Anne van Kesteren

Unicode ☹

Unicode is over twenty years old now and support is still lacking. (Though maybe to be fair we should start counting in 1996 when the 16-bit code point limit was abandoned in favor of surrogate pain and a little more freedom so now many years later U+1F4A9 PILE OF POO fits in Unicode too.) Astral Pain by Jacques Distler highlights some problems with MySQL and code points over U+FFFF. Reportedly a hack has been available since mid-2010 (using utf8mb4 rather than utf8 as encoding declaration for utf-8) when MySQL 5.5 was released, but DreamHost is still on MySQL 5.1.

Twenty (or fifteen) years is a long time to sort out basic character storage.