silnrsi / font-shimenkan

Fonts for the Miao script
Other
3 stars 1 forks source link

[lpo] 16F54 needs raising when following 16F61 #54

Open cheuk879 opened 5 years ago

cheuk879 commented 5 years ago

plrd-lpo-n_-raise Top cell shows current positioning in Taogu. User wants positioning in bottom cell.

cheuk879 commented 5 years ago

Turns out user is still using old build. Verified behaviour matches bottom cell in build #194.

jvgaultney commented 4 years ago

Reopening due to changes from #70

jvgaultney commented 4 years ago

This is now going to require special contextual positioning. So far you've reported these combinations, in which u16F54 and u16F66 need to be centered no matter at which position they are at:

u16F61+u16F54 u16F61+u16F66

Are those the only two cases? Why is u16F61 special? Why not 6A? 62, 63, 64? What about circular ones (such as 59)?

Is it because of the possible visual confusion between u16F61+u16F54 and u16F62 when at F level? If so, then should u16F6A+u16F54 also be changes so as not to be confused with u16F6B?

We can add some special cases for [lpo], but the way we do that may be different if u16F61+u16F54 is truly the only case vs. if we need a general centering mechanism that may need to be turned on for other combinations and positions.

(BTW - I'd advise we don't try to fix this for v1.0, and so I've removed it from the milestone.)

cheuk879 commented 4 years ago

[lpo] is the only one expressing a preference for centring u16F54 after u16F61. They did not actually mention the same for u16F66. But since Issue #70 deals with both when not in sequence & the fix there is what they have confirmed, we'll need clarification for the sequence u16F61+u16F66. And since other groups have so far not indicated anything in this regard, we're seeking input from them as well. Stay tuned.

cheuk879 commented 4 years ago

Now this is what we've got from users recently:

[ywqa] still hasn't replied. Will post update when it arrives.

jvgaultney commented 4 years ago

Thanks - this is definitely post 1.0 though.