Latin: Stylistic Set / Alternates for I and J for titling/all caps usage

[davelab6]

Currently NotoSans-*.ttf (latest version, all systems/apps) lack OpenType stylistic sets.

These are important parts of the highest quality fonts, providing alternates for various use-cases which are important but not the default design decisions, eg, increased legibility, and graphic design.

For legibility, Apple recently updated the San Francisco system fonts with a "legibility set" -
https://twitter.com/Tricertops/status/912623537529204736 - but the design of Noto (primarily made for text and UI) already has default design decisions with this in mind. Here the classic Il1 (cap eye, lowercase el, numeral 1) are distinguishable from each other:

However, conversely I would like Noto to work well in larger size graphic design use-cases, where some alternate design decisions are more suitable.

For example the cap eye I has serifs, and the cap jay J has a tail that descends below the baseline; when but using Noto in larger "title" size usage (and sometimes but not always with an all caps treatment) these are less desirable.

Red/left is the current Noto design, Black/right is the proposed design

Those 2 changes are the ones I'm aware of right now. I propose they be part of both salt and a ss01 features; the latter with a named value of Titling Alternates

(I had considered proposing case feature also, but think this is going too far :)

[davelab6]

I realise that there is a mistake in my above proposal, in that I missed the need for the alternates to also be available in the aalt "Access All Alternates" GSUB feature (https://www.microsoft.com/typography/otspec/features_ae.htm#aalt) so that they can be accessed in the Glyph Panel in desktop design apps like Adobe's

So the lookups for these glyphs should be added to aalt in addition to ss01 and salt

[moyogo]

@davelab6 How does the descender of the cap J relate to the serifs of the cap I? I'm not sure it makes sense to group either cap J with legibility more then the other. If it had to be in two sets, it should be all caps they can have descenders do or none do.

[davelab6]

They are related by usage, not form

[moyogo]

@davelab6 ah, yes I misread, that makes sense for "titling".

[marekjez86]

should be available in https://github.com/googlei18n/noto-fonts/tree/master/phaseIII_only/unhinted/otf