Re: [NANOG] Re: Reasons why BIND isn't being upgraded
On Thu, 01 February 2001, Paul Vixie wrote:
Is there any particular harm from accepting this syntax. It wouldn't be the first time a RFC has been updated to match working code.
No and Yes. No in that an argument could be made that the old parsing routine fell under the "be liberal in what you accept" rules. Yes in that the Master File Format is intended to provide an interchangable database table, so while BIND may have been liberal it was doing so at the expense of some interoperability measures. The real culprit in this story is the script-generator. It should have been cranking out standard-compliant zone files from day one. -- Eric A. Hall http://www.ehsco.com/ Internet Core Protocols http://www.oreilly.com/catalog/coreprot/
On Thu, 1 Feb 2001, Eric A. Hall wrote:
RFC 1122 s1.2.2 may not apply directly to configuration files, but the spirit is good. The bracket acceptance would be classed as a fault-tolerance feature, if BIND had a marketing department and glossy brochures :) I can understand the annoyance felt by a large hosting provider updating BIND in an emergency and finding more than just a security fix. Pim is, I guess, concerned that similar updates in future may have longer MTTR impact. Pete Elke's point about preproduction testing could perhaps be turned from a combative tone to the constructive without loss of information. joshua
On Thu, 1 Feb 2001, Eric A. Hall wrote:
The real culprit in this story is the script-generator. It should have been cranking out standard-compliant zone files from day one.
Absolutely. The guilty party for that script has long left the organization. This is real world, where people have to deal with other people's messes :). Pi
participants (4)
-
Eric A. Hall
-
Joshua Goodall
-
Pim van Riezen
-
Sean Donelan