Sarasa Gothic Mono
26 points by dgv
26 points by dgv
The font took a while to load and I was very confused for about 10 seconds lol
The releases in github are huge, the one I downloaded was 745MB. I don't think this is usable, unfortunately.
I am moderately sure that includes a lot of weight and style variants, which you don't necessarily need. I used it for a while just because it applied the styles I liked to Iosevka, but since I'm not a speaker of any of CJK, I didn't make full use of it.
So, a bit more context from a CJK user: I think this series of fonts is mostly used if you have CJK characters in your code/terminals. In popular monospace fonts like Fira Code or Jetbrains Mono, the width of two Latin characters doesn't match the width of a CJK character (displayed with other fonts, as Fira Code, etc. don't cover CJK charsets), so in a non grid-based editor (that is, not TUI) or some not grid-rendered terminals, the rendered texts can look as disastrous as this:
1 2 3 4 5 6 7 8
一二三四
(Most CJK users, TUI applications and your Markdown / Orgmode tables expect "8" to line up with "四".) This is best illustrated by the screenshots from cnfonts, an Emacs package that helps you adjust the relative size of your monospace fonts and CJK fonts so that they line up. Even with cnfonts, I remember being unable to line things up perfectly however I try (because the ratio isn't a good fixed-point number?), not to mention other non-Emacs editors. So here goes Sarasa, a font bundling Iosevka and Source Sans CJK, ensuring their dimensions match up exactly.
(As other comments pointed out, it's huge in size, and probably more intended to be used in your local IDEs than as a font on webpages.)
I like it, hopefully ligatures for |> and <> are added.
I won’t skip the opportunity to share why these are a bad idea & a misuse of what ligatures are actually for.
[…] Ligatures are special characters in a font that combine two (or more) troublesome characters into one. For instance, in serifed text faces, the lowercase f often collides with the lowercase i and l. To fix this, the fi and fl are often combined into a single shape (what pros would call a glyph).
In this type designer’s opinion, a good ligature doesn’t draw attention to itself: it simply resolves whatever collision would’ve happened. Ideally, you don’t even notice it’s there. […]
Ligatures in programming fonts follow a similar idea. But instead of fixing the odd troublesome combination, well-intentioned amateur ligaturists are adding dozens of new & strange ligatures. For instance, these come from Fira Code, a heavily ligatured spinoff of the open-source Fira Mono.
I agree with the author that what we actually want is better Unicode support in languages so we can type what we mean instead of the likes of trying to shoehorn ASCII Hyphen + Greater-than sign into → when we have the technology (there are multiple solutions to inputting & quite a few (tho not enough) languages that support Unicode).
I'd hope they make builds of this font too like there are for Iosevka, where there are no programming "ligatures", but double-width arrows etc. that can be used instead to replace operators if you want in the editor. This works brilliantly for example with Emacs prettify-mode.
I'm using a tweaked Iosevka myself now, though maybe I can get over my preferences and just go with one of the default builds of Sarasa so I don't have to maintain my custom build config.
I use my own Neovim plugin I wrote a couple years ago that uses Vim’s conceal + tree-sitter to replace some characters in my editor with real Unicode for the languages that don’t support these features. Since it’s tree-sitter, I haven’t found any false positions since you can select on proper node types.
Emacs prettify is similar in that you can select based on the same properties set by the language mode to do syntax highlighting &c. These days the mode may be based on treesitter, though most are not still.
I won’t skip the opportunity to share why these are a bad idea
That's just, like, your/their opinion, man.
I find this a very strong argument even though still struggle to enter umlauts more then needed.
Historically many ligatures merged into own symbols like & originating from the latin word et.
I see coding ligatures as a first step to the arguments you posted.
That’s a good take. Language designers should take note & support Unicode since the masses do tend to like better symbols. Hell, even APL didn’t mind inventing symbols when concise/useful. I have seen some uptick—especially Raku, even OCaml finally got accented Latin characters finally supported now (no other script tho).
I the two languages I want, Thai & Lao, aren’t supported. Abugidas largely render poorly in the terminal & the fallbacks aren’t very good :(
Abugidas largely render poorly in the terminal & the fallbacks aren’t very good :(
Out of interest, would you call uuterm's rendering acceptable?
Is this a terminal emulator from 20 years ago? Can’t say I have heard of it :P I do know that in Kitty, I can get a some Thai fonts that are also about 20 years old to render (some emulators don’t even get that far & choose to hide the vowels or tone markers), but I end up with scenarios like a vowel/tone marker is placed above/below a consonant still takes up 2 character widths which puts the real cursor over an extra space; I’m sure how you can see how this doesn’t scale to text of any length.