Data migration requires disk, plus planning
Laboring Toward Support of SQL

A 3000 emulator needs HP's IP to boot

Last week we reported a couple of stories' worth of information about a new emulator effort for the HP 3000. This one couldn't be more different than the Stromasys Charon product that's now winning customers. We've gotten a heads-up that another Charon site will be going online to replace large HP 3000s, this one in the northeast at a financial services company.

Meanwhile, Piotr Głowacz in Poland is fronting a band of developers who are taking an open source approach -- using publically-available documents.

We did our simulator based 100 percent on publicity-available docs (which is typical for FOSS projects). We've reached the point where the simulator is running, going through the install process for MPE/iX and crashes at the very unspecified moments. As we can't provide an HPSUSAN number for testing, we're just hoping our simulator will do (and we're closer to our goal each day).

We’re still checking to see if Głowacz’ team means that they're getting closer to a non-crash startup for MPE/iX every day. It’s not clear why if they had an HPSUSAN number, it would that help in the testing.

BootWhat your community learned in 2003 was HP's help would be required to emulate PA-RISC processors capable of booting MPE/iX. There's a Processor Dependent Code routine or module that halted the Stromasys work for years. HP's intellectual property lawyers wouldn't cooperate, and the Stromasys development had to get shelved. Until 2008, when HP changed its mind.

Open source has its unique advantages. One of them is finding things available in public domain and modifying that source code to solve a problem. However, PDC has been considered a trade secret by HP. Getting documentation about PDC from a public source is going to be a tough assignment. Stromasys got the information by arranging for a top-down, official relationship with HP. That's led to an HP Worldwide Reseller agreement.

Not even the licensees of the source code for MPE/iX have those internal HP docs about PDC in PA-RISC.

In a commercial arrangement, HP's lawyers might be convinced that it's a good idea to make that data available. It's a leap of faith to imagine that arrangement taking place for an open source project. This project would be the first open source re-engineering of a processor for the HP enterprise user base.

People have ported open source tools. But to open-source an simulator of a proprietary chip, booting a proprietary enterprise OS, has never been done. That distinction of emulator and simulator is important.

If the open source team has a chance getting what it needs from Hewlett-Packard, it might start with Jennie Hou at HP. She ran the 3000 business group at the end, until there was no more group.

We don't know several other things yet. Are any of these developers experienced with MPE? Charon got release-ready when some MPE veterans joined the effort. And I also asked Głowacz "why do this, for a slice of the computer community that's so small?" It's clearly not a simple effort, as an volunteer open source project for a community the size of the 3000’s. The answer sounded like a line from a political speech or a play. “As for your question why, I'd like to answer as simplest as I can — if we're not for money, why not?” Glowacz said.

The line attributed to US Senator Robert F. Kennedy goes, “There are those that look at things the way they are, and ask why? I dream of things that never were, and ask why not?” In truth, the line comes from playwright George Bernard Shaw’s play Back to Methuselah. It was spoken by the Serpent in that play. 

“Why not” is noble and laudable. But there are expenses in most of the simulator/emulator projects -- if nothing else, there's the time of the developers and testers. Those are some reasons why not. Perhaps the nature of an open source project simply absorbs these real costs of labor and systems. 

There are worse places for the 3000 community to find itself, than to be the subject of an open source simulator. Whether it’s suitable for commercial use must still be proven — once the effort gets a version which can boot. The hobbyist part of the 3000 community -- most likely to sieze on a free tool -- is already served by a free limited user and horsepower version of the Stromasys Charon software.