don't need a capslock key if you have separate caps keys.

@th I had one of those when I was little, flea market find, it went to bits when I explored how it was put together and worked .. babillions of cogs.

@th There must be a reason for this layout. The arrows are logical, so must be the rest?

@artelse @th I would guess it is just 1 to 9 left to right, shifted into two rows to safe space and the 0 put in the middle either for some mechanical reasons or because it is a very common digit (similar to the space bar)

@daherb @artelse @th

I see it as

1234 5,0 6789

I imagine accountants complained that putting the five and zero at the far right or left was inconvenient so this is how the machinists responded.

It's the ultimate centrist solution!

@daherb @th @DenOfEarth @artelse I heard from a bloke with a dog on a string that someone got a copy of the New York Times and counted how often each number came up and put the most common ones in the middle.

@Life_is @th why? On calculator you also often see special buttons for 00 and 000.

@wmd @th

A number in the decimal System could be: 1432567. In an abstract visualization: xyyyyyy. Every y can be a digit 0-9. And every digit has the same probability. That is different for the first digit x. It is never 0 (apart from Software development and few other special cases). But for the other digits 1-9 the probability is not the same: in ~50% of cases it is a 1, second most is 2 and so on, with 9 least likely. It is a Feature of number Systems. It can also be used for fraud detection: if a document contains falsified numbers the likelyhood of 1 as the first digit is significantly less than 50%.

@Life_is @th I don't think I see the argument here beyond that the first digit can't be 0. The lowers the chance for 0 assuming no other biases (which sounds unlikely), but doesn't say much about it's relation to other numbers.

@Life_is @th thanks, that does make me wonder more why calculators have buttons for 00 and 000. :)

@wmd @Life_is @th lemme hazard a guess: because (these) calculators were primarily targeted at shopkeepers, who would make use of the 00 every time they needed to type a whole number of their currency (say, $1.00), similarly, the 000 for thousands.

Or in other words: while 1 may be the most common *single digit*, 00 and 000 are the most commonly typed *strings of digits*, especially in currency settings.

@aleks @Life_is @th Oooh, yeah. This sounds likely and a better formulation of the bias I was expecting. I wonder if it's true.

@wmd @Life_is @th Let's verify that this could be the case on written words: I'm using the OANC here. First, the most common *letter* of English is 'e', no surprise there. But the most common *digraph*? It's 'th'. 'ee' is nowhere in sight. If I wanted to make a keyboard for English that makes digraphs easier to type, 'th', 'he', 'in' and 'er' would be my choice.

@wmd @Life_is @th We can, of course, try to do the same for numbers. But I don't have a corpus of (plausibly distributed) numbers available. We can use the OANC again, and then the most common numeric digraphs seem to be 19, 20, 10, and only then 00.

However, this is written language and NOT the kind of numbers you would type into a calculator. I'd need to look for some online price database or something like that.

@wmd @Life_is @th On why written language carries with it certain biases, see degruyter.com/document/doi/10.

tl;dr: culturally salient numbers, such as the numbers of recent years are more frequent in written language use. Makes sense, we talk a lot about recent years, such as 19xx and 20xx. But you wouldn't type those into a calculator very often.

This is painfully obvious when you use trigraphs. The data are obviously rubbish for informing your average calculator layout.

@th @wmd @Life_is for random data, benford’s law holds, but you often have data that is rounded or artificial

@LambdaDuck @th @wmd

But you do not enter rounded or artificial data manually (at least you do not, if you are smart and use a Script for rounding and generation).

@Life_is @LambdaDuck @th this discussion arose from old (mechinical) calculator hardware though.

@th I think it would somehow make me worse at math.

@th@v.st
There's probably some mechanical reason for it to be like that because there's no way a sane person could ever come up with something this wonky
:neocat_0_0:

@th @NanoRaptor Is that a partially Dvorak-ed numpad? 0 in the centre as it’s the most commonly used?

@dymaxion @th A bit of googling tells me it's a very early typewriter, from the experimental days. Going by this video (turn the sound down and just read the subtitles) it looks like a distant ancestor of the golfball typewriter. youtube.com/watch?v=ozFLYTrU8_

@Daveosaurus @dymaxion @th oh that's cool! I was kind of expecting it to have a golfball honestly, not a faceted cylinder, but the faceted cylinder makes sense too.

@Daveosaurus @dymaxion @th oh that looks so cool

love the simplicity of the mechanism

the handedness seems odd (assuming everything back then was right-hand only) but maybe it feels different to use than it looks

@th i saw a video of this working recently, it was so cool. like, i love it.

@th i saw a video of this working recently, it was so cool. like, i love it.

@th ooooh these pointing typewriters are so cool :O

@FlorianTischner
Upper case letters for printing presses were literally kept in an upper case, so why not on a typewriter? 😂

@th

@th Wow, I’ve never seen a typewriter like that! I wonder how old it is?

@th aaaahhhh!!! at first I was expecting Linotype keyboard layout when I saw this and my brain threw a wooden shoe into itself when I didn't see the expected etaoin shrdlu pattern

@th yes! I find it interesting the letter frequencies aren't that far different. (At least, I'm guessing that's what the layout is based on.)

@th @vxo
Actually it's "ETAOIN SHRDLU" for English machines. There's also a small documentary titled "Farewell Etaoin Shrdlu" which follows the last Linotype set print of The New York Times.

See here: archive.org/details/FarewellEt

@bayindirh @th yea that's the one I'm used to. I had never seen the French layout before

@th a fun aside: apparently back in the day, keyboards typically didn't have zero or one keys, with the reason being you could just use the I and O keys which looked similar enough (ex. I92O), here's an early QWERTY layout from 1878, also check out the wikipedia article about QWERTY, the history section is really interesting https://en.m.wikipedia.org/wiki/QWERTY

@mjdxp @th 0 and 1 is missing, because it's analog, not binary/digital 😉

@mjdxp @th woah this thing looks like a frickin altar 😯.

bring your sacrifice to the god of the written word!

@th i am so confused by the lack of 1 and 0. i mean what. why. how

Sign in to participate in the conversation
(void *) social site

(void*)