lower costs to their customers by being a catalyst and amplifier of the free software ecosystem. They even invested in you by paying you to play lobbyist at a time when that did make sense. So yes, sometimes Red Hat makes mistakes, sometimes they invest time, money and resources in things that turn out not to benefit them and the community as much as we hope. But that is just what you do when you take risks and try to innovate and change an industry.

Red Hat invested venture capital for MySQL

Posted Oct 1, 2010 11:38 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (5 responses)

There is no "different interpretation of the word 'seed capital'".

You said a wrong thing -- why can't you admit it? Seed capital is clearly defined and not by any stretch of the imagination would a company founded in 2001, sold in 2008, be at the "seed" stage in 2006.

Red Hat invested venture capital for MySQL

Posted Oct 1, 2010 12:17 UTC (Fri) by mjw (subscriber, #16740) [Link] (4 responses)

> There is no "different interpretation of the word 'seed capital'".

Well, clearly there was, since you meant something different by it than I did.

> You said a wrong thing -- why can't you admit it? Seed capital is clearly defined and not by any stretch of the imagination would a company founded in 2001, sold in 2008, be at the "seed" stage in 2006.

Relax. You "won" your point. Feel free to just read that sentence without the word seed in there. So "Red Hat was one of the venture capital investors providing funding for MySQL AB". There fixed. The point was just a little addition to what others had said. Red Hat invested through various means to help make MySQL a success.

But this is precisely why we are making a little fun of you. You seem to want to win these grammatical/spelling error points over a wrongly used word in a sentence. While missing the bigger discussion completely.

Red Hat invested venture capital for MySQL

Posted Oct 1, 2010 12:23 UTC (Fri) by FlorianMueller (guest, #32048) [Link] (3 responses)

The fact that you meant something different by "seed capital" is irrelevant. You can mean what you want and that doesn't automatically mean that there are different definitions by even remotely reasonable standards. Better look up Wikipedia (deep link leads to relevant paragraph) or any other source.

I said before that seed makes all the difference here. When we talk about creating innovation, the critical stage is early, not shortly before trade sale or IPO.

Contrary to what you claim, I'm not missing the bigger discussion. We're having one big discussion here, and in some subthreads, individual aspects such as Red Hat's investment in MySQL get discussed.

Are you a Red Hat employee by any chance? I believe you (unless I confuse you for another nickname) identified yourself as one in some other thread. It would have been nice to do so in this context in case you are.

Red Hat invested venture capital for MySQL

Posted Oct 1, 2010 12:39 UTC (Fri) by mjw (subscriber, #16740) [Link] (2 responses)

> Contrary to what you claim, I'm not missing the bigger discussion.

Lets disagree, I think this whole discussion shows you are missing why you are wrong in the bigger discussion.

> I believe you (unless I confuse you for another nickname) identified yourself as one in some other thread. It would have been nice to do so in this context in case you are.

Dude, you are repeating yourself. Try to relax and not relive every discussion again and again. Yes, we discussed this already. You fail to identify who funds you and I was open about who funds what work I do:
http://lwn.net/Articles/402409/

Red Hat invested venture capital for MySQL

Posted Oct 5, 2010 22:38 UTC (Tue) by nix (subscriber, #2304) [Link] (1 responses)

It's been four days and Florian doesn't seem to have followed up to this post. Strange that.

Red Hat invested venture capital for MySQL

Posted Oct 6, 2010 3:56 UTC (Wed) by FlorianMueller (guest, #32048) [Link]

During those days there's been a lot of interest in the Android situation (firstly Microsoft's announcement of infringement action against Motorola, then also Oracle's reply to Google's suit) and I had to answer press questions and I wanted to analyze the overall situation on my blog, which was a lot of work (looking at all patents-in-suit from all three Android cases).

I'll try to answer questions to the extent they bring up new points that should be discussed. But I can't always do it as quickly as I'd like to.


Copyright © 2026, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds