diff options
| author | Jeremy Kerr <jk@ozlabs.org> | 2008-09-11 16:42:26 +1000 |
|---|---|---|
| committer | Jeremy Kerr <jk@ozlabs.org> | 2008-09-11 16:42:26 +1000 |
| commit | a72679a9622db66e828e86377f29c9c0c6574d69 (patch) | |
| tree | cae300bb42b1806b09f28ae8251c5775778df760 /lib/sql | |
| parent | 0deabd4014cbc9419d203356786e966c4f803ea3 (diff) | |
| download | patchwork-a72679a9622db66e828e86377f29c9c0c6574d69.tar.bz2 patchwork-a72679a9622db66e828e86377f29c9c0c6574d69.tar.xz | |
Don't just rely on random for UserPersonConfirmation keys
It looks like we're getting identical keys generated for confirmation
keys. Problem has been reported to django, but in the meantime, salt
with the user and email details, then sha1 to give the final key.
This requires an increase in the field size for key, migration script
included.
Signed-off-by: Jeremy Kerr <jk@ozlabs.org>
Diffstat (limited to 'lib/sql')
| -rw-r--r-- | lib/sql/migration/002-extend-userpersonconfirmation-key-length.sql | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/lib/sql/migration/002-extend-userpersonconfirmation-key-length.sql b/lib/sql/migration/002-extend-userpersonconfirmation-key-length.sql new file mode 100644 index 0000000..fa10fba --- /dev/null +++ b/lib/sql/migration/002-extend-userpersonconfirmation-key-length.sql @@ -0,0 +1,4 @@ +BEGIN; +ALTER TABLE patchwork_userpersonconfirmation + ALTER COLUMN key TYPE char(40); +COMMIT; |