I can the logic. I was still surprised. =O I guess this reflects the slight logical incongruity of a "case insensitve, case preserving" collation. Could an UPDATE affecting only case on a COLLATE nocase column be validly ignored as an optimisation?
I can the logic. I was still surprised. =O I guess this reflects the slight logical incongruity of a "case insensitve, case preserving" collation. Could an UPDATE affecting only case on a COLLATE nocase column be validly ignored as an optimisation?