{"thread":{"id":"9776","subject":"git pull fails with http urls?","startedAt":"2007-09-04T14:10:53Z","lastAt":"2007-09-04T19:17:06Z","messageCount":2,"participants":["Richard Purdie","Junio C Hamano"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"52457","messageId":"1188915053.5705.12.camel@localhost.localdomain","threadId":"9776","inReplyTo":null,"subject":"git pull fails with http urls?","fromName":"Richard Purdie","fromEmail":"rpurdie@rpsys.net","sentAt":"2007-09-04T14:10:53Z","receivedAt":"2007-09-04T14:10:53Z","isPatch":false,"sender":{"key":"rpurdie@rpsys.net","avatar":null},"body":"I've noticed a problem with pulling http urls. I can clone them fine:\n\n  git clone http://www.gnome.org/~alexl/git/gio.git/\n\nand I can pull with:\n\n  git pull\n\nwhich works as expected however if I try:\n\n  git pull http://www.gnome.org/~alexl/git/gio.git/\n\nit fails with:\n\n  error: pick-rref: HEAD not found\n  No such ref HEAD at http://www.gnome.org/~alexl/git/gio.git/\n\nThe HEAD file is there at http://www.gnome.org/~alexl/git/gio.git/HEAD\nand:\n\n  git pull http://www.gnome.org/~alexl/git/gio.git/ master\n\nworks. It seems git just doesn't want to see/understand the HEAD file?\n\nIs the above supposed to work or is something wrong with the above\ncommands? I see this with git 1.5.2.3 and 1.5.3, google suggests this\nmight have worked with 1.4.x and broke in 1.5...\n\nRegards,\n\nRichard\n"},{"id":"52505","messageId":"7vsl5umea5.fsf@gitster.siamese.dyndns.org","threadId":"9776","inReplyTo":"1188915053.5705.12.camel@localhost.localdomain","subject":"Re: git pull fails with http urls?","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2007-09-04T19:17:06Z","receivedAt":"2007-09-04T19:17:06Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Pasky seems to have a locally modified version at repo.or.cz to\nadd HEAD to the result of update-server-info, but as far as I\ncan see, update-server-info never placed HEAD in the result, and\nI do not think ls-remote compensated it by adding HEAD by hand,\nso that change is fixing the symptom rather than fixing the\nregression, but I am not sure where the true regression is.\nMost likely we broke git-fetch while trying to deal with\nrepositories with insanely large number of refs (aka\n\"fetch--tool partial rewrite in C\").\n\nHaving said all that, I think it is a good idea nevertheless to\nadd HEAD information across all transports, like Pasky's patch\ndoes.  Currently http/ftp/rsync do not report HEAD while git\nnative protocols do.\n\nPasky, care to submit \"server-info: Add HEAD to info/refs\" for\ninclusion?\n"}]}