w3cping / font-anti-fingerprinting

A system for preventing font fingerprinting
Other
14 stars 4 forks source link

Record Mozilla's position from #7. #9

Open jyasskin opened 4 years ago

jyasskin commented 4 years ago

@bholley, is your comment in https://github.com/w3cping/font-anti-fingerprinting/issues/7#issuecomment-598424965 Mozilla's position?

bholley commented 4 years ago

I think so, unless @annevk has other thoughts.

Is there any reason the proposal for web font caching needs to be coupled with the proposal for restricting system fonts? I think the latter is something we're likely supportive of, and thus would prefer to send disjoint signals on those two things.

bholley commented 4 years ago

also CC @TanviHacks and @martinthomson

jyasskin commented 4 years ago

Unfortunately, the CSSWG is concerned about the impact on minority linguistic communities if we just block system fonts without some other way to let those users cache the fonts. I think #6 (let users "install" fonts to the web) doesn't solve your concern either, because it's unfair to expect users to know that they're picking a font they don't want an attacker to know about.

It is totally reasonable for some browsers to pre-cache all the "common" fonts if they're worried about the privacy impact of caching them on demand, in case that would help your concern.

astearns commented 4 years ago

@bholley the CSSWG is definitely concerned about minority linguistic communities, but I don't think there's consensus that caching webfonts is the right fix for that issue. So to my mind separating the web font caching proposal from restricting local font access (with a different fix to avoid breaking minority linguistic content) is reasonable.

martinthomson commented 4 years ago

I would prefer it if this were taken to https://github.com/mozilla/standards-positions first.

I agree with Bobby (and Anne) here and the comments they have made are likely to stick. However, we do have a process that we're trying to use consistently and I don't want to set a precedent that a comment on a GitHub issue is taken for our formal position.

jyasskin commented 4 years ago

'k, thanks. I'm going to update the proposal tomorrow with a possible compromise, and then I'll poke mozilla/standards-positions.