search

notofonts / symbols

Noto Symbols
SIL Open Font License 1.1
14 stars 4 forks source link

Height of box characters in "Symbols for Legacy Computing" doesn't match height of box characters in "Block Elements" block #26

Open nico opened 3 years ago

nico commented 3 years ago

I was playing with the "Symbols for Legacy Computing" block a bit, and I noticed that https://www.unicode.org/charts/PDF/U1FB00.pdf mentions that "1FB88 🮈 RIGHT THREE EIGHTHS BLOCK" refers to "→ 258D ▍ left three eighths block" as the corresponding box on the other side.

As far as I can tell, these two glyphs have different heights in Noto. Is that intentional? It sounds like the "Block Elements" glyphs in U+258x should have the same height as the glyphs in "Symbols for Legacy Computing"

screenshot

(Same for the other box drawing characters in "Symbols for Legacy Computing" -- affects most glyphs in that block.)

nico commented 3 years ago

(The heights of the "Box Drawing" glyphs do match the heights of the "Block Elements" glyphs, so making "Symbols for Legacy Computing" use the same height would make all of this consistent.)

speters33w commented 3 years ago

This is actually important, though it is more important in a monospace font.

They are a throwback to codepage 437 (CP437), sometimes called "DOS Codepage," or "PC Codepage." This predates Unicode. It is still in use in some places. PuTTy accesses these unicode characters for CP437 emulation where required. These characters should all be the same height and width, and width should match U+0020 space. Lines should match when stacked on top of each other. Strict postscript glyph names are things like: "SF650000."

This UPS dispay screen uses CP437.

images

simoncozens commented 1 year ago

As far as I can tell, these two glyphs have different heights in Noto.

This is true, but these two blocks appear in two different Noto fonts, with different vertical metrics, so I am not convinced they would work together well even if properly aligned. I suspect the right thing to do here is to add all the Block Elements to symbols as well, in a way that's compatible with the Symbols For Legacy Computing.

simoncozens commented 1 year ago

(This is now the same as #70)

Mercury13 commented 1 year ago

I’ll repeat my solution. Support 2580…9F, with the same character cell size. That’s probably enough.

simoncozens commented 1 year ago

Yes, that's what I just said here:

I suspect the right thing to do here is to add all the Block Elements to symbols as well, in a way that's compatible with the Symbols For Legacy Computing.