Oct 1, 2026, 1:14 AM@nyrath @n1xnx @TerryHancock
Oh yes, that's it to a t.
I never did more than one data card repair on the punch at a time, using DUPlicate for a many columns of bad card were readable.
For that client, if we hand keyed data it was on terminal (double keyed and compared).
Cards were only on mark-sense workflow.
So while I knew about tab control (from typing class obviously useful!) I never used it with punch.
Oct 1, 2026, 1:29 AM@nyrath @n1xnx @TerryHancock
I definitely used tab stops in IBM editors that considered files as virtual card decks.!
@n1vux @nyrath @n1xnx @TerryHancock I think one of the first ed macros I wrote was to insert a new line followed by 5 spaces and a c. You can probably guess which language I was writing in.
Oct 1, 2026, 1:31 PM@ianturton @nyrath @n1xnx @TerryHancock
Indeed!
Oct 1, 2026, 2:54 PM@ianturton @nyrath @n1xnx @TerryHancock
40ish years ago I was assisting one of my elders (semi famous) with his second ever C program.
I looked it over and asked him if his first language was FORTRAN, as his C had a pronounced 'accent'. (Yes)
I may have avoided the brain damage oft attributed to both BASIC&FORTRAN (a) by the TA teaching flowcharts being clued on evolving Structured Programming (b) by doing both (& FOCAL) in quick succession after reading a PL/I book from remainders pile.
Oct 1, 2026, 2:58 PM@ianturton @nyrath @n1xnx @TerryHancock
I even had one FORTRAN ii program on the DEC EduSystem (after doing WATFOR / WATFIV dialect of Fort iv), the one with the way jump IF (no THEN, let alone ELSE). Weird!
@n1vux @ianturton @nyrath @TerryHancock
When I was a rocket scientist at Panametrics space instruments in the 90s, our program physicist Fred Hanser still maintained a set of FORTRAN programs he used for particle physics calculations. It wasn't that he couldn't have written new ones in a more modern language, it was that they "weren't broke" so he didn't fix them.
Oct 1, 2026, 3:48 PM@n1xnx @n1vux @ianturton @TerryHancock
I have a vague memory that in certain branches of science they rely upon subroutines written in FORTRAN which the scientists are forbidden to edit. Something about rewriting the subroutines in a different language can subtlety change the output past the decimal point.
@nyrath @n1vux @ianturton @TerryHancock
Spot on.
Numeric precision of FORTRAN calculations is, I believe, specified quite tightly. Just assuming that you can map to a 64-bit float and all's well causes subtle but significant things to happen. And if you've built a whole system based on the results the FORTRAN delivers, you're inviting Charlie to dance the Foxtrot.
@n1xnx @nyrath @n1vux @ianturton @TerryHancock When I was at atomic energy of Canada in 2005 as a high school student with the Deep River Science Academy some of our calculations called a FORTRAN77 fluid simulation library that opened with several pages of comments about its certifications and testing.
I imagine once you've spent the money to certify such things, you really don't want to spend it again if the old code still works.
@Canageek @nyrath @n1vux @ianturton @TerryHancock
I look forward to reading a news story about some AI-holic vibe-coding a translation of such a library and achieving an epic level of humiliating failure.
It's certain to happen; the only question is whether the attempt dies silently at the first verification test, or whether we read about, for example, the next-gen Raptor engine taking out its static test stand in a dramatic fireball.
Oct 1, 2026, 8:05 PM@n1xnx @Canageek @nyrath @ianturton @TerryHancock
Oh yes. And yes.
@arclight is the online authority on AEC-certified FORTRAN libraries wrapped & still in use down south in 🇺🇸.
(His and @nyrath 's tenure with SJG may have overlapped?
IIRC, Neither overlapped with my sister-in-law, art-dir for 1st gen GURPS)
Oct 1, 2026, 9:35 PM@n1vux @n1xnx @Canageek @ianturton @TerryHancock @arclight
Back in 1977, when I was in high-school, I did the artwork for Steve Jackson's famous table top war games Ogre. So I figured I was an artist.
But as a scifi fan, I wanted an accurate 3D star map. So from a cold standing start, knowing only the definition of tangent, sine, and cosine, I derived a polar to rectangular coordinate transform. This can turn a star's right ascension, declination and distance into x,y,z starmap coords.
Oct 1, 2026, 9:48 PM@nyrath @n1xnx @arclight
IIRC Steve mentioned OGRE! when he introduced us at PoGo that once.
(I had similar trig mania, but for Fuller Geodesic Domes, and later on, goniometers of all sorts #instrumentsOfMathsInstruction ).
(I cheated mathswise on my 3D star chart: inscribed spherical coords (RA & Dec) on 6" craft Styrofoam sphere and inserting measured pipe cleaners for range to stars.
Also tried Holmes stereoscope with chart views plotted with T/NROFF on a Diablo Daisy wheel printer.)
Oct 1, 2026, 9:52 PM@n1vux @n1xnx @Canageek @ianturton @TerryHancock @arclight
The trouble was, using my fancy scientific calculator, it took about 15 minutes to calculate each star. And I had hundreds to process.
Meanwhile it was time to sign up for college. I figured I was an artist, so I became an art major.
Oct 1, 2026, 9:59 PM@nyrath
My darling son-in-law tried Art School; took him a couple terms to determine whatever they were teaching wasn't what he called art!
(It's nice to have someone else in the household who understands symmetrical vs preceptual color solids! He's found a career that makes use of his aesthetical vision, he's a haberdasher now.)
Oct 1, 2026, 10:01 PM@n1vux @n1xnx @Canageek @ianturton @TerryHancock @arclight
In college was my old friend Tex. We were in high school together. We ruled science fiction class, because we knew more scifi than all the students and the teacher.
One day in college I complained to Tex about taking 15 minutes to calculate each star. He asked for the equation, and scribbled something. It was a Basic program.
Oct 1, 2026, 10:18 PM@n1vux @n1xnx @Canageek @ianturton @TerryHancock @arclight
"I cannot use this! It is a computer program, only usable by people with big brains" I complained.
Tex said "Nope", lead me to a DecWriter, and showed me how to type into the computer the program, the data, and "RUN".
It printed out the 3d coordinates about ten per second.
I was in love!
Oct 1, 2026, 10:20 PM@n1vux @n1xnx @Canageek @ianturton @TerryHancock @arclight
Alas, it was too late to change my major. I took as many computer courses that I could squeeze in, but my degree is in fine arts.
Oct 1, 2026, 10:04 PM@nyrath @n1xnx @Canageek @ianturton @TerryHancock @arclight
Secret from 1950/60s EE sliderule guide, a sliderule with a very specific set of Trig scales can do Rect ↔️ Polar conversion in one setting, faster than a calculator. The needed cheat is adding 1 mentally on B scale to find scaled hypotenuse.
Oct 1, 2026, 10:37 PM@n1vux @n1xnx @Canageek @ianturton @TerryHancock @arclight
I do remember being annoyed. While glancing through a calculus textbook, I stumbled over the very equations I had derived for polar to rectangular coordinate transfer.
Oct 2, 2026, 3:01 AM@nyrath that's confirmation!
You got it right from first principals! ⭐️
In my maths world, that's GOOD!
(Those same equations are in a lot of textbooks and reference tomes. Possibly not astronomy tho, they're so wedded to their mislabeled spherical coords. They convert among spherical origins and axes and supposed inertial references, don't seem to believe in right angles! 🤷♂️)
@nyrath @n1xnx @n1vux @ianturton
Yeah, I think Fortran is still king when it comes to pure number-crunching.
@n1vux @ianturton @nyrath @TerryHancock
Fred also had two-foot-high stacks of papers on his desk, and could find EXACTLY the document he needed within those stacks in seconds.
We also had a 20-foot wall shelf we called "the wall of Fred" with nigh onto 100 years pf Physics Review Letters and other journals. He still referred to them, and could likewise immediately find a specific article from 50 years ago that had the calculations he needed.
In a very rare combination with that polymathic genius, Fred was also a fully functional social human being and an extremely nice guy.