OK - let me see if I understand what you are saying: $a - is the commit-id of the uploaded artifacts $b - is the commit-id that was in the vote email $c - is the commit-id of the tag in git
Which of those three don't match. (When I read your description - a and b match, but c does not) If that's the case, delete the tag, and recreate the tag so that it matches what was in a and b. If that isn't the exact case, we need to figure out where the failure is, and decide how to handle it. --David On Sat, Nov 1, 2014 at 6:24 PM, Daan Hoogland <[email protected]> wrote: > I am talking about a mismatch between the commit-id that the tag > 4.4..1 points to and the artifact that was uploaded. I think they are > functionally the same but the internal version is different. The > artifact is 4.4.1-SNAPSHOT hence probably the root of the release > branch and the commit is the tip where the version strings were > changed to 4.4.1. > > not sure if this relates to your links. I made a bug in the > build_asf.sh and uploaded soething other then the release commit id > > On Sat, Nov 1, 2014 at 10:22 PM, David Nalley <[email protected]> wrote: >> -private >> >> We are talking about whats in: >> https://dist.apache.org/repos/dist/release/cloudstack/releases/4.4.1/ >> ? >> If so, I am not sure how we arrive at the fact that the release >> artifacts aren't what we voted on. I looked at the commit log for the >> part of the tree and found [1]. I also checked it out, and svn log-ed >> it and get the output in [2]. >> >> So let me step back and ask - how did you determine that what's in >> dist.a.o doesn't match what we voted on? OR -are we talking about >> something else? >> >> --David >> >> [1] http://markmail.org/message/vjoxa6lg7o6dxmsl >> [2] https://paste.apache.org/agtf >> >> >> On Sat, Nov 1, 2014 at 3:12 PM, Daan Hoogland <[email protected]> >> wrote: >>> People, >>> >>> It seems the uploaded artifacts on the release repo of apache does not >>> contain the commit-id of the voted release. I don't know if this can >>> be remedied. I think we should pull it, right? can we replace it by >>> artifacts that do contain the right version? Or should we have a new >>> version? >>> >>> thanks, >>> -- >>> Daan > > > > -- > Daan
