Using table aliases prevents such problems by disambiguating the column
names. For example, if you had table aliases in just one part of this
query as in the below, you would have gotten an error instead of
deleting all those rows:
select respondent_id from respondent where respondent_id in (s
I've got a SELECT with a WHERE clause using IN against a BYTEA column
which works when I'm querying against the table directly. However when I
substitute an equivalent view in my query (a view made of several inner
joins which simply carry over that same BYTEA column), I get the error:
operato
