You've reached my desktop machine.
I've spent most of my career as a software entrepreneur but for the last three years, I've been enjoying a completely unexpected late but unbelievably satisfying career as a lecturer in electrical engineering.
What follows is a sampling of the fingerprints I seem to have left on the web.
My tiny software business, Hamilton Laboratories. I’d always wanted to start my own business and create something all my own.
In 1987, the time was right. I was 36, I’d just finished my MBA (a 2-year program I did in 3 years while working fulltime), we had our first kid on the way and I wanted to work from home. I spent a year planning my escape, lying awake many nights, worried about all the ways I could fail. To my surprise, that ended the day I turned in my notice. I never again had another sleepless night (at least, not over this.) The 15 months I spent creating the first release of my product and perhaps the next 10 years following turned out to be most idyllic and satisfying of my life.
My product, Hamilton C shell. I loved operating systems and compilers (but no one had ever let me work on a compiler) so this was just the perfect choice for me. Today, my C shell is a very stable product but I still enjoy working on it as the occasion presents and I continue to release new versions. (Yes, it works fine on Windows 10.)
My C shell on Wikipedia. It's notable!
My solution (1995) to the multiple reader/single writer problem. I'd already been using it for seven years, but that was still in time to make it one of the earliest published C language solutions I've been able to find. It's also the most performant, assuming contention is low enough to avoid starvation, requiring readers to take and immediately release a single lock to acquire or free the resource; writers simply take the lock to acquire and release to free. Other early publications I know of were the GNU (1997) lock, which refuses new readers if writers are waiting but requires both readers and writers to take two locks and immediately release one to acquire, and the Zimmerman (1999) lock, based on mine, similarly modified to avoid writer starvation, which requires that readers take and release two locks and writers take two and release one to acquire.
My elevation in 2007 to Senior Member in the IEEE. I’m about 1/3 of the way down at line number 115 in the Excel spreadsheet for that month. This is how they announced it. But the plaque is very nice.
My PE license in Texas. My father insisted I should get it done soon after college while the all material on the exams would be fresh. That part was good advice, but I’ve never done anything where it mattered if I was registered until now, as an instructor, where I need to set a good example.
My first patent, for a display processor I built for IBM’s early standalone word processing machines. The entire processor took 10 “Dutchess” chips, each with 100 NANDs and 34 NORs. Believe it not, it had a real instruction set. I went to IBM out of college because they were the only ones willing to let me try hardware design.
The patent I’m most proud of, listing me as sole inventor of the extraction technique I used in the ranker I wrote for Microsoft’s search engine. It ran against every matching page, generating blizzards of statistics to feed my heuristics for comparing a query to a page. Early on, our project architect had me terrified that my ranker would single-handedly blow us out of the water on perf. By the time we went live in Jan 2005, it was taking less time to find the 10 best pages out of 5 billion than it was to generate the snippets to go with them.
Missing seems to be most of my online writing from the entire decade of the 1990s when I moderated the OS/2 and Windows discussion groups on BYTE Magazine’s BIX network. I wish I could point to some of the articles I posted there discussing RISC, threads, why OS/2’s “Better Windows Than Windows” strategy was doomed, etc. All of that appears to be gone forever.
Speaking at my 35th reunion at Stanford in 2007. Go to the iTunes store and search for “stanford class of 1972 panel discussion.” It’ll be the first item, a free download: “Class of 1972 Panel Discussion: Hell No, We Won’t Go (Quietly Into the Night).” Or just click here. This improbable panel included a chief justice, an ambassador, a musician, a man building a spaceport, a geneticist worried about global warming and me. I’m introduced at 29:37.
In the news for stirring up debate on Wikipedia by proposing that the article on "cannabis" (the drug) be renamed "marijuana" as the more common term based on search engine statistics. (It failed.)
My Amateur Extra license KD1UJ.
My experiences in the late 90s with some difficult surgery and LSR. (You have to find those URLs yourself.)
A short autobiography on
my Hamilton Labs site,
finally updated once again out-of-date.
Arguing with the neocons on the WSJ site.
Commenting on the RealDVD litigation, the spoliation issues that arose and about the court’s preliminary injunction order posted to attorney Benn Sheffner’s Copyrights & Campaigns blog. Turns out I was right.
Offering my first impressions of my Pardini free pistol, posted to a popular forum for discussion of international-style shooting, the kind of events held at the Olympics. The guns are quite exotic, as you’ll see in my photos, and probably not like anything you’ve seen. At times, I’ve been a very active club-level competitor in everything from air pistol to high power rifle. I’m also an NRA-certified firearms instructor and (as of a few years ago) one of 3 women training counselors in Washington.
Inciting trouble by asking, “Why are shooting events still sexist?” The last time I checked, this was still the longest-running, most contentious debate thread ever in the history of that forum.
Establishing my bona fides as a car nut with a trip report on the Euro delivery of my BMW 540i/6 in August 2000. I was shifting into 5th at 80 and into 6th at 110. Here in the US, the most fun I’ve found so far is drifting the hairpins on Highway 1 through California. The turns are so tight, you can get the slides going at surprisingly slow speeds. By the end of a day on Highway 1, my arms are tired from turning the wheel, my feet are tired from braking and shifting and my face is tired from having been contorted into a silly grin all day long. Too bad it’s only safe to drift the right turns and only going south. (Think about why.)
My very first post to USENET on Jul 18, 1985, discussing the limited selection of CDs. On Jul 23, I decided I’d go out on a limb to explain how oversampling works. I was pretty sure I was right, but I can remember being terrified at the possibility that if I wasn’t, easily tens of engineers worldwide would point that out over the next week or so as my message slowly hopped from one machine to the next and their responses slowly hopped back. See, kiddies, what you say online can follow you forever.