On Mon, Apr 8, 2013 at 3:36 PM, Robbie Gemmell <[email protected]> wrote: > On 8 April 2013 18:17, Justin Ross <[email protected]> wrote: > >> These are all improvements, even the first one. No doubt it's an >> important security enhancement, but it's not a regression, > > > Some might argue there is a regression or sorts as you couldnt previously > do things like change the SSL, authentication, ACL, etc settings that you > can now.
Yes, I agree, it is a regression in that sense. >> and it >> seems reasonable to release note that you should enable acls if you >> wish to use the web console. I'd be more inclined to accept it if the >> change were small, but it isn't. >> > For what its worth, the changes really are quite 'small'. Much of the diff > is simply moving code that would have otherwise had to be needlesly > duplicated into a utility class and so isn't actually all that great of a > change. Okay. If you think the risk to stability and the schedule is minimal, then I defer to you. > It's hard for me to reconcile these changes with the goals for a >> second release candidate. I hope you can tell that I ordinarily try >> to be very flexible, but this doesn't make sense. >> >> Honest question: if you have a compelling reason to introduce these >> changes into a stable stream before 0.24, should we have a distinct >> branch for that purpose? > > I'm a little unclear on what you mean here, is it a '0.22.1' release or > simply putting the changes on a branch before introducing them to 0.22? Sorry, I meant the former, 0.22.1. I mention it because it's a way to deliver something out of sync with the current cadence, if that's what you need. Justin --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
