Possible regression in ref advertisement

From: Carlos Martín Nieto <cmn@elego.de>
Date: Feb 25, 2013, 16:58 UTC
Message-ID: <1361811516.3212.14.camel@centaur.cmartin.tk>

Hi all,

When testing to see if a different implementation was in shape, I came across something odd where newer git doesn't advertise one of the refs in the git repo.

Running `git ls-remote .` or `git-upload-pack` in my git repo, newer git versions omit peeling the v1.8.0-rc3 tag.

The diff between the command above when ran with 1.7.10.4 (from Debian) and current 'master'

--- old 2013-02-25 17:31:29.583526606 +0100 +++ new 2013-02-25 17:31:36.783526559 +0100

@@ -1379,7 +1379,6 @@
 c15295d7477ccec489953299bd03a8e62f86e611       refs/tags/v1.8.0-rc2
 cd46259ebf2e624bcee2aaae05c36663d414e1a2       refs/tags/v1.8.0-rc2^{}
 22ed067acc84eac8a0a72d20478a18aee4e25571       refs/tags/v1.8.0-rc3
-87a5461fa7b30f7b7baf27204f10219d61500fbf       refs/tags/v1.8.0-rc3^{}
 bfeb8b9ae0012cb61e026cbcd29664876abf5389       refs/tags/v1.8.0.1
 ed9fe755130891fc878bb2433204faffb534697b       refs/tags/v1.8.0.1^{}
 63add1fb45e1ab7a76bb38bbb9467c91fdfaaa7e       refs/tags/v1.8.0.2

Diffing with the output from next, diff tells me it's binary for some
reason, but looking manually, the peeled v1.8.0-rc3 tag isn't there
either. I haven't had time to bisect this, so I'm putting it out there
in case anybody wants to investigate before I have time to dig into it.


   cmn