Sloppyness smells also tobacco

Posted Aug 20, 2010 14:11 UTC (Fri) by FlorianMueller (guest, #32048) [Link]

A common perception is that TH is a company recently set up for the competition complaint as its business model. I share this view. You fail to convince me of your alternate teaching.

It's really hard to argue with someone who behaves so unresaonably in a discussion. Just the part I quoted contains again a number of unfounded and unsubstantiated assertions, untrue also. Why a "common" perception? Apparently "common" doesn't include the European Commission, the Wall Street Journal, the Financial Times and so many other media? You say me I fail to convince you of something. In the past when I failed to convince you of something related to software patents, it was actually better for the cause that I didn't care.

You just ask a bunch of question. I'll answer them, but I challenge you to substantiate your claims. You called the TH case a "staged" drama, and after repeated requests on my end you haven't given any explanation as to why. On the contrary, whatever I said concerning the details of the case and the players, you basically had to acknowledge and then disclaim that you meant any of that to be "staged". Still you haven't said what is staged in any reasonably specific way.

I'll answer these questions, but I refuse to play a game of having to bring up more and more facts that convince reasonable people only to deal with a, say, person who simply doesn't want to be convinced no matter what I say. I do this here not for you but for the readers of LWN anyway.

* When was the TH set up?

I don't know the exact date, but I found this on their 'About' page:
Roger Bowler, Co-founder
Roger created the Hercules mainframe emulator in 1999. He lives in Paris, France where he works as a senior mainframe developer and z systems operator. He serves as the company’s technical visionary and its liaison with the Hercules open source project.

* When and for what reasons was H published as open source

It was open source from the start in 1999.

(under the worst imaginable OSI approved license)

You say that from today's perspective but they made the choice a long time ago. It's not only OSI-approved but also FSF-approved by the way.

Proliferation of licenses is a general FOSS problem, can't blame such an old project for its tradition.

* Did TH seriously expect to get an operating system license?

There are two answers, and both of them I had already pointed you to, so your asking reaffirms my impression you don't want to understand.

1. I had given you the link to TH's offerings. That very page explains at the bottom that the operating system license is available, but currently limited to the use case of disaster recovery. Disaster recovery/business continuity is a pretty significant business model in IT by now. z/OS can be used for that purpose in accordance with the excerpt from the z/OS EULA as quoted on that page. TH indeed already conducts its business on the basis of being limited to disaster recover. There's a video available on that page showing a reference customer.

2. Apart from that, there's nothing unreasonable about the request for a license to z/OS in emulation. The European Commmission stated in its recent announcement that "IBM is alleged to have engaged in illegal tying of its mainframe hardware products to its dominant mainframe operating system." The Commission looked into these complaints for a long time and then decided to launch these antitrust probes. So reasonable people can indeed believe that it's fair and reasonable to ask IBM for a z/OS license on non-IBM hardware.

* Why did I.B.M. tolerate H for so long?

That's a question to put to IBM. I don't know the thinking of IBM executives. From the outside I can only try to come up with plausible explanations. When we talk about mainframes, we talk about highly mission-critical purposes and customers in that area need a company to turn to for professional support and similar services. So having TurboHercules versus just having Hercules alone makes a positive difference, just like Linux would never have become as successful in enterprise use at it has without companies providing specialized support.

The other factor I would imagine is the evolution of performance. If you look at the total number of MIPS you can build into an Intel-based system now and compare it to a mainframe, the gap has never been narrower. Thus the time has come for Hercules to become an interesting choice for a significant number of mainframe purposes. The software is mature and the hardware became powerful enough. Great combination.

* Who is behind OpenMainframe?

If you mean the openmainframe.rog website, you have to ask them. I'm not involved with it. They link to mainframe-related articles from all over the web. I sometimes use that site as a starting point for my own research. Occasionally they link to my blog, like so many others do. The idea of opening up the mainframe market is a good one if you consider the vast size of the market and the way IBM milks its customers, exploiting their $5 trillion lock-in.

* Furthermore, why are you so passionate to refute it? Why the drama?

I'm always passionate about refuting things if I believe I have acquired enough knowledge to do so and if it relates to what I do. Since my blog discussed the TurboHercules matter on multiple occasions, I care and I have done a fair amount of research. There's no drama. It's all about the truth that must come out now during the antitrust investigation.

Why campaign against OFE?

I don't "campaign" against OFE because they're OFE. I call them out on hypocrisy, also in connection with the contradiction between Oracle/Sun's Java Community Process and OFE's demands for open standards. OFE misses the point by talking only about one particular definition of "open standards" when the real issue facing the IT sector is that some de facto standards have emerged because of pervasive technologies and the question is what to do about them. The European Commission, in particular its vice-president Neelie Kroes, is well aware of that and sets the right priority with this initiative about de facto standards, about which OFE is conspicuously silent. I would view OFE much more favorably if they truly advocated openness all the way. I would, however, always be skeptical of OFE's chief executive because he lobbied for softwaer patents in the name of the open source community.

Why is TH's complaint framed as an OSS issue?

When TH's antitrust complaint was lodged, I didn't even comment on it. When I obtained IBM's threat letter (at a time when some others in the community did, such as Eric Raymond), I saw the patents-against-FOSS angle there. Now concerning the antitrust case, it certainly is a David vs. Goliath, free software start-up vs. proprietary monopolist, type of battle. TH is a FOSS startup, started by the founder of the FOSS project in question. IBM, as far as the mainframe legacy business is concerned, is at least as closed as Apple and much more closed than Microsoft.

There is no reason for a small player (actually Opera is no SME anymore) to deny to have more powerful allies, except that suspected allies don't like the rumours because they get into trouble.

I agree that if it helps to make the truth come out, there's nothing wrong with little guys getting help from bigger one with an interest in the same matter. I guess that generally there's now a lot of anticipation in the IT industry among IBM's various competitors (thinking of hardware as well as software companies) in capturing a part of a huge market(twice as big as the Linux-based software market), and we'll see if any alliances and/or investments result from that. The antitrust probes were launched just before the primary vacation month of the year, so it's not surprising that nothing hasn't happened yet. We'll see.

Posted Aug 20, 2010 14:14 UTC (Fri) by foom (subscriber, #14868) [Link]