I took the " out of the password and it works. The other install worked with the " in the password. On 04/09/2015 02:40 AM, Steffen Kaiser wrote:> -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > On Thu, 9 Apr 2015, James wrote: > >> Both use PAM authentication on Linux. >> openssl s_client -connect localhost:993 >> a login test_user test_secure_password >> The response I got back was: >> a BAD Invalid characters in atom > > Does your password contain the " quote? If not, enclose the password > in "". Or try the literal form: > > 1 login user {##} > password > > where ## is the number of octets of password > >> The secure password used to work but it was a different architecture >> (x86_64 I think). >> The current architecture is armhf. > > architecture of the server or client? > Maybe you need to re-create the password store? > > - -- Steffen Kaiser > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1 > > iQEVAwUBVSYeynz1H7kL/d9rAQIsqwgAgVzzXjZj1hAO22hoamDM0mPJovqnNlIF > YQ9/pR421GOca8YOebh9YHqqMbxuj8DFADS06EAdcKQdyONKd0CsxtHsjqIrV/FB > uxEMY/MFdTto9ToMjj33YtQFz0MjACvrqO/hzepgCK3Pgt3HK0UHyN36qYmwnpNs > Vzbm/YYb+4xnciYq2sLlAKvrvIn+LQT25OP7toGiXnOkGQSTbwxRtysmLe+f3Cvv > XlnYTZsKTIuwAJNcDphKpd0LSts5k6Fua4pw1LtxP42W63xE5iq3KTqGdi4u6nQS > jVRdKwAKVjihbbLFN1K1nlVyEEQP2dpEk4B+smRwGQTaxRrDFlki8A=> =L7CD > -----END PGP SIGNATURE----- >
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thu, 9 Apr 2015, James wrote:> I took the " out of the password and it works. > The other install worked with the " in the password.pass"word is no atom and Dovecot should reject it. Maybe the former version did not checked the atoms 100% RFC conform. You can always use a litral to pass these words.> > On 04/09/2015 02:40 AM, Steffen Kaiser wrote: >> On Thu, 9 Apr 2015, James wrote: >> >>> Both use PAM authentication on Linux. >>> openssl s_client -connect localhost:993 >>> a login test_user test_secure_password >>> The response I got back was: >>> a BAD Invalid characters in atom >> >> Does your password contain the " quote? If not, enclose the password in "". >> Or try the literal form: >> >> 1 login user {##} >> password >> >> where ## is the number of octets of password >> >>> The secure password used to work but it was a different architecture >>> (x86_64 I think). >>> The current architecture is armhf. >> >> architecture of the server or client? >> Maybe you need to re-create the password store?- -- Steffen Kaiser -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEVAwUBVSdqQXz1H7kL/d9rAQI//ggAyAADFM6ct/cBiE6uebHIfFFWtP4yi3GW ghG+6w58yNguekrBYl7D5K8tGLVHwCMsLagJ0OWU30H8LiEY8w4uV33B8lit3XFD RL6RPCPW8imEgKx2cPI4JLPFimRuTZN/tsHB6ej4QsAJSJ6KiyNx6o9XYYfR6oQ6 JARROQ7v266wBPx1U6HnVfbZPpM32I5fUE4FKd88JkfBODYrZNlSULqWEh9XVnWX 1ZdUmb3gi+5zCZpXMZT49EMpLGwArHrLtJyf24KPMUATtQoZ1n8GyJJ3zajSZL0F eceQ4c/5gza4dQJVwpHozkR2fSrwFR8iCgCcD43qQBQvX/fc9g64fg==yYqF -----END PGP SIGNATURE-----
On Apr 10, 2015, at 12:14 AM, Steffen Kaiser <skdovecot at smail.inf.fh-brs.de> wrote:> pass"word > > is no atom and Dovecot should reject it.No, not passwords should be rejected.> Maybe the former version did not checked the atoms 100% RFC conform.Where is the definition that passwords have to be atomic? And if it is there, it should be changed. Nothing anywhere at anytime should care about the contents of a password once it passes basic ?not password of monkey? complexity checks. -- "You see, in this world there's two kinds of people, my friend: Those with loaded guns and those who dig. You dig."