Projects in computing history - Jay Maynard

> Recent entries
> Calendar view
> Friends page
> User info
> Jay's web page

Wednesday, 30 October 2002


Previous Entry Share Next Entry
1407 - Projects in computing history

I'm working on a project in computing history for fun.



SHARE is the user's group for IBM mainframe systems. (Well, they say that they're for any IBM systems, but it's overwhelmingly mainframe.) They've been around since the early 1950s, and have been involved in many, many generations of computing technology from the perspective of the user.

In the mid-1960s, IBM bet the company on the introduction of a whole new line of computing systems with the radical concept that they should be software-compatible across the whole line: System/360. The system was, of course, fabulously successful. What's often lost in this bit of history is that the originally planned operating system wasn't delivered on time, or (at least the first several versions) very well debugged. OS/360 wasn't really usable until 1969 or so, four years after the delivery of the first 360 system. Nevertheless, it did eventually reach its potential, and its basic design and interfaces are still quite recognizable today as the basis of IBM's flagship mainframe operating system, z/OS.

OS/360, unlike nearly all commercial operating systems today, was both in the public domain and distributed in source form, so it could be freely modified and those modifications shared. (The slogan, "SHARE: It's not an acronym, it's what we do.", reflects this culture, which greatly predates the modern Open Source movement.) Copies still exist today, and Hercules, the open source IBM emulator for Linux and Windows that I maintain, run it quite well; in fact, better than the original systems. My Pentium III/500 laptop (the Armada M300 I've been trying to replace) runs OS/360 four times faster than the fastest System/360 machine ever built, and twice as fast as the fastest System/370 (the next generation).

OS/360 came in several flavors, to fit differing requirements and sets of machine capabilities. The most complex and full-featured version was called MVT; it only ran well on faster systems, but could run varying workloads better than the other versions.

In 1973, IBM was introducing System/370, which improved on System/360's price/performance, while adding significant new features, first and foremost the ability to provide virtual storage. (All modern PC operating systems do this, but in the early 1970s, this was quite a feat.) Recognizing that IBM was concentrating its efforts on the new, virtual storage operating systems, SHARE launched a project to provide ongoing support and enhancements from users who would remain with MVT even after IBM got away from it, This was called the MVT Project.

Next February in Dallas, SHARE will hold its 100th conference, and as part of that, they will celebrate the 30th anniversary of the MVT Project. I'm working on putting together a fully-tricked-out MVT system, with all of the representative mods that people used most, as well as making others available for folks to integrate as they might want. This is turning out to be a real trip through history. I'm digging through collections of mods from 1975 and 1977, and getting a real understanding of just how much people changed in the standard IBM code. I never got to do this professionally, as my entry into the field of IBM mainframes came in 1980, after MVT faded into history for all but a few shops. It's an interesting project, and I'm learning quite a bit (though how useful it'll be is another question...)


I need to get the project done by the beginning of February to have CD-ROMs ready for the conference. There's still a lot of work ahead, but it's coming along.

current mood: [mood icon] geeky

(1 comment | Leave a comment)

Comments:


[User Picture]
From:michaelmink
Date: - 0000

IBM, and finance footnotes

(Link)
Two interesting footnotes, closely related. IBM was selling computers like hotcakes in the 1960s, and their star salesman was none other than H. Ross Perot. There are a number of very funny stories about how IBM kept raising the bar for Perot, and Perot kept breaking records. Eventually, he left IBM to form the company that made him very rich. Another person who tumbled onto computer leasing as a major financial tool was Saul Steinberg, who ran his company up into a largish conglomerate, moderately powerful in the 1980s, before cratering a few years ago. IBM's computers didn't just make computer history; they also made financial history, too.


> go to top
LiveJournal.com