New: add more name IDs and make sure the correct settings are fontRev…

[yisibl]

1. The sfntNames now support more fields:

• designer
• designerurl
• license
• licenseurl

2. fontRevision It is no longer always 1.0.
   Based on the recommendations[1] in this document, we can now support better version number settings. It can be displayed correctly in the font editing software.

e.g. Version 5.678 beta3

TODO: add test cases.

[1] https://silnrsi.github.io/FDBP/en-US/Versioning.html

[yisibl]

@puzrin please take a look?

[puzrin]

I'd say, project tests should be reorganized with use of opentype.js instead of binary compare. Because everything flies to nightmare :).

Without normal testing it's very difficult to move forward.

[yisibl]

In name id 5, the minor version should always be three digits and should refer to the UFO specification（openTypeNameVersion）: https://unifiedfontobject.org/versions/ufo3/fontinfo.plist/#opentype-name-table-fields，I can also consider adding openTypeNameUniqueID

[yisibl]

In the implementation of ufo2ft, the default version is always 0.000. Use Python's .zfill(3) to fill 0 in front of the minor version, similar to .padStart(3, '0') in JS.

https://github.com/googlefonts/ufo2ft/blob/742b3a54408dbba99f24a9d21e32b0c499ca6ffe/Lib/ufo2ft/fontInfoData.py#L169-L175

[puzrin]

IMO this is not enough flexible and should be reworked.

1. Split for 2 parts: pure string add and version magic.
2. Api for string add clutter options significantly. IMO this can be added via more abstract interface, list of [ data, id, platformID, encodingID, languageID ]. That will allow to add any vendor string user can imagine.
3. I don't like existing magic with version. It can exists, to extract 2 major numbers, when possible, but --version-num also required to define header value explicit.

[puzrin]

@yisibl here are 2 PRs and 1 issue i'd like to close/replace: #86, #102, #105.

Reason: topics were mutated multiple time, and tend to be "endless".

My last opinion #86 (comment) is to create 2 new issues if that's still actual. AFAIK, all other critical things have been landed to master. Let me know what you think.

[yisibl]

I'm very sorry for the late reply. When I have time, I will investigate how the APIs of other libraries are handled. You can close these issues first. Thanks!

[puzrin]

Closed in favor of #111 & #112