On 7/08/09 18:07, Markus wrote: >> On Thu, 2009-08-06 at 22:53 -0700, Luke Kanies wrote: >>> It's a bit unclear from this patch, but I don't think we're downcasing >>> on the hashing side, either, so unless everyone declares the hostnames >>> as lower-case, this still won't work. > > On Fri, 2009-08-07 at 09:20 +0200, Brice Figureau wrote: >> Are you sure? > > I'm not sure either way (Luke may be though). On the hypotheses that > this was a regression I found where in the chain the case-behavior had > changed. I didn't find anywhere that the hash key case would have > changed, so (by my hypothesis) that wasn't the problem.
To be sure of what introduced the behavior change and fix it, I suggest you git bisect the case, provided you have an integration test (you can even run an automated git bisect). > That leaves several possibilities: > > 1. Hash key casing did change, and I just didn't find it. > 2. The underlying hash implementation changed (e.g. older versions > had used some flavour of indifferent hash) > 3. The regression hypothesis is wrong, and there is a long lived > but heretofore unnoticed bug > 4. The proposed fix is sufficient > > The third alternative is more plausible than it might seem, since by my > reading it doesn't require all lower case names, just _consistent_ > capitalization, if neither the keys added and the keys used as probes > are downcased. That's right. -- Brice Figureau My Blog: http://www.masterzen.fr/ --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Puppet Developers" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/puppet-dev?hl=en -~----------~----~----~----~------~----~------~--~---
