{"thread":{"id":"21219","subject":"git-pack-objects gitattributes?","startedAt":"2009-10-12T23:00:08Z","lastAt":"2009-10-20T16:34:00Z","messageCount":4,"participants":["Nasser Grainawi","Michael J Gruber"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"124773","messageId":"4AD3B4F8.6030007@codeaurora.org","threadId":"21219","inReplyTo":null,"subject":"git-pack-objects gitattributes?","fromName":"Nasser Grainawi","fromEmail":"nasser@codeaurora.org","sentAt":"2009-10-12T23:00:08Z","receivedAt":"2009-10-12T23:00:08Z","isPatch":false,"sender":{"key":"nasser@codeaurora.org","avatar":"https://avatars.githubusercontent.com/u/757421?v=4"},"body":"Hello,\n\nI'm trying to avoid doing delta compression on a number of large binary files. \nI got a suggestion to use $GIT_DIR/info/attributes and a line like this:\n*.bin -delta\n\nThis doesn't seem to show a big improvement (and honestly I can't find where in \nthe git-pack-objects source the value of this attribute is used).\n\nCould someone shed some light on this attribute and any other improvements I \ncould make for efficiently serving up a repo over git-daemon with near-weekly \nrevisions of 100MB+ files?\n\nThanks,\nNasser\n"},{"id":"125406","messageId":"4ADCB457.8050601@codeaurora.org","threadId":"21219","inReplyTo":"4AD3B4F8.6030007@codeaurora.org","subject":"Re: git-pack-objects gitattributes?","fromName":"Nasser Grainawi","fromEmail":"nasser@codeaurora.org","sentAt":"2009-10-19T18:47:51Z","receivedAt":"2009-10-19T18:47:51Z","isPatch":false,"sender":{"key":"nasser@codeaurora.org","avatar":"https://avatars.githubusercontent.com/u/757421?v=4"},"body":"Nasser Grainawi wrote:\n> Hello,\n> \n> I'm trying to avoid doing delta compression on a number of large binary \n> files. I got a suggestion to use $GIT_DIR/info/attributes and a line \n> like this:\n> *.bin -delta\n> \n> This doesn't seem to show a big improvement (and honestly I can't find \n> where in the git-pack-objects source the value of this attribute is used).\n> \n> Could someone shed some light on this attribute and any other \n> improvements I could make for efficiently serving up a repo over \n> git-daemon with near-weekly revisions of 100MB+ files?\n> \n> Thanks,\n> Nasser\n\nping? any help? anyone?\n"},{"id":"125410","messageId":"4ADCBAC7.9010601@drmicha.warpmail.net","threadId":"21219","inReplyTo":"4ADCB457.8050601@codeaurora.org","subject":"Re: git-pack-objects gitattributes?","fromName":"Michael J Gruber","fromEmail":"git@drmicha.warpmail.net","sentAt":"2009-10-19T19:15:19Z","receivedAt":"2009-10-19T19:15:19Z","isPatch":false,"sender":{"key":"git@grubix.eu","avatar":"https://avatars.githubusercontent.com/u/233215?v=4"},"body":"Nasser Grainawi venit, vidit, dixit 19.10.2009 20:47:\n> Nasser Grainawi wrote:\n>> Hello,\n>>\n>> I'm trying to avoid doing delta compression on a number of large binary \n>> files. I got a suggestion to use $GIT_DIR/info/attributes and a line \n>> like this:\n>> *.bin -delta\n>>\n>> This doesn't seem to show a big improvement (and honestly I can't find \n>> where in the git-pack-objects source the value of this attribute is used).\n>>\n>> Could someone shed some light on this attribute and any other \n>> improvements I could make for efficiently serving up a repo over \n>> git-daemon with near-weekly revisions of 100MB+ files?\n>>\n>> Thanks,\n>> Nasser\n> \n> ping? any help? anyone?\n\nWell, describing a reproducable test case would help... as well as\ntelling us your git version.\n\nbuiltin-pack-objects.c certainly refers to the delta attribute, see\nno_try_delta() and its callers.\n\nHave you checked your attrs with git-check-attr? How do you measure the\nimprovements you expect?\n\nMichael\n"},{"id":"125490","messageId":"4ADDE678.4030001@codeaurora.org","threadId":"21219","inReplyTo":"4ADCBAC7.9010601@drmicha.warpmail.net","subject":"Re: git-pack-objects gitattributes?","fromName":"Nasser Grainawi","fromEmail":"nasser@codeaurora.org","sentAt":"2009-10-20T16:34:00Z","receivedAt":"2009-10-20T16:34:00Z","isPatch":false,"sender":{"key":"nasser@codeaurora.org","avatar":"https://avatars.githubusercontent.com/u/757421?v=4"},"body":"Michael J Gruber wrote:\n> Nasser Grainawi venit, vidit, dixit 19.10.2009 20:47:\n>> Nasser Grainawi wrote:\n>>> Hello,\n>>>\n>>> I'm trying to avoid doing delta compression on a number of large binary \n>>> files. I got a suggestion to use $GIT_DIR/info/attributes and a line \n>>> like this:\n>>> *.bin -delta\n>>>\n>>> This doesn't seem to show a big improvement (and honestly I can't find \n>>> where in the git-pack-objects source the value of this attribute is used).\n>>>\n>>> Could someone shed some light on this attribute and any other \n>>> improvements I could make for efficiently serving up a repo over \n>>> git-daemon with near-weekly revisions of 100MB+ files?\n>>>\n>>> Thanks,\n>>> Nasser\n>> ping? any help? anyone?\n> \n> Well, describing a reproducable test case would help... as well as\n> telling us your git version.\n\n1.6.5\n\n> \n> builtin-pack-objects.c certainly refers to the delta attribute, see\n> no_try_delta() and its callers.\n\nOops, somehow missed that while looking at the code right above it. Thanks.\n\n> \n> Have you checked your attrs with git-check-attr? How do you measure the\n> improvements you expect?\n\nI did check, it returns 'unset' like one would expect.\n\nI guess the big problem is that I don't have a good test case. I would have\nexpected a 'git repack -adf' to spend less time saying \"Compressing objects\",\nbut that doesn't seem to be happening...\n\nOh, wait, never mind.\nI was missing some of the binaries I was trying to skip. Adding some more \nexceptions to the attributes file dropped the \"Compressing objects\" time from \n20 minutes (or more) to maybe 10 seconds.\n\nThanks Michael!\n\nNasser\n"}]}