Ten Years Ago - A Tale of Two Teams
I didn't realize that this is the 10th anniversary of Linux on the Mainframe until Meike Chabowski (Novell/SuSE) posted about it in the System z Linux discussion group on LinkedIn. The group, interestingly was started by Colin McCabe (RedHat), so ... it's all about teamwork, right?
In the summer of 1999, I was working for BMC. I got email from a certain friend in Australia. Two things were happening. Some guy in Austin was working on a from-scratch port of Linux to the S/370 (really XA, but did not require G2 or even ESA). Also, there were rumours about an IBM effort.
The guy in Austin was none other than Linas Vepstas. He had been doing Linux for a while and also knew the mainframe and had enough savvy to port and cross-compile the kernel plus a few essential utilities. (Linus, meet Linas. He's taking your baby to IBM land.) "Doing Linux for a while" is an understatement; Linas was deeply involved in some very exciting projects most of you have heard of by now (OpenGL, GnuCash) and is a math heavy. He was using graveyard shift time on an LPAR. Several of us VM aficionados introduced him to VM and PER, and progress got better.
Sometimes rumours turn out to be truth. We learned that IBM's Boeblingen lab was in fact porting Linux to S/390. Their work required the G2 instruction set which thankfully my P/390 was just barely sufficient to handle. (I say "my P/390"; I actually acquired it from Jan Ott, a brilliant colleague at BMC. More about Jan later.)
The Boeblingen port was released through Marist College. It was really nothing more or less than a tarball of a recent RedHat distribution. But it was sufficient for the time being and became known as "the Marist distribution". Marist is literally just up the road from IBM Poughkeepsie, land of the z and home of the ... [kind reader, fill in something funny here]. A limitted number of us got our hands on it early. Linux is Linux, and we knew Linux so we were totally happy with it. (We don't need no steenkin local X display; we don't need no steenkin installation tool.)
"They say" that within IBM, the independent work porting Linux to the mainframe was affirmed by the effort Linas was heading up. "They say" that it forced IBM's hand to either put up or shut up.
More intrigue follows tomorrow, same net time, same net channel.