If you want to delete duplicates, you don't have to select them. The most efficient way I know is to just add unique key ignoring any errors (if you want to keep the first record from any set of duplicates):
create table tbl_new like tbl;
alter table tbl_new add unique index a_b_uniq (a,b);
insert ignore into tbl_new select * from tbl order by id asc;
rename table tbl to tbl_old, tbl_new to tbl;
drop table tbl_old;
-- optionally, drop unique index
alter table tbl drop index a_b_uniq;
Because I made the mistake of allowing as many as they want, and want only to keep the first one.
Tho you do have a point; if they don't realise the others have gone, why does it matter.
Love PHP. Love CSS. Love learning new tricks too.
All the best from the United Kingdom.