UTC/1999-018 Subject: Re: agenda item - Raised decimal Yes. This affects the bidi property, and is thus normative. Should we chose to have MIDDLE DOT be the recommended dot for old-style numbers, then it needs to be CS. (That won't change it in non-numeric contexts). 002E;FULL STOP;Po;0;CS;;;;;N;PERIOD;;;; 00B7;MIDDLE DOT;Po;0;ON;;;;;N;;;;; 2027;HYPHENATION POINT;Po;0;ON;;;;;N;;;;; Mark Peter Westlake wrote: > > At 15:56 1999-05-30 -0700, Markus Kuhn wrote: > >John Cowan wrote on 1999-05-30 16:46 UTC: > >> Markus Kuhn scripsit: > >> > >> > Is there a Unicode character to represent a decimal point? > >> > >> No. That was debated in the past and rejected. There is a fairly > >> full writeup at http://anubis.dkuug.dk/jtc1/sc2/wg2/docs/n2002.pdf . > > > >It seems my question has been misunderstood, because this (and similar > >answers) slightly missed the "point". > > > >I was talking about a character DECIMAL POINT and not a (more general) > >character DECIMAL SEPARATOR. A character DECIMAL POINT would *always* be > >represented by a point (hence the name), and it would be left to the > >font designer whether this point would be located on the base line (US > >convention) or centered (UK convention). A character DECIMAL POINT would > >never be represented by a comma and would not have the purpose of > >unifying the different locale conventions. > > Markus, > > When I asked about this a couple of years ago, someone suggested > that MIDDLE DOT was the closest thing to a decimal point. He didn't > know whether that was a valid usage, though. > > Peter.