{"thread":{"id":"27111","subject":"spurious fail of git merge after rebasing","startedAt":"2011-04-15T22:45:12Z","lastAt":"2011-04-18T07:37:33Z","messageCount":3,"participants":["D. A. van Delft","Jeff King","Danny van Delft"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"165937","messageId":"201104160045.12633.davd@bart.nl","threadId":"27111","inReplyTo":null,"subject":"spurious fail of git merge after rebasing","fromName":"D. A. van Delft","fromEmail":"davd@bart.nl","sentAt":"2011-04-15T22:45:12Z","receivedAt":"2011-04-15T22:45:12Z","isPatch":false,"sender":{"key":"davd@bart.nl","avatar":null},"body":"Hi,\n\nnot sure if this is the right place, but I couldn't find a \nbugzilla or something for git.\n\nPoint is, I have found a scenario where sometimes git fails on \na specific merge.\nEach time I start with the same identical git repository \n(archived once and restored each time before a run) and run \nthe same fixed series of rebases and merges on it. Once in a \nwhile it fails with a merge conflict, sometimes at first try, \nsometimes after 10+ times. I have a canned set of testcases \nwhich reproduces this behaviour. I have also narrowed it down \nsomewhat on how to work around, or prevent it from occurring. \n\nHowever, given that it doesn't always fail or succeed, is \nunexpected.\n\nregards,\nDanny A. van Delft\n"},{"id":"166025","messageId":"20110418061027.GA26221@sigill.intra.peff.net","threadId":"27111","inReplyTo":"201104160045.12633.davd@bart.nl","subject":"Re: spurious fail of git merge after rebasing","fromName":"Jeff King","fromEmail":"peff@peff.net","sentAt":"2011-04-18T06:10:27Z","receivedAt":"2011-04-18T06:10:27Z","isPatch":false,"sender":{"key":"peff@peff.net","avatar":"https://avatars.githubusercontent.com/u/45925?v=4"},"body":"On Sat, Apr 16, 2011 at 12:45:12AM +0200, D. A. van Delft wrote:\n\n> Point is, I have found a scenario where sometimes git fails on \n> a specific merge.\n> Each time I start with the same identical git repository \n> (archived once and restored each time before a run) and run \n> the same fixed series of rebases and merges on it. Once in a \n> while it fails with a merge conflict, sometimes at first try, \n> sometimes after 10+ times. I have a canned set of testcases \n> which reproduces this behaviour. I have also narrowed it down \n> somewhat on how to work around, or prevent it from occurring. \n> \n> However, given that it doesn't always fail or succeed, is \n> unexpected.\n\nYeah, git should generally be deterministic. Can you show us your test\ncase?\n\n-Peff\n"},{"id":"166028","messageId":"201104180937.33461.davd@bart.nl","threadId":"27111","inReplyTo":"201104160045.12633.davd@bart.nl","subject":"Re: spurious fail of git merge after rebasing","fromName":"Danny van Delft","fromEmail":"davd@bart.nl","sentAt":"2011-04-18T07:37:33Z","receivedAt":"2011-04-18T07:37:33Z","isPatch":false,"sender":{"key":"davd@bart.nl","avatar":null},"body":"Added attachment contains testcases, in three separate dirs.\n\nI came about this problem during development of a script that \ngenerates git commands to rebase a graph, with all its wanted \nsub graphs, starting at one commit onto another commit. The \n\"runtest\" script is the output of this script, hence the \nstrange looking branchnames during its operation.\n\nregards,\n-- \nDanny\nOp zaterdag 16 april 2011 om 00:45 schreef D. A. van Delft:\n> Hi,\n> \n> not sure if this is the right place, but I couldn't find a\n> bugzilla or something for git.\n> \n> Point is, I have found a scenario where sometimes git fails\n> on a specific merge.\n> Each time I start with the same identical git repository\n> (archived once and restored each time before a run) and run\n> the same fixed series of rebases and merges on it. Once in\n> a while it fails with a merge conflict, sometimes at first\n> try, sometimes after 10+ times. I have a canned set of\n> testcases which reproduces this behaviour. I have also\n> narrowed it down somewhat on how to work around, or\n> prevent it from occurring.\n> \n> However, given that it doesn't always fail or succeed, is\n> unexpected.\n> \n> regards,\n> Danny A. van Delft\n"}]}