>>I find that if I do a basic parent-child-grandchild join like the example with indexes on A.id, B.id, B.id_A, C.id_B, I get Rushmore optimization of None on all tables, but it says it joins on the indexes given.
>>Also, according to the docs, if I index on UPPER(char_field) the following should be fully optimizable, but I only get partial optimization.
SELECT * FROM test WHERE UPPER(char_field) = "SOMETHING"
>>Does this make sense to you?
>
>Is SET DELETED ON and no index on deleted()? This with cause partial optimization, but may or may not affect performance depending on the percentage of deleted records.
Duh! (It's been a long week already...)
SET DELETED OFF got me to it on the character query.
But I still get no optimization on the parent-child-grandchild JOIN.
Insanity: Doing the same thing over and over and expecting different results.
