40 Years of VM Community—and Going Strong!
The dawn of VM was a different time. In the early 1970s, VM/370 arrived, along with the first System/370 models. It replaced the relatively obscure CP/67 operating system, which ran only on System/360 Model 67 systems. But VM was joyously flexible and commercial: supporting a family of processors, actively marketed by IBM, and offering elegant cost-effective timesharing as an alternative to then-clunky TSO. Even better, in many ways it was a wide-open system (site- and vendor-customizable and -extensible, fully source-code based) allowing easy application and tool development, as well as all sorts of system modifications and extensions.
As large and small installations in diverse industries adopted it, an era of personal mainframe computing began. And even though mainframe user groups—SHARE, GUIDE and WAVV in the U.S., and others around the world—had existed for some time, VM’s proliferation and dynamic nature elicited increased synergies among customers and with IBM. This created local, regional, national and worldwide VM communities that made the system much more than just another vendor OS, and which remain vibrant and valuable today, more than 40 years after VM’s introduction.
From the Beginning
Since its initial release in 1972 (which I installed!), VM has had a community of kindred enthusiasts who’ve used, extended, supported, loved and evangelized for it. But VM-land was always something of a cult, and because it was undervalued, it never quite achieved mainstream status. Accordingly, it was often rumored to be—and sometimes was—on IBM’s cut list. The user community kept VM alive during those dark periods. Good thing, too, for where would today’s Linux on System z and mainframe-based cloud computing be without today’s z/VM?
VMers—often experienced on other platforms—have traditionally been open minded and interested in VM’s potential, rather than using it as shipped by IBM. Its virtual machine environment has long been a test bed for experimentation/innovation/enhancement. So it's no surprise that as the popularity of Linux grew, VM was the foundation for its mainframe birth and popularity.
VM, of course, was the mainframe’s first virtualization tool and platform. So it was adopted not just as a production tool with its Conversational Monitor System for personal computing and system automation, but also for hosting production and testing VSE, MVS and Transaction Processing Facility (TPF) work. VM and internal variants supported developing IBM’s other OSs and prototyping new architectures before they existed in hardware; PR/SM and LPAR facility are hardware implementations of some VM functions. So even sites not explicitly using VM benefit from its decades of use and evolution.
Any discussion of the VM community must begin with the epic paper by Melinda Varian, formerly of Princeton University. Titled “VM and the VM Community: Past, Present, and Future,” its 70 pages detail—with citations, quotes, photographs, technologies and battles won/lost—VM's epic history and evolution, from elegant but niche tool to today’s strategic platform for Linux on System z. Read it to meet the multitude who shaped VM into the resource it is today.
VM’s 2012 40th anniversary has been a time to look both back and ahead. It prospered in large measure because it was source-code based, and uncounted contributors saw it not as shipped but as it might be. Then came object-code only (OCO) eliminating the source for large chunks of VM, which IBM implacably pursued, in spite of prolonged fervent objections from the VM community. And VM survived multiple “Death of VM” rumors and attempts, in no small part due to user group presentations such as Guide’s “VM/370 in 1978—a Status Report” and SHARE's “Why VM?”
comments powered by