U+276F prompt character (❯) got smaller between 3.1.1 and 3.2.1

[gibfahn]

U+276F prompt character (❯) got smaller between 3.1.1 and 3.2.1

[Finii]

Where is the PR @Finii !!!?

[gibfahn]

Thanks for the detailed explanation and the solution! I don't know if it's helpful, but I filed your proposed solution as a PR. I think that's right, don't have much experience in this area. #1617

It's your change, so if you want to file your own PR for it fine by me, just trying to reduce some cycles for you 😁 .

[Finii]

Most probably you already read it, but if not... the cryptic rules have some explanation

# Stretch 'xz' or 'pa' (preserve aspect ratio)
# Supported params: overlap | careful | xy-ratio | dont_copy | ypadding    
# Overlap value is used horizontally but vertically limited to 0.01
# Careful does not overwrite/modify existing glyphs
# The xy-ratio limits the x-scale for a given y-scale to make the ratio <= this value (to prevent over-wide glyphs)
# '1' means occupu 1 cell (default for 'xy')
# '2' means occupy 2 cells (default for 'pa')
# '!' means do the 'pa' scaling even with non mono fonts (else it just scales down, never up)
# '^' means that scaling shall fill the whole cell and not only the icon-cap-height (for mono fonts, other always use the whole cell)
# Dont_copy does not overwrite existing glyphs but rescales the preexisting ones
#
# Be careful, stretch may not change within a ScaleRule!

'align': 'c', 'valign': 'c', 'stretch': '^pa1!', 'params': {'ypadding': 0.3, 'careful': True}

• Align horizontally and vertically in the center
• Scale it with preserved aspect ratio (x and y scale is equal) into 1 cell (default would be 'up to 2' for pa)
• Take the whole line-hight (^) and subtract 30% from the height (ypadding)
• Do the scaling also for proportional fonts
• Do not change preexisting glyphs

[Finii]

Sorry to hear of your size problem.

I can see it directly in the font 🤔

Suspect

• #1564

Because from these heavy brackets people DO expect excessively big icons 🤔

[Finii]

In this example the icon is (far) less tall than the small letter T, which is strange

I will example in detail.

[Finii]

The cell heights difference is quite huge here?

Not the case 'globally':

68% vs 76% ?

Hmm, thanks for reporting. This needs some consideration and a systematic check. I see the effect by the reason is not quite clear; and maybe that should have been marked with the ^ (fit to full height) modifier.

[Finii]

Ahh, one should read ...

[Finii]

Ah, the heavy brackets are now two times shrunken, there was an old rule that already sized them down relative to the full cell:

The rule needs to be changed to ^pa1! ....

[Finii]

@allcontributors please add @gibfahn for bug

[allcontributors]

@Finii

I've put up a pull request to add @gibfahn! 🎉

[martinpelikan]

Apologies if this is unrelated or intended, but I noticed an increase in the size of some symbols in ttf-firacode-nerd, going from 3.2.0 -> 3.2.1 in my i3 status bar.

3.2.0	3.2.1

U+F120 (nf-fa-terminal) and U+F11B (nf-fa-gamepad)