Re: Setting min/max TLS protocol in clientside libpq - Mailing list pgsql-hackers

From Michael Paquier
Subject Re: Setting min/max TLS protocol in clientside libpq
Date
Msg-id 20200425002948.GA1474@paquier.xyz
Whole thread Raw
In response to Re: Setting min/max TLS protocol in clientside libpq  (Daniel Gustafsson <daniel@yesql.se>)
List pgsql-hackers
On Fri, Apr 24, 2020 at 02:03:04PM +0200, Daniel Gustafsson wrote:
> That was brought up by Michael in the thread, but none of us followed up on it
> it seems.  The current name was chosen to be consistent with the already
> existing ssl* client-side settings, but I don't really have strong opinions on
> if that makes sense or not.  Perhaps use ssl_m{in|max}_protocolversion to make
> it more readable?

There was no hard push in favor of this comment so I did not insist,
but I am not wedded to the existing connection parameter names.

-       {"sslminprotocolversion", "PGSSLMINPROTOCOLVERSION", NULL, NULL,
+       {"ssl_min_protocolversion", "PGSSLMINPROTOCOLVERSION", NULL, NULL,
Shouldn't that actually be "ssl_min_protocol_version" with one extra
underscore?

> The attached renames the userfacing setting, but keeps the environment variable
> without underscores as most settings have env vars without underscores.

There are two in this case: PG_COLOR and PG_COLORS.  For readability
it could make sense to use something like PG_SSL_MIN_PROTOCOL_VERSION
or PGSSL_MIN_PROTOCOL_VERSION, but like Daniel I'd rather keep the env
variables without underscores.
--
Michael

Attachment

pgsql-hackers by date:

Previous
From: "David G. Johnston"
Date:
Subject: Re: psql - pager support - using invisible chars for signalling endof report
Next
From: James Coleman
Date:
Subject: Re: Binary search in ScalarArrayOpExpr for OR'd constant arrays