Hypocritical much, PJ?

The reason that there hasn’t been a new post to this blog in a year is that nothing has happened with IBM and Hercules for a year. Behind the scenes, we’re taking some steps, but those will be announced when they happen.

In the meantime, though…

I was reading a post on Groklaw about Oracle vs. Google, a software patent lawsuit. This line struck me:

Of course, even one patent is capable of doing harm, which is why software and patents need to get a divorce.

I agree, PJ. Why, then, are you defending IBM’s patent attacks on Hercules?

Why this is a patent attack on Hercules

A common refrain from the folks who support IBM in this fight is “They’re attacking TurboHercules SAS, not the open source project itself”. This is not correct.

TurboHercules SAS’s product is services, support, and optionally hardware around the open source Hercules emulator. They have no code of their own. They do provide binaries to their customers, but those binaries are compiled from the same source code that anyone and everyone can download for free from hercules-390.org. All of the emulation technology is provided by the open source emulator.

Given this fact, IBM’s claims that TurboHercules SAS’s product is infringing can only apply to the open source Hercules emulator itself. There is nothing else there that can infringe.

IBM’s attack, therefore, is a patent attack on open source software, something they explicitly pledged not to do for two of the patents in question, and even if those two patents are removed, the larger threat sill remains.

Whether you agree with TurboHercules SAS’s choice of Windows Server 2008 as the OS of choice on their turnkey solution (and, for the record, I would have chosen Linux), the fact remains that they are a user and distributor of open source software. Attacking a user of open source software in this way can only serve to discourage others from adopting and using it.

Those who argue that IBM is justified in attacking TurboHercules SAS in this way should stop and think. The ones doing Microsoft’s work here are not TurboHercules SAS or myself; it’s IBM and their supporters.

PJ vs. Hercules

PJ’s now posted, further:

Someone has sent me the page you link to, one you put up just to criticize Groklaw and to insult me personally.

No, PJ, it’s not all about you. This blog exists to defend Hercules against the patent attacks from IBM. It just so happens that most of what I’ve been moved to comment about so far has been your one-sided, IBM-can-do-no-wrong statements attacking TurboHercules SAS and me for daring to disagree.

Now I am VERY glad I returned your money.

As you wish; however, I still believe that you provide a valuable service, and would genuinely like to further that service.

I see know you didn’t come here to help us, Jay. You came to do what you are doing now, I believe, and in fact others have been warning me that this was your plan, to set me up, to try to create a phony “division” in the FOSS community.

Horse exhaust and hogwash. I genuinely came to Groklaw to try to help, by sharing what I knew firsthand. Unfortunately, you appear to think that, despite IBM’s own statements and actions, that they’re a total, unconditional friend of the open source community, and therefore nothing they do can in any way be construed as an attack on that community. The facts in this case are different, no matter whether you agree or not. As has been said in other contexts, “You’re entitled to your own opinions. You’re not entitled to your own facts.”

And now, what a coincidence, you sound like prior SCOfolk and Microsoft helpers on the subject of Groklaw. It’s not good to sound like Darl, you know. But it’s your choice.

I refuse to accept your attempt to smear me by association. Then again, you’ve been doing it since the very first words you spoke in reply to anything I said, so why should this be any different?

I am my own person, PJ. I speak only in my own interest and, I feel, in that of the open source community. If I sound like those on the side of SCO, it’s because I’ve been attacked as though I were – even though I most assuredly am not. As far as I’m concerned, SCO should sink beneath the waves and Darl should have to personally reimburse those he’s wronged.

All that happened is we don’t agree with you. We certainly listened to you ad nauseum, as the many, many comments addressing your comments in detail attest. So your accusation is false. We don’t think you’ve told it like it is, and that is our opinion after listening to what you had to say.

If “listened to you” means “cherry picked the facts and used them to slam you for being a Microsoft shill”, sure. The unbiased observer will see it rather differently.

It is you who have ignored the community, which did actually try to help you, and which doesn’t see IBM as being the problem in this fact pattern.

No, it’s that part of the community that believes IBM can do no wrong, not the community as a whole, that has twisted my words and ignored inconvenient facts to try to paint me as a Microsoft shill and SCO supporter, when in fact I am neither.

You’ll probably say this is a personal attack. That’s also nonsense. I’m attacking your actions, which show a determined, repeated, clear-cut bias toward IBM and away from the open source community when you must choose between the two.

You’ve decided that I’m the bad guy, and in the process, attacked an open source project. Are you proud of what you’ve done? I’m sure you are. Should you be? In my opinion, no. You’re standing up for software patents just because your favorite company has used them in an attack on an open source project.

In the process, you’ve shown to any truly impartial observer that you do not deserve the geek cred you earned from following the SCO case any more.

Update: As just one example of PJ cherry picking facts, consider this. An anonymous commenter posted:

Here’s where you’re wrong:

From what I’ve read, TurboHercules doesn’t develop the software. The Hercules project does. TH sells support for the Hercules software.

Assuming this is true (there is nothing I’ve seen anywhere to show it isn’t), the patent attack CANNOT be targeted at TH, but only at the open source project.

Saying this is not saying that one agrees with TH’s decision to file an anti-trust complaint. That is a different issue entirely. The only thing that is relevant is that asserting the patents against the open source project will almost certainly encourage Microsoft to do the same against FOSS projects.

I wish that IBM and TH would sit down with SFLC and figure out a way to have their dispute without bringing patents into it.

Realize that:
* Intel and AMD also have patents.
* VMWare, VirtualBox, VirtualPC, Qemu, and Bochs probably tread on some of those patents. Maybe sponsoring companies have licenses to permit it and maybe they
don’t.
* Figure out where this is going. There is no possible way to spin this as anything other than bad news.

PJ’s reply, in full:

There is no way if you posit the “facts” as you have, but the actual facts are different.

Just what facts are different? The only one she could possibly be disputing is that TurboHercules SAS doesn’t develop the software. (The ones in the original poster’s last paragraph are beyond serious dispute.) How, exactly, does she know what TurboHercules SAS develops? How, especially, does she know better than I do? I’ve asked Roger Bowler what, if anything, TurboHercules SAS does to the Hercules source as published to produce the binaries they ship, and thee answer was that they use compiler options to take the unmodified source and produce their binary modules. That’s all. No changes to the source code are made.

I’ve said that to PJ several times, and yet she continues to say that TurboHercules SAS’s code is not the same as the open source product. If PJ had any integrity at all, she’d either retract those statements or else say just how she knows something that neither Roger Bowler nor I do about TurboHercules SAS’s product.

Booted from Groklaw

Well, what I expected has happened: Groklaw’s PJ has had enough of my disagreeing with her and objecting to having most of my words ignored and the few remaining ones twisted, and has revoked my Groklaw account:

So, you come full circle, attacking me and Groklaw. I assume that means you don’t wish to be a member of Groklaw now, taking into consideration that membership implies a desire to help out here?

As I said in my reply:

Well, I still wish to help out Groklaw, but it’s obvious that you, and the folks here, don’t want my help – and it’s been obvious ever since update #3 to this article. I was just too optimistic to see it.

That’s why I have my own blog: so I can tell my side of the story without having you ignore the majority of what I say and twist the rest to attack me. Enjoy the company of your fellow Kool-Aid drinkers.

I had to post that reply as Anonymous, since I was unable to log in.

The writing was on the wall when she picked one statement out of my long post, replied to it with a vicious attack on my credentials as an open source community member, and ignored all the rest of what I wrote. I can now only assume she did so because she found the facts inconvenient.

She is of course entirely within her rights to kick me out. Nevertheless, I think that she did so, after doing to me what she has, speaks far more to her true integrity and support for the community – especially those parts of it that dare to disagree with her in even the slightest detail – than it does to mine. In one comment, early on, the poster expressed pleasure that I had come to engage them, unlike SCO, which refused to do so despite repeated requests. I now understand why they didn’t, and cannot recommend in good conscience that anyone else who Groklaw decides is a mean, nasty, Microsoft-backed enemy of open source – whether the accusation is true or not – go there and try to set the record straight.

Nevertheless, I recognize that the site still has a role to play, as long as PJ’s obvious pro-IBM bias is understood up front. Accordingly, I’ve sent them a $25 contribution. We’ll see if they keep it. No, I don’t expect a thing in return.

Update: PJ refunded my contribution with no comment.

Update 2: Well, no comment directly to me in the PayPal transaction. She posted:

All right, Jay, what kind of hypocritical game is this? You just made a contribution to Groklaw financially.

I seriously don’t want your money, considering all the circumstances, and I want you to tell us why you sent it, given your remarks.

I posted the last paragraph of this posting in reply, and added:

Despite your accusations that you don’t think I told it like it is, and despite your continued twisting of my words, I still believe that. My contribution is yours for the asking, no strings attached.

It is, too.

What I’d like to see from IBM

One question I’ve been asked lately is “What do you want from IBM, anyway?”

Here’s what I’d like to see:

1) IBM issues a direct, unequivocal statement that they will not pursue Hercules on patent grounds. This would go farther toward defusing the current controversy than anything else. It’s been argued that they’ve done so by reiterating their patent pledge, but in light of their other statement openly questioning TurboHercules SAS’s motives and claiming that those doubts entitle IBM to carve out an exception to their pledge, they need to do more.

2) IBM agrees to license its software on reasonable, nondiscriminatory terms on Hercules. This doesn’t mean they need to support Hercules, nor does it mean they need to fix apparent problems with their code that cannot be reproduced on real IBM hardware – just as was the case before IBM drove its competition from the market – or that do not clearly depend on IBM hardware behaving in ways that do not conform to IBM’s published architectural documentation. It also doesn’t mean that IBM can’t charge a reasonable and fair price for it.

3) IBM provides low-cost licenses for developer editions of its software systems, on a completely non-support basis, to anyone. I’ve been arguing for this for over a decade. They’re already building the package (called the Application Developer’s CD, or ADCD) for their PartnerWorld for Developers program; all they have to do is distribute it. I’ll even volunteer to run the licensing operation for it at no cost to IBM, along the same lines as the OpenVMS hobbyist license program, if they wish. I’ve also suggested in the past that the user group SHARE do so. This would help to revitalize SHARE, as well.

4) IBM provides complete interoperability documentation for its hardware to any and all, on similar terms to those that apply to Microsoft’s technical interoperability  documentation.

Points 2 and 4 are not new to IBM. That’s how they operated under the terms of the consent decree, and they thrived. IBM’s complained about investing billions of dollars into z/Architecture – yet they did so with System/360 originally, and we all know how well they did from that investment, which was generally regarded as a bet-the-business proposition. If they can invest five billion 1965 dollars in System/360 under the terms of the consent decree and do as well as they did, surely they can handle operating under the same constraints while investing five billion 2001 dollars in z/Architecture.

IBM the monopolist

My previous post invoked antitrust law to explain why forcing IBM to license its mainframe software on Hercules is not the bad thing some have painted it as. I’d like to expand on why it’s an antitrust matter.

Put simply, a customer who uses an IBM mainframe cannot escape dealing with IBM.

To see why, you need to understand that the IBM mainframe computing environment is fundamentally different from that of the now-ubiquitous Wintel or Linux server. It’s not at all uncommon – in fact, it’s the usual and expected case – to have custom, in-house-written code that dates back 40 years and for which the source has long since vanished. This is possible in the IBM mainframe environment because IBM, to its credit, has been religious about maintaining backward compatibility all the way back to the original OS/360, announced in 1964.

That much is good for IBM’s customers. What’s bad for its customers is that they’re now faced with a choice: deal with IBM, or face ruinous expenses to change platforms – with the success of such an effort very much in doubt. The landscape is littered with the stories – and the corpses, in many cases – of companies that tried and failed to get away from the IBM mainframe platform, spending immense sums in the process to reengineer their entire corporate data processing infrastructure only to find there are parts they can’t recreate and can’t reengineer.

This wasn’t a problem as long as IBM was subject to the terms of a landmark consent decree originally signed in 1955. Under the terms of that decree, IBM shared its technical documentation with all who wanted it, and licensed its software – that is, once it began selling software in the 1970s (previously, all of it was freely available and in the public domain) – to any customer willing to pay the money. If the customer ran IBM software on a non-IBM processor, they would first contact the vendor of that processor for technical support, and only if the problem could be recreated on an IBM system would IBM fix it. This was the way things worked for the remainder of the 20th century.

In 2001, IBM was released from the consent decree. They immediately moved to consolidate their control of the market, driving the two remaining hardware competitors, Amdahl (which was owned by Fujitsu by then) and Hitachi, out of it by raising the cost of licensing its upcoming 64-bit technologies to the point that remaining in the market became uneconomical. It then unilaterally terminated its licensing of its software on the one emulator platform on which it had done so, Fundamental Software’s Flex-ES, leaving hundreds of customers in the lurch. It completed its takeover of the market by buying up Platform Solutions, Inc., making an antitrust lawsuit go away in the process.

Those are all classic tactics of the monopolist. The Bush Administration’s Justice Department ignored IBM’s slap in the face. The European Union’s competition authorities never got a chance to rule on any of those complaints.

TurboHercules SAS’s antitrust complaint at the EU is being painted as Microsoft getting back at IBM for its own antitrust loss there. Quite aside from the complaint being pressed by a small company with no ties to Microsoft, this seems like a chronic speeder complaining about speed traps. If IBM doesn’t want to get hit with antitrust complaints, it has a simple remedy: it can behave like a company that wants to remain within the law.

“They want to force IBM to change its license!”

One of the most common misapprehensions about IBM’s attack on Hercules is that it began because TurboHercules SAS asked IBM to license its software on the Hercules open source emulator. The complaint is that “TurboHercules wants to force IBM to change its license! This is a threat to the GPL!”

Uh…no.

There’s a crucial difference here. Unlike anything GPLd – or under any other license in the open source world – IBM’s mainframe software license is a weapon used by a monopolist. That places it under the control of antitrust law. In the antitrust world, there’s a term for requiring that a customer buy one product in order to be allowed to buy another: tying. It’s illegal.

If it weren’t for that, I would agree wholeheartedly that IBM could license its mainframe software in any way it chose. However, antitrust law changes the game.

This is only a threat to open source software if, somehow, open source software were to gain monopoly power in a particular market. I just don’t see that as a possibility, because it’s antithetical to the very nature of open source. Almost by definition, an open source project cannot gain monopoly power. If it did, all it would take is for someone to fork it, something provided for by every Open Source Definition-compliant license.

The complaint that it’s all about forcing someone to change their licensing terms ignores inconvenient facts. That’s the hallmark of one side in this debate, and it’s not the one I’m on.

Whither Groklaw?

I was inspired to start this blog in large part by the treatment I received over at Groklaw. The site rose to prominence for its coverage of the seemingly interminable war SCO waged against Linux, in the form of its lawsuits against IBM, Novell, Red Hat, and others. The site, and its owner Pamela Jones, better known as PJ, tackled the monumental task of making the whole long, drawn-out saga understandable to people who knew a lot about computing and little about the law. In the process, it earned a massive amount of geek cred.

Even after reading the first grossly slanted story posted there about TurboHercules SAS, I thought that it was just a matter of PJ being misinformed about the facts. I posted a fairly long comment rebutting much of the story, and presenting the truth as I know it. PJ picked one tangential point I’d made and posted it as update 3 to the main article – and totally ignored everything else I’d written.

I should have taken that for what it proved to be: notice that PJ wasn’t interested in hearing facts that didn’t uphold her view that IBM could do no wrong, and that anyone who dared to so much as disagree with them was automatically a shill of Microsoft. I didn’t. I poured a lot of time into commenting there over the next several days, only to have my facts ignored and my motives questioned. I thought I was engaging in a dialog with people who sought the truth. What I was actually doing was trying to convince a bunch of folks who all drank the same Kool-Aid.

No more. Instead of continuing to try to teach the pig to sing, I’m going to stop annoying the pig.

I’m certain the denizens of Groklaw will take this posting to mean that they’ve beaten me into submission, and that I’m running away crying. It’s nothing of the sort. It’s simply a refusal to try any longer to convince people whose minds are made up with facts that prove inconvenient to their theories. I’ve got better things to do with my time.

PJ will, no doubt, take this as an attack on her integrity, and probably revoke my account on Groklaw. That’s her choice to make, as it’s her site, and if she no longer wishes me to have an account there, then it’s entirely proper for her to revoke it. For what it may be worth, I do not subscribe to the rabid conspiracy theories claiming PJ’s bought and paid for by IBM, much less that she doesn’t exist except as a front for a committee of IBM lawyers. I think she honestly believes IBM can do no wrong, and simply ignores anything that might shake that belief.

Whatever the cause, I do not believe Groklaw can be looked at as an impartial reporter of the facts. It is a partisan forum for those who feel IBM can do no wrong to tout their beliefs, secure in the knowledge that the site’s credibility will give their words added weight – whether or not that credibility is justified any longer.

Another voice in the debate

Hello and welcome to IBM vs. Hercules. There’s a lot of misinformation being spread around the geek blogosphere concerning the recent brouhaha between IBM and TurboHercules SAS, and how it affects the open source Hercules emulator. Here, you’ll find my take on it, including the facts as I know them.

I’m the project manager of the Hercules emulator for IBM mainframe systems. Hercules, despite what some people claim, is truly open source software. It bears the OSI Certified Open Source Software service mark, owned by the Open Source Initiative. Theirs is the only opinion that counts on that matter. Whether or not you think the license Hercules is released under, the QPL, is truly an open source license doesn’t matter. OSI does, and that’s enough.

I’m not beholden to TurboHercules SAS, or OpenMainframe.org, or CCIA, or Microsoft, as some folks seem to think. I’m looking at this from the viewpoint of the open source developer staring down the barrels of a cannon loaded with patents and aimed squarely at me by a big corporation that’s been a good friend to open source in the past.

I’m going to say things that folks don’t want to hear. That’s just plain tough. I don’t care if you don’t like it. All I ask is that you don’t distort or misrepresent my word or my positions. I don’t think that’s an unreasonable expectation. Disagree with me all you want, but at least give me the basic courtesy of arguing against what I say, not what you want me to have said.