I see no problem with adding features to point releases, so long as they are unlikely to cause security/stability issues and don't change existing functionality.
Is there a protocol somewhere about this? -Damien On Aug 3, 2010, at 11:46 AM, J Chris Anderson wrote: > > On Jul 31, 2010, at 6:16 PM, J Chris Anderson wrote: > >> >> On Jul 31, 2010, at 6:09 PM, Noah Slater wrote: >> >>> I saw a conversation on IRC tonight about bumping this thread. I noticed an >>> email from J. Chris in another thread saying he wanted to hold back from >>> 1.0.1 until something was fixed. Waiting for the all clear. Let me know. >> >> >> All clear! I don't remember suggesting we should wait, but I'll take your >> word for it. >> > > Did I say all clear? Whoops! ;) > > It's come to my attention that this commit would like to be backported to > 1.0.1, despite not meeting the procedural requirements for backporting. (it's > a new feature) > > http://svn.apache.org/viewvc?revision=980985&view=revision > > "Add support for replication through an HTTP/HTTPS proxy." > > The reason is that our releases are a leaky abstraction, and Ubuntu will be > freezing to a CouchDB release for their next release in a few days. > > Since Ubuntu is our largest install base, and they would love to be able to > offer sync to users behind proxies, I am +1 on bending the rules for them. > > The patch itself is not technically risky, as it has no effect unless the > user provides the new replicator option, so the chance of introducing bugs is > very small. > > I am backporting this now, but of course I am open to discussion. The > preemptive backport is meant to ensure that we don't forget to discuss this. > If anyone is -1 on the idea, please let us know, so that we can find common > ground. > > Chris > > >> I definitely reverted and redid one of the patches that I backported to >> 1.0.x, but that's all cleaned up now. >> >> I think we are in the clear. Plus I'm excited to start working on 1.0.2. >> >> We should also roll 0.11.3 at this time. >> >> Chris >
