Collation of INFINITY vs. EMPTY SET?

[Joseph Myers]

On Tue, 17 May 2016, Carlos O'Donell wrote:

> Is the solution to write automation to create iso14651_t1_common
> and list all the unspecified elements in code point order?

We should have automation, which should follow current Unicode data except 
where there are deliberate differences (bug 14095).  I expect both making 
automation to follow the current data, and identifying existing deliberate 
differences, to be substantial amounts of work.

(That's a framework for answering the question of how glibc should behave, 
rather than an answer for how to achieve particular results.)

-- 
Joseph S. Myers
joseph@codesourcery.com