Digit 0 height doesn't match other digits in some applications

[Reedbeta]

Unsure if this is a font issue or a renderer issue, or an intersection of both, but I thought I'd start here.

Example string: [Version 10.0.19045.3693]

Here's how it renders in Windows Terminal, with Monaspace Neon:

The 0 in 19045 is visibly shorter than other digits. But the 0's in 10.0 appear to be the right height.

Looks similar in VSCode:

Looks more correct in Visual Studio 2022:

Looks good in Firefox as well:

[Firestar-Reimu]

Why the 0's in 10.0 is much wider than other digits? Is it a ligature like 0.0 -> eyes?

[alexeyten]

@Firestar-Reimu that's “Texture Healing” https://monaspace.githubnext.com/

[Reedbeta]

Aesthetically, I don't love the wide zeroes - they look incongruous next to the other digits for me - but that's another issue. 🙂 The mismatched heights, though, seem like an actual bug somewhere.

[ToxicFrog]

I just came here to report an issue that I think might be the same thing. This is in VScode with ligatures enabled:

And disabled:

In both cases we get a lot of unusually short zeroes, but if ligatures are enabled the 0 following the leading 1, and only that 0, is the correct height. This is in vscode and you get the same effect with any non-false setting of fontLigatures, including ones that should be turning off texture healing.

[ToxicFrog]

On further reading, the "any non-false setting" is #118 which is a vscode bug, so this is in fact a texture healing bug with monaspace. 'calt' off gives short zeroes everywhere, 'calt' gives a mix of short and normal zeroes.

[Finii]

It would be interesting to know if you use the static or the variable font version?

[Finii]

The different zeros do have different heights, just by one unit, but nevertheless, that rounds either up or down when "hinting".

In this regard it would be also interesting if the behaviour you encounter changes if you change the font size by 1 or 2 pt. Or maybe look at ridiculously big letters like 60 pt, is the difference still one pixel?

Ignore the different position, the XY display on the bottom of the glyhs shows the green point at 1484 and 1485

[idan]

Fixed in ca8015b27bdc6d8e1201b1ad722f4ae3c38970d0