> I'm not sure either what's the character between INFINITY (∞) and ELEMENT OF (∈). I checked an old MSX Technical Data Book and it seems to be lowercase phi (φ), but it certainly doesn't look like it.
The last three letters in the Danish and Norwegian alphabets are Æ, Ø, and Å. Given that both the lower and upper case version of 'æ' and 'å' are included in the MSX font, it would make sense that they would also include the 'ø' symbol which the resulting font is currently missing. Given the limited number of pixels it would also make sense that the same graphic is used for both the lower and upper case version of this symbol.
In a pinch, maybe (see bottom paragraph for a better alternative), but ö is Swedish and German. It's "the same letter" but at the same time it is not.
In Sweden, the modern alphabet ends with "å, ä, ö". In Norway and Denmark the modern alphabet ends with "æ, ø, å".
Although the order is different, å is the same in both Swedish and Norwegian and Danish, and ä is "equal to" æ, and ö is "equal to" ø.
I've seen some people aged 60+ use ö in their handwriting in Norway.
I was born in Sweden and am a Swedish citizen but have been living in Norway since I was four years old. (Swedish father, Norwegian mother.) My last name contains and ö in it, but to the Norwegian authorities and everyone else here, my name is written with an ø. I always write my name with an ø, because that's what I'm used to and that what everybody else is used to, except when I'm in Sweden, then I use ö.
Aside from my last name, I would never interchange ä with æ or ö with ø, because doing so either for a Norwegian word using Swedish letters or for a Swedish word using Norwegian letters would simply look absurd to me. In fact I think even some people might find it offensive to do so. The "å, ä, ö" and the "æ, ø, å" are part of the Swedish and Norwegian peoples identities.
If you need to type words that have these letters in them but you don't have the letters accessible on the computer you are using, it's much better to perform Anglicization IMO. Type å as aa, ä/æ as ae and ö/ø as oe".
It looks like the MSX font is very closely based on the original IBM PC font (these days known as CP437), so for the characters that are ambiguous in 8x8 I recommend checking out the 8x14 versions from the PC.
> I wasn't sure whether some letters are upper or lower case, like pi (π), or theta (θ)
Wikipedia says "227 (E3hex) is the Greek lowercase pi (U+03C0, π), but early fonts such as Terminal use a variant of pi that is ambiguous in case, and therefore can be used for the Greek capital pi (U+03A0, Π) or the n-ary product sign (U+220F, ∏)."
> I'm not sure either what's the character between INFINITY (∞) and ELEMENT OF (∈). I checked an old MSX Technical Data Book and it seems to be lowercase phi (φ), but it certainly doesn't look like it.
Wikipedia says: "237 (EDhex) is supposed to be used as Greek lowercase phi, but is mainly used as the empty set sign (U+2205, ∅) and was also used as the Greek phi symbol in italics (U+03D5, ϕ) to name angles, diameter sign (U+2300, ⌀), and as a surrogate for the Latin lowercase O with stroke (U+00F8, ø)."
The joys of low-res, approximate typefaces - glyphs so weird and obscure you can use them for just about anything.
They're really not - the IBM PC set is double-pixel wide and with quite different choices for many of the letters because the MSX is 5x7 while IBM PC used 8x14 (EGA) or 8x16 (VGA).
It is far more likely they took inspiration from the Apple ][ font and tried to add serifs in the limited space.
CP437 is the codepage - chars outside the ASCII range of 0-127. The font you are thinking of is probably the BIOS font from either the IBM EGA or VGA adapter.
By "based on" I meant the character set, not the actual glyph designs, I absolutely agree the MSX characters don't look much like the PC ones.
...but the character sets match surprisingly well; outside the basic ASCII range they have mostly the same accented and Greek characters in the same places, many of the same glyphs for control codes, and some of the symbols right at the end. The box-drawing characters are very different, and the MSX has more accented characters in total, but nothing that would raise eyebrows in a European MS-DOS code-page.
(off-topic: thank you very much for Envy Code R, it's one of my favourite monospaced typefaces ever!)
I haven't used an MSX before but the font really reminds me of fixed6x13, which is the one used in the screenshot in the above link. I've modified mine with a slashed zero, which makes the resemblance even closer.
related: UNSCII[1] is a font set attempting to combine characteristics for all the 8-bit microcomputers. this person has found a whole pile of various box drawing characters and oddities that don't map to anything in unicode. I should think there would be an argument here for a Unicode Consortium submission, to make it possible to represent "texts" from this machines in unicode proper.
If I understood correctly, the characters corresponding to the first row of CP437 are actually 2-byte sequences (01 xx) on MSX, and the prefixed values as individual bytes are shifted down a row to correspond to the uppercase alphabet rather than being values 0-31, so iconv would have to know about the prefix to not mix those up. The resemblance to CP437 probably isn't a coincidence, though. Microsoft was heavily involved in MSX, so it was undoubtedly indirectly influenced by IBM PC.
The last three letters in the Danish and Norwegian alphabets are Æ, Ø, and Å. Given that both the lower and upper case version of 'æ' and 'å' are included in the MSX font, it would make sense that they would also include the 'ø' symbol which the resulting font is currently missing. Given the limited number of pixels it would also make sense that the same graphic is used for both the lower and upper case version of this symbol.
https://en.wikipedia.org/wiki/Danish_and_Norwegian_alphabet
Size does matter: https://www.youtube.com/watch?v=f488uJAQgmw