Currently, we overload "indent" meaning both "unix socket
authentication" and "ident over tcp", depending on what type of
connection it is. This is quite unfortunate - one of them being one of
the most secure options we have, the other one being one of the most
*insecure* ones (really? ident over tcp? does *anybody* use that
Should we not consider naming those two different things?
If not now, then at least put it on the TODO of things to do the next
time we need to break backwards compatibility with the format of
pg_hba.conf? Though if we're going to break backwards compatibility
anywhere, pg_hba is probably one of the least bad places to do it...
pgsql-hackers by date
|Next:||From: Tom Lane||Date: 2010-11-17 15:39:04|
|Subject: Re: Indent authentication overloading |
|Previous:||From: Ross J. Reedstrom||Date: 2010-11-17 15:32:53|
|Subject: Re: contrib: auth_delay module|