Non-Vietnamese stacked diacritics overlap

[Shihali]

Arbitrary vertically stacked diacritics overlap with illegible results. Top is Andika New Basic, which is designed to handle stacking diacritics exceptionally well. Bottom is Public Sans.

Common Vietnamese stacking diacritics stack legibly, although not in the preferred manner. Top is Andika (with Vietnamese diacritics activated via language), bottom is Public Sans. Andika is not perfect, but closer.

[davelab6]

@crystaltype please could you take a look

[moyogo]

Note that Vietnamese is not the only language or transcription system that uses ắ, ề, ẫ, etc. In some of those, the vertical stacking is preferred, or acceptable like in Vietnamese, so it would be nice to still have them available.

[thisisdano]

Apologies if this is a silly question, but are the arbitrary stacked macron diacritics actual characters? Either way, I'll learn how to do this! But curious to know if this is language support or more of a best practice.

And also: I can improve the Vietnamese to be closer to the expected output. Can you provide examples of what you expect for these characters — or, even better, a range of expected output?

[Shihali]

The arbitrary stacked macron diacritics are not used in the standard orthography of any language, as far as I am aware. They are used in historical linguistics and transcription. For example, ā̆ is used to indicate an "a" of unknown length or an "a" that is either short or long, and ā́ is very common in transcriptions of Vedic Sanskrit to indicate a long, high-pitched "a".

https://vietnamesetypography.com/design-challenges/ has more on Vietnamese typography than I can possibly provide. http://software.sil.org/wp-content/uploads/sites/19/2015/12/AndikaNewBasic-features5.5.pdf isn't perfect, but shows some differences between "standard" arbitrary combining diacritics and preferred Vietnamese usage.

[moyogo]

Beside historical linguistics and transcription systems there are a few orthographies that use macron with another diacritics. While ū́ is used in Lithuanian and is in Unicode as a single character other macron letters are used like ā́ used in Kaska language or ḗ used in Kaska and Southern Tutchone which can only be represented by using combining characters in Unicode.

There are several orthographies and transcriptions systems that use other stacked diacritics. Depending on the scope of the fonts, you may want to allow for some or any stacked diacritics.

[thisisdano]

Useful Vietnamese typography resource for future reference: https://vietnamesetypography.com/

[thisisdano]

I've not yet improved the macron stacking, but I am trying to improve the display and spacing of the Vietnamese characters here: https://federalist-proxy.app.cloud.gov/preview/uswds/public-sans/dw-diacritics/test/

[crystaltype]

Apologize for the late reply! Hi @thisisdano, I'm a native Vietnamese speaker, hopefully I can help you improve Vietnamese in your typeface ^^ Anyway, I notice some of the things that you might want to improve:

• All the diacritics should be at the same weight as the stroke weight (for example, I think your tilde is a little too thin in the lighter weights).
• You can scale the diacritics but please don't squeeze them (keep the ratio intact), or else it won't look natural.
• In the stacked diacritics in bolder weights, you can scale both of the diacritics (not keep the first one un-scaled, and then cut out the second one like in circumflex+acute). In small sizes it would look like the acute is missing, and/or affect readability. Does that even make sense XD

These are just some thoughts after first looking at your typeface! ^^ I can help you fix them though, if you want, I can make a fork and work on the improving Vietnamese there.

[thisisdano]

Thanks @crystaltype! I've tried to fix the weights a bit and improve the scale and position of stacked acute/grave. (Trying to get closer to what I see in Adapter, here.) Does this seem like an improvement?