Mastodawn

A thought and analysis about the OpenType's fsType fields relevance.

[...]

I still think that sofware implementations should be tolerant and have proper flag settings enable licensed usage outside a legacy bitwise mess.

https://dnl.reibenschuh.info/Summary_of_the_OS2_fsType_field_across_a_corpus_of_shareware_fonts.pdf

#typst #pandoc #pdf #truetype #opentype #ttf #otf #fonts

Show thread

Omikhleia - Dragon Brumeux 3d ago

@terefang74
"In open-source typography, the field is often unset even when licenses are explicit elsewhere"
Interesting. This would perhaps be actionable?

Show thread

Terefang74 1d ago

@omikhleia probably not, because the license/copyright text as implemented in opentype is freetext, so usually nothing that would allow easy parsing upon.

also it becomes technically tricky if permissable licenses are provided outside the opentype specification, like a README/LICENSE file paired with fonts in a zip archive.

Show thread

Omikhleia - Dragon Brumeux 1d ago

@terefang74 The license is indeed the authoritative legal document...
You are right that fsType field acting both as a hint of intent and enforcement mechanism is problematic -- but this does seem defective by design, as most technically-based enforcement systems... It's also very messy if the foundry disappears etc.

Show thread

Terefang74

@omikhleia

observe that the specification was first created before the internet.

while you had different foundries, there were basically only three suppliers (MS, Apple, and Adobe).

a historical oversight that can still cause grief and opinions on both sides 35 years after.