To:
Brad Knowles <brad.knowles@skynet.be>, Ed Sawicki <ed@alcpress.com>, Randy Bush <randy@psg.com>
Cc:
DNS Operations <dnsop@cafax.se>, namedroppers <namedroppers@ops.ietf.org>, <ngtrans@sunroof.eng.sun.com>, IPng <ipng@sunroof.eng.sun.com>
From:
David Conrad <david.conrad@nominum.com>
Date:
Tue, 16 Jul 2002 18:50:51 -0700
In-Reply-To:
<a05111b0eb95a3d13cbc1@[10.9.8.228]>
Sender:
owner-dnsop@cafax.se
User-Agent:
Microsoft-Entourage/10.1.0.2006
Subject:
Re: (ngtrans) Re: RFC 1886 Interop Tests & Results
Brad, On 7/16/02 2:17 PM, "Brad Knowles" <brad.knowles@skynet.be> wrote: > Moreover, if you are someone looking to actually use products > such as the ones under test, Who says the things under test are products? The DS implementation Nominum did was done on a snapshot of the ISC BIND tree that has not yet been released, even in alpha. It shouldn't be too much of a surprise that that version of BINDv9 contains bugs. The implementation was done to prove that the DS spec was actually implementable and to provide input back to the standardization proceess to tighten up the ambiguous bits. > it would be a benefit to know which > products worked and which ones didn't, or which ones had what > problems in what environments, so that you would have a better idea > as to what products might be suitable for use in your own environment. When products are available, such testing would indeed be useful. Of course, this isn't particularly relevant to what the IETF is doing. > IMO, tests like this without full disclosure are meaningless. To be blunt, your opinion would be wrong. Of course, there is meaning -- just the fact that multiple implementations of a protocol exist is useful information (indeed, an IETF protocol cannot advance to full standard without this fact being true). Rgds, -drc