Re: Cleaning up the INET/CIDR mess - Mailing list pgsql-hackers

From Greg Stark
Subject Re: Cleaning up the INET/CIDR mess
Date
Msg-id 87slrctbnx.fsf@stark.xeocode.com
Whole thread Raw
In response to Re: Cleaning up the INET/CIDR mess  (Andrew - Supernews <andrew+nonews@supernews.com>)
Responses Re: Cleaning up the INET/CIDR mess  (Greg Stark <gsstark@mit.edu>)
List pgsql-hackers
Andrew - Supernews <andrew+nonews@supernews.com> writes:

> On 2006-01-25, Bruce Momjian <pgman@candle.pha.pa.us> wrote:
> > Agreed.  10.1 as 10.0.0.1 is an old behavior which has been removed from
> > most modern versions of networking tools.

On the contrary not only is it still widely used but it is *required* by POSIX
for the relevant functions, inet_aton and getaddrinfo. Note that getaddrinfo
was created from whole cloth by POSIX so there was no backwards compatibility
need for it.

This isn't an obscure old-fashioned thing. People really do use this syntax.

> Indeed so. However the current behaviour has neither the merit of being
> traditional nor the merit of being logical:

Well for networks (cidr datatype) people do frequently refer to things like
10.1/16 and intend it to mean the network prefix. Sure you could argue having
the netmask default to the old class-based addressing is anachronistic but
what other default netmask would you suggest anyways? The only other
reasonable default would be the longest 0-bit suffix which would produce some
odd surprising results like '10.1/16' but '10.2/17'.

-- 
greg



pgsql-hackers by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: Cleaning up the INET/CIDR mess
Next
From: Thomas Hallgren
Date:
Subject: Re: Adding a --quiet option to initdb