To:
dnsop@cafax.se
From:
James Raftery <james-dnsop@domainregistry.ie>
Date:
Thu, 8 Feb 2001 12:32:31 +0000
Content-Disposition:
inline
In-Reply-To:
<20010207145127.C15994@nic-se.se>; from mansaxel@nic-se.se on Wed, Feb 07, 2001 at 02:51:27PM +0100
Sender:
owner-dnsop@cafax.se
User-Agent:
Mutt/1.2.5i
Subject:
Re: Bogus nic.fr behavior
On Wed, Feb 07, 2001 at 02:51:27PM +0100, Måns Nilsson wrote: > Do we operate the net as if it was still only operated > by skilled people who know what a RFC is and even has read one, or do > we accept that we might have to use a little less elegant solution in > order to increase yield? Most of us wearing the Protocol hat would go for > perfectionism. But some who do actual work might opt for function.. But when we run checks for user or software stupidity that look outside of what is /required/ and into the area of what we consider to be /desired/ we tread a very thin line. Highlighting common mistakes is good. Assuming that a common mistake for BIND is a common mistake for all implementations is not good. In this case, the AFNIC check script considers it prudent for a BIND server to have an accurate view of the root servers. Fine. That's their choice. But they are blindly applying that check to every request. They shouldn't punish those who use software which isn't BIND, especially when that alternative software is acting correctly. Regards, james -- James Raftery (JBR54) "It's somewhere in the Red Hat district" -- A network engineer's freudian slip when talking about Amsterdam's nightlife at RIPE 38.