crypt auth
I notice our docs have:
If you are at all concerned about password
<quote>sniffing</> attacks then <literal>md5</> is preferred, with
<literal>crypt</> to be used only if you must support pre-7.2
clients. Plain <literal>password</> should be avoided especially for
At what point do we just remove the support and say that people need to
upgrade their clients? Sure, it's up to ppl not to configure it that
way, but security-wise it's a foot-gun that I think is completely
unnecessary.
//Magnus
Magnus Hagander wrote:
I notice our docs have:
If you are at all concerned about password
<quote>sniffing</> attacks then <literal>md5</> is preferred, with
<literal>crypt</> to be used only if you must support pre-7.2
clients. Plain <literal>password</> should be avoided especially forAt what point do we just remove the support and say that people need to
upgrade their clients? Sure, it's up to ppl not to configure it that
way, but security-wise it's a foot-gun that I think is completely
unnecessary.
AFAICT, removing an authentication method requires a protocol version
bump. If you think it is worth dealing with those complications, then
go for it. I think it might be worth it.
Peter Eisentraut <peter_e@gmx.net> writes:
AFAICT, removing an authentication method requires a protocol version
bump.
Why would it require that? There would just be some auth method codes
that remain reserved but aren't used anymore.
regards, tom lane
Tom Lane wrote:
Peter Eisentraut <peter_e@gmx.net> writes:
AFAICT, removing an authentication method requires a protocol version
bump.Why would it require that? There would just be some auth method codes
that remain reserved but aren't used anymore.
Yeah, I was mistaken. AuthenticationCryptPassword would remain in the
protocol definition, but the server would just never send it.
Peter Eisentraut wrote:
Tom Lane wrote:
Peter Eisentraut <peter_e@gmx.net> writes:
AFAICT, removing an authentication method requires a protocol version
bump.Why would it require that? There would just be some auth method codes
that remain reserved but aren't used anymore.Yeah, I was mistaken. AuthenticationCryptPassword would remain in the
protocol definition, but the server would just never send it.
Since I've seen no actual objections to this happening, I will go ahead
and do it unless someone objects explicitly :-)
//Magnus
Magnus Hagander wrote:
I notice our docs have:
If you are at all concerned about password
<quote>sniffing</> attacks then <literal>md5</> is preferred, with
<literal>crypt</> to be used only if you must support pre-7.2
clients. Plain <literal>password</> should be avoided especially forAt what point do we just remove the support and say that people need to
upgrade their clients? Sure, it's up to ppl not to configure it that
way, but security-wise it's a foot-gun that I think is completely
unnecessary.
Here's a patch that does this. Will apply unless there are objections.
//Magnus