Topic: utf8mb4_unicode_ci And mysterious reduction of a limit of characters
I transfer the data from an old site on the updated. Suddenly ceased to be located text in text in mysql.
Now I use utf8mb4_unicode_ci, therefore it was not especially strained, the ratio should be 3 to 4, admitted it.
But started to consider characters in the text that does not get. Them it appeared ~36k (from 65k). It seemed to strange.
Started to experiment: the digit and a Latin letter - get on 65535, "" 32767 times get. Twice less!
In varchar all three gets on 255. Equally.
Database did not change old, and created from zero, the coding of the mysql-server, basis and tables utf8mb4_unicode_ci
[quote =] SHOW VARIABLES WHERE Variable_name LIKE ' character \_ set \_ % ' OR Variable_name LIKE ' collation %'