Thank you for helping maintain the Unix lore.

As the old joke goes:

"For your birthday, I wanted to get you a pocket calculator ... but then I thought you'd already know how many pockets you have."

> was Metafont the only outline-based font technology

Surely Karow's Ikarus was earlier than that.

One of the main innovations of Metafont was the use of "pen"s, so that one would describe a single path and the software would trace it and imitate the use of one or more pens, to end with an outline of something with thickness, and essentially more curves. It mimics how drawing and writing actually happens.

AFAIK, Zapf did not like this approach at all, as he was used to design typefaces the traditional way, by specifying all the curves. Richard Southall embraced the new paradigm and used Metafont as it was supposed to be used, but produced only a couple of demo typefaces (mainly the nmt family) and a handful of commercial ones (I can now only remember Colorado, with Ladislas Mandel, used in the phone directories of US West). I think he also implemented Melior, but of course this was never distributed as it was a proprietary Zapf design.

Note: all the above are based on recollections of my discussions with Zapf, Southall, and Knuth, in the distant past. All my relevant printed materials are in a different country right now, and I don't have easy access to them.

Yeah, _Digital Typefaces_ by Peter Karow was first published in 1986 (in German) and the company had been running for a long while before then, the software having been introduced at ATypI in Warsaw in 1975.

_TeX and METAFONT_ (the book) was published in 1979 (I still vividly remember checking a copy out of the local college library as a high school student in 1983) having its initial release in 1978, after being precipitated by the infamous second edition galley proofs on TAoCP 30 March 1977.

Ah, I suspected it wasn’t the first, it’d just been a long time since I dealt with some of this history. 30 years ago, I had the entirety of Gutenberg to Emigre at the front of my brain but much of it has vanished since.

RIP. I've actually found the Handbook of Programming Languages (4 volumes) that he edited much more useful than his Unix history.

I met him in a few conferences, back in the day. We ended up talking more about linguistics than Unix history, somehow.

They definitely also had NUMA (Non-Uniform Memory Access): on multi-CPU systems, you had a variety of ways to specify where you wanted your data to live (stay close to one CPU, for example).

People who look for solutions in this space (Time Series DB for measurements, for example) can also look at the good ol' RRDtool https://oss.oetiker.ch/rrdtool/ first released more than 25 years ago...

And thank you for this. In my professional setting, this is a very valuable addition -- provided it works correctly, of course ;-).

Now if we could also have comments inside the code ("BEGIN/END snippet by Copilot"), that would also be great!

I concur completely. Back in those days, the very basic stuff you mention (awk, sed, make) were being built by a handful of people, all sitting together, and the few outsiders who were submitting enhancements (even before these were called "patches") knew the email addresses to send these to. For Sendmail, you should contact the people at Berkeley, for most of the others you sent to Bell Labs.

Then software started appearing from other points. We were getting new versions of software after email announcements -- and later on, on comp.sources.unix -- and we were reading the comments to see that other people were contributing, too.

The way you publish your software (especially today) essentially boils down to how much you are looking for contributors vs. users (vs. no one at all).

Most of this is based on Copyright legal framework, which is surprisingly homogeneous around the world. The discussions about ownership of AI-generated material are exactly the same in EU.

Yes, and if the same come ends up in someone else's hands, they can state "we didn't steal it, a GenAI generated it for us, the same as it did for you". Given the non-deterministic operation of current GenAI systems (a major difference from compilers), it would probably be hard to prove either position.