On Thu, 9 May 2019 at 14:10, Andrew Stitcher <astitc...@apache.org> wrote: > > On Wed, 2019-05-08 at 21:25 -0400, Andrew Stitcher wrote: > > ... > > > Whats the nature of the issue/fix? Is it only these examples that > > > would > > > be affected? > > > > I'll know when I find the fix! It's not entirely clear yet - but it > > looks like the existing code might only accidentally work. > > Phew - the existing code is correct, but the reconnect work has changed > the implicit protocol between proton and the select implementation > (usually internal to the proton binding, but wth an external loop like > tornado the external loop). To fix this and maintain the capability of > external loops isn't trivial - at least a days work. > > This change supports an important change (and IMO fix) to the python > bindings of correctly trying *all* the addresses/IP protocols > associated with a given DNS name that I think is more important than > the external loop abiility. > > This being so - I'd recommend releasing this as is with a release note > and being sure to fix this issue for 0.29. > > Andrew >
A vote might be more compelling than a recommendation :) As things stand, no one has yet voted against releasing, and there are currently 3 binding +1 votes, so the vote would pass tomorrow as-is. I'm thus far inclined to let it and proceed, since a fix is not yet in place and it will likely be awkward for me to actually do a respin next week, so doing that will likely significantly delay the release yet further. I'll happily do a 0.28.1 or 0.29.0 in similar time frame once a fix for this or other issues/improvements becomes available. Robbie --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@qpid.apache.org For additional commands, e-mail: users-h...@qpid.apache.org
