Clearly Better: A thread on superiority in ELEMENTARY systems of notation, nomenclature, etc

Note the loud emphasis on the “elementary” modifier in the thread title! This is not a thread for arguing about arcane design choices in super-esoteric contexts.

This is about systems of conventions used in tasks that lots and lots of people deal with all the time, and why some of those systems are just objectively better than others, according to practical and rational criteria.

De gustibus is still non disputandum, of course! Nobody is saying you can’t personally prefer any system you like to any other system. But functional superiority is a thing, and we’re here to identify it (although not necessarily unanimously in every case, of course).

Ready? Some examples:

  • Metric units >> English units. Decimal, simple, consistent, case closed.
  • Euro coin denominations >> dollar coin denominations. Actually I’m accepting argument on this one, though I still support this ranking.
  • Phonetically ordered alphabets >> randomly ordered alphabets. If you ever studied an Indic language and subsequently found yourself absentmindedly flipping to the front of an English dictionary to look up an English word beginning with “u”, you know what I’m talking about. Phonemes have inherent structure, people! It is kind of ridiculous to just toss them (or their corresponding graphemes) into a random sequence like the Latin alphabet (or any of its Phoenician-derived cousins) when you could organize them to reflect that structure!
  • American musical note terminology >> UK musical note terminology. Whole, half, quarter, eighth, sixteenth, and so on. Binary, simple, consistent, case closed. Get outta here with that archaic hodgepodge of minims and crotchets and hemidemisemiquavers, for Pete’s sake.
  • American or UK crochet terms >> continental crochet terms. Okay, the US and UK crocheters have an off-by-one discrepancy based on where you start counting at “single crochet”, but at least it then increases by rational sequence through half-double, double, half-treble/triple, triple, etc., as opposed to random names like bride and punto alto and what-not.

QWERTY vs. Dvorak? I actually have no idea, but feel free to make a case either way.

snake_case >> CamelCase for identifiers.

Depends on the language. Python favours snake_case, Kotlin and Java camelCase, with TitleCase for class names.

I just follow the convention or documentation, I have been bitten too often by overly strict linters (SonarQube, for example)

Funny you should mention crocheting right next to a paragraph into which “crotchet” (a quarter note) would have perfectly fit.

No, I’m saying the Python way is superior. That some languages have settled on the wrong convention just makes them wrong.

Ungendered languages ==> gendered languages

DD-MM-YYYY>>MM-DD-YYYY Day Month Year going up in a logical order, small to big, vesus some meaningless hodgpodge.

W-D-L>>W-L-D for same reason

Even better is YYYY-MM-DD since it goes in the same order as numbers (and therefore sorts correctly). As far as coinage, the use of 25c coin instead of a 20c coin is only a minor annoyance. Although it would become more important if we wanted to get rid of the 5c coin. Canada has already rid itself of the cent coin.

Yeah, the argument for a 20 cent coin is that it maintains the same pattern as the bills. Everything is 1, 2, or 5 times a power of 10. But while that pattern might be nice, I’m not sure it makes any real practical difference.

Of course, US coinage also lacks a coin between 1 and 5 cents, and for all practical purposes also lacks a 50 cent coin. But then, there’s an argument that too many different coins also has drawbacks.

ISO-standard date >> British style >> American style

2023-11-12

Biggest unit of time down to smallest. Need to sort some filenames by date? Nope, you don’t; they are already sorted <chef’s kiss>

I still prefer using letters for the months, like 2023 Nov 13 or 13 Nov 2023. It doesn’t auto-sort, but it’s completely unambiguous: An ISO date can be misread as an American-style date, but a date with letters for the month can’t be misread (well, except maybe for early-first-century dates, but those don’t come up much).

To add one, closing punctuation in a quote: The proper system is to put things inside the quotation marks if and only if they’re part of the quote. This is far superior to the American system of just jamming all of the punctuation inside of the quote just because.

How come?

NATO phonetic alphabet >> everyone’s personal “A as in Albert, B as in Boy” system.

Heh, yeah; in fact, I did mention the “crotchet” in that paragraph! Possibly that’s what subconsciously inspired my choice of the next example.

snake_case is more readable. It is especially more readable when you have abbreviations and the like. For a made-up example, HTTPSSOAPWSDLParser is no way as readable as https_soap_wsdl_parser.

If we are going all nerdy…

GraphQL >> REST >> SOAP

kilometers per-liter is better than liters per 100 kilometers.

American spelling is better than British spelling (except for aluminum)

I feel there ought to be a better system for pitches too. Given that equal temperament has been standard in western music for quite a while now, the staff notation with certain notes being ‘first class’ while others are indicated by a rather clumsy and arbitrary mechanism of sharps and flats seems messy. I guess it was devised as a kludge to extend a simpler earlier system and just became entrenched?

Let’s just agree that all conventional musical notation is pretty out of date.
There’s a reason that modern teaching and visualization materials if they don’t have to use conventional notation all largely choose something akin to a piano roll / Dodeka system.

You can take my hemidemisemiquaver grace notes out of my cold dead piping hands!