Semicolons; revisited

[josh-hadley]

pre-16.0 beta public feedback pointed out some issues with the first pass at addressing UTC-179-A113. After some discussion with other PAG members, we decided to make the definition of "Semicolons" to be U+003B and its compatibility equivalents.

[Ken-Whistler]

Looks o.k.

[markusicu]

Hi @josh-hadley are you ready to merge this? I would like to get a snapshot of the UCD proper for UCDXML and the beta.

[josh-hadley]

Hi @josh-hadley are you ready to merge this? I would like to get a snapshot of the UCD proper for UCDXML and the beta.

Almost ready. I wanted to give the author of the feedback that initiated this change an opportunity to look it over first. If I don't hear anything soon I'll go ahead and merge and get you un-blocked.

[dscorbett]

My original feedback was specifically about the non-common-script semicolons. This PR also removes two common-script semicolons: U+204F REVERSED SEMICOLON and U+2E35 TURNED SEMICOLON. I don’t know how those should be categorized, but I wanted to check with you that their removal was intentional. FWIW, L2/12-168 shows U+204F REVERSED SEMICOLON used as a standard semicolon in right-to-left Old Hungarian. That’s all the feedback I have on this PR.

[markusicu]

Thanks, David.
We "inherited" a less-than-clear task and may have gone too far.
I like what Josh and Robin have done here: Dial it back to a principled, defensible subset of characters.
I think it would be useful to merge this PR as is and leave it at that for the start of the 16.0 beta.
Also retaining two more sc=Common characters could be ok too if @josh-hadley feels like making those changes.
During beta (and beyond) we can discuss the various "semicolons" further.

[josh-hadley]

it would be useful to merge this PR as is and leave it at that for the start of the 16.0 beta

Agreed, and I'll proceed with that.

Thanks for responding @dscorbett!