Re: CDDB 1.11 test

From: Ilya Zakharevich (nospam-abuse_at_ilyaz.org)
Date: 02/12/04 

Date: Wed, 11 Feb 2004 23:41:17 +0000 (UTC)

[A complimentary Cc of this posting was sent to
Joerg Hevers
<news2.2004.joergh@spamgourmet.org>], who wrote in article <MPG.1a89122c37b3ee079899bc@ID-23289.user.uni-berlin.de>:
> > Thanks, this is what I suspected... Was it introduced recently, or
> > was the test of CDDB *always* failing? It dates Aug 2003...
>
> This was introduced with server software version 1.5, which was
> installed on the freedb-servers in late December or January, depending
> on the server.

A related question: is it possible to introduce two fake queries: one
which is guarantied to return only one match, and one which is
guarantied to return (e.g.) exactly 3 matches? And guarantie the
contents of the results - so one can test CDDB retrievers on these
entries?

One can use queries which would never appear in real life (e.g., with
total length less than the last track). Of course, one must also make
sure that the entries are never updated via submit interface...

Thanks,
Ilya

P.S. Are there plans to fix this broken DISCID issue? E.g., allow
queries of the form DISCID-MD5, where MD5 is some ;-) less-conflicting
signature?