Comments by "Lawrence D’Oliveiro" (@lawrencedoliveiro9104) on "Why is @ on your computer keyboard?" video.

  1. Early ASCII had “↑” which was replaced with “^”, and “←” which was replaced with “_”. DEC’s assemblers for both PDP-10 and PDP-11 families used “@” for indirect addressing modes. The 18-bit and 12-bit machines, being simpler, just put an “I” modifier in the instruction.
    7
  2. Every computer from about the 1970s/80s onwards supports ASCII, except old IBM mainframes and other mainframes that wanted to be compatible with them. Certainly ASCII support has always been universal among microprocessor-based computers, as well as “minicomputers” from the like of DEC, DG and others. And of course if a computer was going to run a Unix OS, then ASCII support was a requirement.
    6
  3. I use “|” every day. It’s used to create command pipelines in Linux, and as part of union/disjunction operators in several different programming languages. “`” has a few different uses, like quoting reserved words being used as table/field/variable names in SQL, or indicating TTY output in Markdown. And how many people type “\” when they really mean “/”? I like to call that “DOSlexia”...
    5
  4. Here in 🇳🇿 we mainly use the US keyboard. But on Linux I can define a Compose key, which gives me access to hundreds of extra characters via mnemonic multi-key sequences. E.g. compose-lessthan-doublequote and compose-greaterthan-double quote for “ and ” characters, compose-l-slash for £, compose-slash-equals for ≠ and so on.
    2
  5. Remember that the seeds of the emergence of Europe from the Dark Ages came from Arabic learning, which in turn started from classical Greek texts that the Europeans had forgotten. Of course, the Islamic Golden Age came to an end once the fundamentalists got the upper hand ...
    2
  6.  @beeble2003  Using the grammar of some arbitrary human language as the basis for designing a programming language doesn’t seem like a wise choice. Look at what happened to Strachey’s CPL.
    2
  7. There is only one “pound sign”, and it’s “£”.
    2
  8.  @mattwo7  Pound signs not really relevant to the subject of pound signs?!?
    2
  9. They tried to trademark that symbol, as I recall.
    2
  10. Sounds similar to how “$” was used in DEC’s DCL. A program could read from standard input, but if that happened to be a batch file, then when it hit a line beginning with “$”, it would normally get an EOF indication instead of reading that line, which would remain for the command interpreter to read once the program had exited. There was a “$DECK” command that could be used to change this behaviour; its presence meant that EOF only happened on a line that said “$EOD” rather than just beginning with “$”. Also you could specify some other prefix character than “$” if you wanted. Interestingly, DCL used “@«filename»” as the command for “read batch commands from «filename»”.
    2
  11. “¬” (U+00AC NOT SIGN) was used for logical negation in PL/I. Being an IBM-designed language, of course it was going to depend on the EBCDIC character set.
    2
  12. LEO ... that’s going way back. Imagine a computer system created by a chain of tea shops! Because their chief statistician saw it as a great way to improve their business, and nobody else had a suitable solution, so they came up with their own.
    2
  13. In Unicode, I can find “≈” (U+2248
    1
  14. ALMOST EQUAL TO),
    1
  15. “≃”
    1
  16. (U+2243 ASYMPTOTICALLY EQUAL TO) and
    1
  17. “≅” (U+2245 APPROXIMATELY EQUAL TO), among others.
    1
  18. You mean “°”? I type that with compose-o-o.
    1
  19.  @dalen.tenney5210  The compose key is a standard *nix thing.
    1
  20. Here in 🇳🇿 we got rid of all coins smaller than 10¢ years ago. It didn’t lead to any extra inflation. Prices are still given to the nearest 1¢, but if you pay cash, then Swedish rounding normally gets applied to the total at the till. But most of us pay by EFTPOS, in which case what we pay is the price shown.
    1
  21. Pascal used a postfix-“^” character to indicate dereferencing. I used one compiler which allowed “@” as a synonym for “^”, which I thought was a good option for a bigger and more easily visible symbol. I think C made a mistake in indicating dereferencing with a prefix instead of a postfix operator. And also in using the symbol for mathematical equality to denote assignment.
    1
  22.  @beeble2003  It is just silly. Deciding whether operators should be prefix or postfix based on some grammar wank is just silly.
    1
  23.  @beeble2003  This is why, calling a programming notation a “language” is a misnomer, and leads people into traps like you have fallen into.
    1
  24.  @beeble2003  Think of the parallels with mathematical notation.
    1
  25.  @beeble2003  How little you know.
    1
  26.  @davidwuhrer6704  Chomsky’s formal language theory works for computer notations, not for real human languages. That should give you a clue as to the fundamental difference.
    1
  27. It’s actually a stylized representation of the Latin word “et”.
    1
  28. The recitation of the alphabet would end with “... X, Y, Z and per-se-and”. The “per-se-and” being this symbol. You can see where the name “ampersand” came from.
    1
  29. I think EBCDIC was introduced in 1964 with the System/360 machines. Interestingly, ASCII came out the same year. But IBM’s excuse for not adopting it was that it wasn’t ready in time.
    1
  30.  @mattwo7  Yet somehow 🇺🇸 ended up with a smaller gallon than the 🇬🇧 gallon.
    1
  31. The religiotypical will never understand those of us who are religiodiverse, will they.
    1
  32. 5:52 Note this is list template expansion, not string template expansion. The input is a list structure, and the output is a list structure. This is also part of the macro facility, which expands program text that is already in AST form (homoiconicity). This is a much more robust and powerful macro facility than the conventional ♯define that does string expansion, that you might be familiar with from C and other languages.
    1
  33. “` ...`” for command substitution in POSIX shells is kind of old-fashioned. The main problem is that the same symbol is used for both opening and closing the substitution sequence. The more modern equivalent is “$( ... )”, which can be easily nested, and also uses larger symbols that are easier to see.
    1
  34. He was careful to distinguish the two.
    1