April 24, 2009

Sun Set

To be honest, I probably could've just sent this out as a 140-char tweet and let it fade away into the Internet aether. On the other hand Oracle's acquisition of Sun Microsystems marks the end of a significant chapter in computing history, with some personal relevance, too. Not to mention it's been a while since I've updated the blog so a new entry seems warranted.

Thanks to family friends with Apple employee discounts I was indoctrinated early in The Macintosh Way, with mice and GUI and all that, which minimized my exposure to the joy of MS DOS. School's computers were still Apple IIs with some Commodore 64s with some PETs languishing in the back. Played around with BASIC programming on the 8-bit machines but that's hardly a real command line, nor real programming, for that matter.

Of course, I was super excited going off to college, with their banks of computer labs, speedy new 486 PCs in the dorms, and my very own e-mail account to keep in touch with high school friends. E-mail was hosted on the time-share system, but thankfully they offered the more friendly PINE for us newbies, instead of generic Unix mail. Nevertheless it was an introduction to the Unix command line system. Living amongst all the Tech geeks, it wasn't long before I was gobbling up the documentation and customizing my shell environment. Of course, this was all hosted on the mysterious but surely mighty SunOS servers.

When I got my foot in the door at JPL, the standard job of the summer undergrad intern is data processing, on Sun workstations, of course. It was so awesome to have a station all to myself, with a full X-Windows environment so I could run multiple terminal windows and maybe even a Mosaic instance on another virtual desktop. My mentor was nice and helpful on the actual scientific tasks, but I was expected to do the data-munging and other tedious tasks myself. Time to learn emacs to do the programming, sed and awk to crunch the data files, not to mention the specialized astronomical applications with their own inscrutable config files and scripts. The Sparcstation 20 was more than up to the task, however. I could even work from the dorm via a remote X-window session or Telnet into the shell.

By the time I got to graduate school, the transition to from SunOS to Solaris was complete, and Windows machines were infiltrating the halls of academia. The Sparcstation was relegated to e-mail server duty, but the Ultra 5's and Ultra 10's were still the dominant platform at the observatory, to run all the Unix-based programs. There was this new thing called Linux, but that's still way too much trouble to install and make work on commodity PC hardware.

Meanwhile at home I was still steadfastly holding onto the Mac, all through the decline of MacOS 9. But when Apple made the big jump to the Unix-based MacOS X, I was fully prepared for the transition thanks to my SunOS experiences, so much so that I made sure the Terminal would open on login, just like good old xterm. The Sun machines were steadily obsoleted and replaced by Linux on PCs in scientific computing, and just plain old Windows when I got a real job. And Apple has all of a sudden become the flag-carrier for Unix-on-the-desktop, as Sun fades away into the maw of the Oracle acquisition machine.

Both Apple and Sun have helped me keep an open computing mind and kept me from becoming yet another Windows drone. So here's to the original Stanford University Network and all people it touched, I amongst them. Hopefully Sun technologies such as ZFS will live on in MacOS X, and Oracle will continue making Sun hardware into the best database appliances money can buy, since Sun Micro is still a good customer of ours.

Posted by mikewang on 09:14 AM