I had to have a stern word with one of our crew for bigotry that blatant, directed at our Amish builders. He figured out he should keep it to himself, but still has go at them from time to time when he feels threatened by them in other ventures.
That's funny. The guy I hired to lay the concrete in our creamery was a guy in his early 60's, with beautiful french manicure, plucked eyebrows and long flowing blonde hair.
I did not ask Mike about his personal life. He certainly didn't have a problem with the "burly builders" on the job, and they did not have a problem with him.
Oh, and he and his "burly builder" crew did a great job on a very complicated pour, which had to be done in three phases due to several different zones of underfloor heating (walk-in cooler, incubator, make room and pack/ship room), and the loading dock having to go down an extra four feet. In fact, it was such a complicated pour that the first "burly builder" that claimed he could do the job ran off after he screwed up the measurements on the forms for the trench drain, and they had to be re-set.
The next guy that bid on the job wanted five thousand dollars more than Mike wound up doing it for, probably to make the payments on that brand-new F-350 dually he drove up in. Mike was a godsend when he drove up in his beat-up old F-150, and I already knew he'd been the primary on a big milking floor with an underfloor manure pit for a thousand cow dairy down in Cobleskill, so I asked him about that, and he showed a lot of insight into how to pour a complicated dairy floor with a lot more interesting drainage than ours, with its T-trench and 5 circular drains going down to the VTA.
Mike both solved and prevented a great number of problems due to his actual experience, and from the layout knew exactly what order to do the different floor segments in. Mr. Chargemore in his brand-new F-350 dually might have done just as good a job, but I know for a fact it would have cost at least 5 grand more -- and he might have done a worse job. It might have needed to be re-done, and he might have pierced the PEX tubing we'd already laid out for the underfloor heating, and it might have cost us a whole lot more due to that.
Mike and his guys did the pour right around the underfloor heating system, as revealed by the pressure test we ran after the concrete was set up -- and you could see how careful he was being dancing around the tubing during the mad scuffle that is a complicated pour. He didn't even break a nail.
I'd say your argument that "Hiring a homosexual to work with a group of burly builders is a bad idea" belongs down in that manure pit under that milking floor of the thousand cow dairy Mike poured with his French manicure, plucked eyebrows and long flowing blonde hair, Mr. "Hucker75."
OMG I still have the troff source for that paper, and all the code. That was a fun project, setting up the RPC client on a Sun in Ithaca to prompt remote machines to run my ODE solver and automagically asynchronously send the results back to the Sun. I was able to sweep initial condition space a lot faster that way, despite the network latency. It's pretty hilarious that, writing that paper in '87, I had this diagram of 'compute servers' with the remote Convex, Sequent Balance and the Intel Hypercube at Argonne and the Cray-2 in Minneapolis drawn in -- not to mention the VAX 9000 with an array processor across campus, calling the Sun RPC library in C to do the client-server communications.
(When the first version of XMosaic came out five years later and people were like "look! it browses files over the internets!" I was like, "So? I've been doing that with ftp and telnet and rcp and rsh for like a decade. I've written code to do that. I wrote a paper for a refereed publication with someone at Caltech while I was in France over ARPAnet. I'm supposed to get excited about browsing files?" But I did set up httpd and Xmosaic on my SGI box, and started making and little html files, just to see how it all worked. I had to admit, it was pretty nifty. I realized I could have done the whole RPC ODE project with a CGI script instead of Sun RPC. Today, I'd probably call it from a server-side python platform, and use Numpy for the numerics. Could be a good way to get up to speed with the Enthought suite. I might yet, lol!)
Anyway, the hardest part of the RPC ODE project back in 87 weirdly enough, was that there were different C and Fortran compilers on each of the remote and local Unix boxes, and doing mixed-language compiles (linking the Fortran solver to the C wrapper) was a little different on each, and not very well documented. I used m4 macros to make the source a little cleaner and consistent across platforms, but even 'make' was subtly different on each one. Plus, the CRAY was a 64 bit architecture, so single precision floating-point IEEE numbers on its version of XDR still had to be cast into double precision when shipping it back to the Sun 2. You can imagine how hilarious I found Microsoft's conniptions about how to deal with a 64 bit architecture like, just a few years ago.
Funny you should mention lambda calculus, as the CS department at Cornell was really theoretical, and yah -- the grad students there did mostly theory, code not so much. Oh well, missing out on all the fun, I guess. And, I was in charge of more iron and had courtesy accounts on a lot more big iron around the country, as a UNIX sysadmin than they could ever hope to use. The CS grad students had to like, apply for time on big iron and write grants to pay for it. Maybe that's why most of them did theory.
They didn't even have an undergraduate program at Cornell in CS until I was about to graduate with my BSc in Engineering in 82, and weirdly enough I would have had to transfer out of Engineering and into the Arts college to major in CS. And, as a new major, the engineering faculty doubted its value. No way was I going to CS in the Arts college after busting my ascii for three years in the College of Engineering! With a BA there would be nothing to distinguish me from an English major! Yikes!
Oh, well, thanks for looking up my first refereed publication, ForkBomber! Maybe I'll include the whole thing in my memoirs, with a data key holding the source. Never did release it anywhere. I'm pretty protective of my source code.
One new co-worker started going on about my 're-training later in life' and the 'big gaps in your resume'.
I disputed the first, having graduated on time (and under budget -- I came in to Cornell with a lot of AP credits, so I could go part-time my fourth year, which allowed me more time to work as a programmer) in '82, not to mention the three years off between my MSc and PhD in computational geophysics to work as a Unix sysadmin -- 4.2 and 4.3 BSD on a dozen different Sun 2 and 3 boxes I'd set up the 3 disk servers for and others as diskless nodes (first running SunWindows, then NeWS, then X), 2 Celerity boxes and an old VAX 11/750 I converted from VMS to 4.3 BSD using the Mt. Xinu tapes (thanks, Melinda!) oh and SysV and GL programming on our IRIS workstations. Having started programming at 15 in High School on a time sharing account our Computer Club wrangled with Hazeltine Corporation on their UNIVAC and going straight on to working on a TERAK networking it to the IBM 360 and the Phoenix array processor through a terminal connection (oh Kermit!) and programming the Terak in UCSD Pascal, the Phoenix AP in a slightly weird vector Fortran, and the 360 in PL/1 as an undergraduate (it's how I put myself through engineering school) I was curious as to what he meant by 're-training later in life' -- unless he meant the Unix systems, C programming and Bash scripting I started on a 4.2 BSD box over in astrophysics ("oddjob" -- you have heard of it, no?) in my first year of graduate school -- '82-'83.
Before I could dispute the second claim, he interrupted me with, "But of course you took time off to have kids..."
I said, "Sorry, no, I never did have any children. Too busy programming."
He was genuinely confused after that. He thought he knew all about me from my age and gender, and he didn't know a blessed thing. Poor dear.
So, I'd say your hypothesis has some merit -- that people do make assumptions based on their stereotypical views, and it's only through conversations like these that they discover any actual facts. I think in most cases, the conversation never happens: they just make incorrect, biased assumptions.
It's kind of a shame that they do, because it leads to a lot of misunderstandings. On their part.
I think it would be a lot easier for everyone if people did not make assumptions.
Beware the new TTY code!
