Firacode: Is ◻ instead of [] a good idea?

[vi]

It makes empty list/vector look like some entirely new object instead of just a list without any elements.

[MadcapJake]

I personally like it. The unicode square you have in the title is quite a bit smaller than the bracket-square that Fira Code provides.

[sjmielke]

It _really_ threw me off at first as well. Is there an easy way to remove it from the font file?

[xcambar]

Agreed, I find it very eye-catching and it doesn't quite recall an empty vector to me.

[vyp]

+1 on having an option to remove it if it's possible.

[weavejester]

Maybe there should be a subtle break between the two [], similar to the way -- is rendered.

[xcambar]

@weavejester :+1:

[surrealist]

The square for [ ] confuses me too.
I'd like to know what are scenarios or usages that square make good senses of [].

[bootmii]

Java uses it as a dimension symbol.

[tp]

Was also confused by this when using array type shorthand:

Maybe it would improve when my theme would use italics for types, but using the default behavior I find the square box way to prominent.

[banshee]

The box for [] always makes me think it's just a missing unicode character box.

[tobia]

@surrealist in some languages, such as Haskell, [] is a distinct object with its own identity.

That being said, I agree with @weavejester that a subtle break between the two [] would be useful.

[joelalejandro]

+1 , agreed with @tobia

[Juerd]

Suggestion: make the subtle break diagonal.

[joelalejandro]

Diagonal break number two looks beautiful.

El lun., 18 de jul. de 2016 a la(s) 14:32, Juerd Waalboer <
[email protected]> escribió:

Suggestion: make the subtle break diagonal.
[image: suggestion 1]
https://camo.githubusercontent.com/3de9831a46a68dfee72f104bd04b42041c1099e2/68747470733a2f2f6a756572642e6e6c2f692f30363461653032353366323330343761313131336137346439656538633139642e706e67
[image: suggestion 2]
https://camo.githubusercontent.com/210679335cd9b2331521f73b0b4a9e98654b56ce/68747470733a2f2f6a756572642e6e6c2f692f32346637393464326631646635363666353137636365313864303863313034632e706e67

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
https://github.com/tonsky/FiraCode/issues/92#issuecomment-233399598, or mute
the thread
https://github.com/notifications/unsubscribe-auth/AAHQga0nnLEhFkJr977svhOLoA80pXnYks5qW7i6gaJpZM4G5O9b
.

_Joel Alejandro Villarreal Bertoldi_
Desarrollador de software / Diseñador gráfico

GitHub / Twitter: @joelalejandro
Tel.: [phone]

[bobmagicii]

i support the diagonal broken version. [] is quite frequent in PHP and i thought it was the "unicode done got knackered" character and not legit code.

[xcambar]

Oooh, how subtle! 👍 for the second variation.

[easyaspi314]

+1 with the second one too. Goes great with Fira's subtly tapered ends.

[nesteruk]

Diagonal break will likely not be particularly visible on smaller font sizes.

[xgenvn]

@Juerd +1

[thany]

@Juerd :+1:

[polymeris]

I agree that [] looks bad as a box. To me it neither expresses "empty list" nor "unspecified index" like in int x[];.
It is also very distinctive, and pops up everywhere, e.g. in every (defn foo [] ...).

I would just remove it from the ligatures. Or maybe render it as [·] if you insist on having one.

[bobmagicii]

i like both the / cut [] and the [·] ideas.

[surrealist]

[·] looks like it has a thing in it, not empty.

[mpaluszkiewicz]

@Juerd +1

[MikkyX]

@Juerd +1, that second one definitely looks like the best way to do this

[banshee]

Personally, I think the diagonal idea isn't good; it implies there's some asymmetry involved, and that's not true for many, many cases (all cases?).

I think the dot idea is also suboptimal. The problem is that [] is used to mean many different things in many different languages. Adding something in the middle just isn't going to work across many different contexts. You see that in several comments; people are talking about "empty" - but there is nothing about [] that implies empty. Or that doesn't imply empty. Emptiness just isn't a concept involving []. That's my problem with the box, too - the box is very, very different than two matching square braces. It's making invalid semantic assumptions.

[bobmagicii]

@banshee but look at the <= and >= operators. part of it is visual flare. that is what makes the [/] work.

[Spontifixus]

I personally think the [ ] ligature should go away. The reasoning behind this is the following.
The sign for an empty array, list or vector is the two square brackets, not one square. This always interrupts me reading the code asking me what I did wrong there.

So +1 for this request.

[banshee]

Turns out it's super easy to remove []'s ligature, so if you love the font and just want [] gone, it's straightforward. I did it and submitted a pull request, or you just just build it yourself from my fork (https://github.com/banshee/FiraCode/tree/remove_leftbracket_rightbracket_ligature)

[clairernovotny]

@banshee can you build them (ttf at least) and put them somewhere for us to download? Many of us don't have Glyphs or a Mac to build it on.

[banshee]

You don't actually need a mac, you can build them just by installing Clojure and then figuring out which Clojure plugins to install then if you're on Windows installing the new Ubuntu support in Windows 10 so you get bash and then...

Hmm. OK, here are the font files :-) https://github.com/banshee/FiraCode/releases/tag/noLigs

I want to be clear - I think FiraCode is fantastic, and I think it's totally fair to call what I did a "nasty hack job." I'd love for my little branch to die off quickly. I'd even like a nicely done ligature that didn't look like a box, but my skills don't include making fonts look good.

[clairernovotny]

@banshee any chance of ttf versions? OTF can't be used in text tool windows like PowerShell/Command prompt :/

[tonsky]

Gosh you’re boring. I’ve just released 1.201 version where [] is rendered as circle instead of square. That should solve it

[thedevsaddam]

@tonsky previous comment was fun of the year[2016] :smile: