@simoncozens has set up a CI-driven, automated PDF proof generation process for a nastaliq project:
- Uses shaper and language-specific wordlist to determine set of words which can be correctly shaped.
- Allows both font designer and Google design side to evaluate progress of fonts under development
- Currently deployed on a private Github repo
As part of the long term plan to run all of Noto from this github org (https://github.com/notofonts), I'd love to see a generic proof generator project made out of the nastaliq one, which will require us to gather together language-specific wordlists for all languages that Noto supports, and then deployed to all font repos in the notofonts org repos.
There is a distinction between "proofs" and "specimens", the former being used by font designers to evaluate extremely specific things during font development, and the latter being used by prospective users to evaluate if/when they want to use the font family - but there is some overlap in these things, as a specimen with extreme depth will contain a lot of late-stage proofs.
What makes Simon's approach here important is that it fills a gap for complex scripts between something like https://www.adhesiontext.com and a specimen generator like https://variable.typespecimens.xyz
It also uses UX mocks to show the fonts in context of use, and since Noto is used for UIs, especially Material Design on Android, we will also need to gather together mocks for all languages.
@simoncozens has set up a CI-driven, automated PDF proof generation process for a nastaliq project:
As part of the long term plan to run all of Noto from this github org (https://github.com/notofonts), I'd love to see a generic proof generator project made out of the nastaliq one, which will require us to gather together language-specific wordlists for all languages that Noto supports, and then deployed to all font repos in the notofonts org repos.
There is a distinction between "proofs" and "specimens", the former being used by font designers to evaluate extremely specific things during font development, and the latter being used by prospective users to evaluate if/when they want to use the font family - but there is some overlap in these things, as a specimen with extreme depth will contain a lot of late-stage proofs.
What makes Simon's approach here important is that it fills a gap for complex scripts between something like https://www.adhesiontext.com and a specimen generator like https://variable.typespecimens.xyz
It also uses UX mocks to show the fonts in context of use, and since Noto is used for UIs, especially Material Design on Android, we will also need to gather together mocks for all languages.