When Allegro's Stan Sieler announced he'd completed 30 years of employment at the firm, it seemed to spark our curiousity about how things have changed over that period for the creator of so much MPE software -- and parts of IMAGE/SQL, for that matter.
He joined HP in 1977, after working on Burroughs systems. Over the years both with HP, and then later, he’s left many fingerprints on the 3000 identity. He proposed multithreading that HP finally implemented for DBPUTs and DELETEs. Wrote STORE on the Classic 3000s, plus can see various aspects of MPE/iX because of his work on the HPE operating system [the MPE/XL predecessor using an instruction set called Vision] before he left HP. A lot of the process management stuff that was his code is still running today. Sieler assisted on Large Files. IMAGE/3000 on the classic systems has intrinsic-level recovery he designed. A week after he left HP, they canceled the Vision project and ported 95 percent of his work to MPE/XL.
Then came the Allegro work during the era when the 3000 division called the company Cupertino East: Jumbo datasets in IMAGE/SQL. Master dataset expansion. B-trees. By that time he was already in the Interex User Group Hall of Fame. We interviewed him for the Q&A in our November printed issue, and spoke via Skype. Stan used his iPad for the chat.
Second of three parts
How are you coming to terms with being really well-versed with a work that fewer people not only know about, but even use?
Yes, that’s a hard question. I know the two places I’d go if I wasn’t doing Allegro anymore. In both places I think I’d be applying knowledge I’ve learned. It may not specifically be MPE, but it’s things like being careful about maintaining data structures of filesystem and the users’ data. These are lessons we’ve learned for 34 years on the HP machine. I think as we get older, we ought to be able to go up the technical ladder. The problem is that there isn’t enough of a ladder, in most places.
What makes the higher rungs of the corporate ladder hard to reach for someone who’s as experienced as you?
I have a friend who’s a fellow magician, and a senior scientist at Apple. I eat in their cafeteria and we talk magic, and I look around and they’re all young enough to be my kids, except for a smattering of people. He agrees that Apple needs more older people, because we’ll point to things and say, “See this? That shouldn’t have happened. We saw that kind of problem 20 years ago. We’d know better than to do that.” Apple is one place I think I’d want to work, except I don’t think I could stomach their policies. I could see going to Google, too.
My dream job? Being CTO of Tivo. They have the best DVR, and it’s crap. But everyone else’s is worse. It’s so easy to look at theirs and say they could do this and this better, and they haven’t. I’d like to improve it, so I could use it. It’s a lot like the 3000. A lot of the things I’ve helped push over the years are things that I wanted: The ability to properly handle bigger disk drives, and things like that. But sometimes you don’t get your way
What is the current mix of MPE work in your week, versus all other work?
It varies from day to day, and sometimes it’s hard to tell, because there are a few things that I do that run on MPE as well as HP-UX, three or four products plus a couple of internal tools that run on both platforms. There there are things like Rosetta — where all of the work is done off the 3000, but it’s supporting reading from STORE tapes, so it’s 3000-related. But definitely more than half of the work I do is off the 3000. We’ve got a proposal or two out to enhance our 3000 X-Over tape-copying product for them, and then we could use the enhancements ourselves. We’ve identified a relatively major new feature we could add.
People can see Apple seems to be losing its steam. Does it seem like an echo of what happened to HP and the 3000 in the late ‘80s and early ‘90s?
I remember when HP was first abandoning the 3000 in favor of Unix. To some extent, Apple’s doing the same thing with touch computing and changes for the interface on the Mac to be more tablet-like. Also, Apple is putting in more restrictions on what your apps can do if you buy through the App Store. In terms of hardware, Apple’s very quick to roll things over. At least with the 3000, things tended to be backward-compatible for a lot longer time. You didn’t really have a problem with a new version of an OS using more resources and rendering the older machines useless.
On the flip side, a lot like the 3000 came out with the A-Class machines and you’d ignore their crippling, that was pretty cool hardware. Apple’s doing the same kind of leap. I used to bring people into the office to see my Mac Pro, and I’d say, “this is technology aliens dropped down to Earth, it’s so advanced.” The Mac Pro black tower? That’s a more advanced race of aliens.
HP’s been through more than a decade now of no futures for the 3000. How much worse has that been for you than the decade leading up to the 2001 announcement? What have we really lost?
It’s a lot worse in terms of number of customers, income, the ability to fund doing interesting projects. That’s why we’ve branched out to do other things.
Do you handle Unix support calls, for example?
Allegro does. I tend to get pulled in when they’re hard problems. Same thing goes for the 3000, where I tend to not handle the frontline call.
For next time: redefining 3000 migration, Allegro's emulation considerations, and how practicing magic can impact a tech career.