"David G. Johnston" <david.g.johnston@xxxxxxxxx> writes: > This seems to fail to answer the OPs question. Specifically, do these > rules automatically, or at least if #1 is not true, cause typemod > information to be lost? IOW, is it because of the unknown that both end up > up-casted to typemod-less text? Any mismatch of typmod will result in the merged column being considered to have no typmod. There is no provision for identifying a "common superset" typmod. It would have to be some type-specific API, since the encoding of typmod is type-specific; and we don't have one. regards, tom lane -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
