On Thu, Aug 28, 2014 at 6:11 AM, Jan Slodicka <j...@resco.net> wrote:
> > If NOCASE is a custom collation, could a bug in that collation explain > corruption of indexes FK_account_transactioncurrencyid and > FK_account_ownerid? (Both have NUMERIC affinity and store blobs such as > X'001C2300C5DF8BEA11DF8834FBDCD77E'.) > > So far I supposed that the answer is NOT and consequently excluded a > collation bug. > Correct. If corruptions are appearing in indexes that do not use custom collations, that would tend to rule out a problem with the collation function. -- D. Richard Hipp d...@sqlite.org _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users