On 06/11/2009, at 8:48, Raimon Fernandez wrote:
I'm blocked ...
On 06/11/2009, at 6:27, John DeSoi wrote:
On Nov 5, 2009, at 12:35 PM, Raimon Fernandez wrote:
at least, my first md5 (psw+user) is the same as the pg_shadow
(wihtout the 'md5') ...
should I md5 the first md5 as I get
I'm blocked ...
On 06/11/2009, at 6:27, John DeSoi wrote:
On Nov 5, 2009, at 12:35 PM, Raimon Fernandez wrote:
at least, my first md5 (psw+user) is the same as the pg_shadow
(wihtout the 'md5') ...
should I md5 the first md5 as I get it as string (like username) or
byte by byte ?
John DeSoi writes:
> ... But it is unclear to me what happens when
> the user or database name has non-ascii characters. The client
> encoding is not established until after authentication.
No encoding conversion will happen on those names. If you consistently
use the same encoding in all cl
On Nov 5, 2009, at 12:35 PM, Raimon Fernandez wrote:
at least, my first md5 (psw+user) is the same as the pg_shadow
(wihtout the 'md5') ...
should I md5 the first md5 as I get it as string (like username) or
byte by byte ?
As far as I know, a string. But it is unclear to me what happens
