{"thread":{"id":"6387","subject":"git-svn: does anybody care for sub-optimal merge tracking?","startedAt":"2007-01-15T06:56:35Z","lastAt":"2007-01-15T06:56:35Z","messageCount":1,"participants":["Eric Wong"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"31752","messageId":"20070115065635.GA939@localdomain","threadId":"6387","inReplyTo":null,"subject":"git-svn: does anybody care for sub-optimal merge tracking?","fromName":"Eric Wong","fromEmail":"normalperson@yhbt.net","sentAt":"2007-01-15T06:56:35Z","receivedAt":"2007-01-15T06:56:35Z","isPatch":false,"sender":{"key":"e@80x24.org","avatar":null},"body":"First off, parental information on the creation of new tags and branches\nwill stay, meaning --follow-parent will remain in git-svn; and will\nprobably even become the default for fetching.\n\nHowever... there are some experimental features in git-svn\nthat I'm considering dropping:\n\n* tree matching: -b/--branch <refname> and -B/--branch-all-refs[1]\n* the graft-branches command\n\nTree matches are usually done at fetch time, and will add parents to an\nimported commit if the tree matches the one made in a commit.  I\noriginated this feature myself, but have not used it often.\n\nTree matching likely has problems with a dcommit-based\nworkflow (I haven't heard any complaints, either because nobody is using\nit or they're still using set-tree (formerly 'commit').\n\nTree matching will almost certainly go away unless somebody raises some\nvery good objections.\n\ngraft-branches attempts to do merge-tracking via regexes on log\nmessages, analyzing parenting information in logs, and also via tree\nmatching (like -b/-B).  git-svnimport does somewhat the same things\n(minus tree matching), but at import time (not after-the-fact with\ngrafts).\n\nDespite spending a reasonable amount of time creating graft-branches, I\nhave hardly ever used it on real projects; and the times I did use it I\ndidn't think it was very useful.  The regex matching on log messages and\nparenting information in logs should not be harmful to a dcommit-based\nworkflow.\n\n[1] - apologies for not documenting -B/--branch-all-refs in\n      the manpage, it's basically shorthand for:\n      `git-rev-parse --symbolic --all | xargs -n1 echo -b`\n\n-- \nEric Wong\n"}]}