From owner-freebsd-hackers Thu Jun 19 11:46:54 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA10808 for hackers-outgoing; Thu, 19 Jun 1997 11:46:54 -0700 (PDT) Received: from pandora.hh.kew.com (ahd@kendra.ne.highway1.com [24.128.53.73]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA10803 for ; Thu, 19 Jun 1997 11:46:52 -0700 (PDT) Received: (from ahd@localhost) by pandora.hh.kew.com (8.8.5/8.8.5) id OAA03931; Thu, 19 Jun 1997 14:46:48 -0400 (EDT) Date: Thu, 19 Jun 1997 14:46:48 -0400 (EDT) From: Drew Derbyshire Message-Id: <199706191846.OAA03931@pandora.hh.kew.com> To: craig@gnofn.org, hackers@FreeBSD.ORG Subject: Re: sad UT_HOSTSIZE of 16 Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > From: Craig Johnston > If we're even going to store the remote hostname in utmp, in this day and > age of ridiculously long hostnames, shouldn't we go for a bit more than 16 > chars? . . . > 16 for display purposes might be just fine, but it'd be nice to > have 255 chars or so of remote hostname stored in the utmp file. > Sure, we can find out more other ways, but I think it should be > in utmp. I've been considering this. Many of the programs automatically write the dotted quad address if the name is too long -- I have some fixes to apply to make this to more standard if as threatened I get commit privs. :-) If the programs automatically put dotted quads IP addresses in when needed for length and you can assume a fast DNS, perhaps the better apporach is to add support to the various viewing programs (last, finger, etc.) to resolve the dotted quad back to the full host name. This doesn't change the data format or waste space, at the expense of additional wall clock when formatting. > Of course the question is: what breaks? Nothing, if you fix normally the data and enhance the data consumers as I describe. Unenhanced consumers still work as well. Will a printable IPv6 address fit in 16 bytes? -ahd-