- From: Willy Tarreau <w@1wt.eu>
- Date: Thu, 3 Nov 2016 08:01:28 +0100
- To: "Roy T. Fielding" <fielding@gbiv.com>
- Cc: Cory Benfield <cory@lukasa.co.uk>, Julian Reschke <julian.reschke@gmx.de>, Kazuho Oku <kazuhooku@gmail.com>, HTTP Working Group <ietf-http-wg@w3.org>
On Wed, Nov 02, 2016 at 03:50:02PM -0700, Roy T. Fielding wrote: > In my experience, most developers (especially open source developers like me) > are happy to fix bugs in their software, particularly when they are backed > by specification text that is now 21 years old (and still counting). Agreed. When I first noticed haproxy was not properly handling 1xx, I fixed it and redeployed. The only thing (especially with intermediaries) is that these components are rarely upgraded, so the ones facing bugs are replaced and the ones without bugs stay in place for 5+ years. But the good thing is that when something new breaks your old deployment, you quickly notice it, realize you're running an outdated, unsupported version, upgrade to the next one and notice it fixes your problem for free. So yes, adding new 1xx status codes will definitely break existing bogus code, and by the time their users notice, fixes will be available and the issue will only last for the time it takes to update the package. Willy
Received on Thursday, 3 November 2016 07:02:08 UTC