{"thread":{"id":"565","subject":"gitweb wishlist","startedAt":"2005-05-11T01:26:26Z","lastAt":"2005-06-04T08:29:21Z","messageCount":102,"participants":["Petr Baudis","YOSHIFUJI Hideaki / 吉藤英明","Kay Sievers","Jan-Benedict Glaw","Junio C Hamano","Jonas Fonseca","Linus Torvalds","Thomas Glanzmann","H. Peter Anvin","Matthias Urlichs","David Mansfield","Martin Langhoff","Edgar Toernig","David Greaves"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"3018","messageId":"20050511012626.GL26384@pasky.ji.cz","threadId":"565","inReplyTo":null,"subject":"gitweb wishlist","fromName":"Petr Baudis","fromEmail":"pasky@ucw.cz","sentAt":"2005-05-11T01:26:26Z","receivedAt":"2005-05-11T01:26:26Z","isPatch":false,"sender":{"key":"pasky@ucw.cz","avatar":"https://avatars.githubusercontent.com/u/18439?v=4"},"body":" Hello,\n\n I would be very happy if you could extend the gitweb scripts a little.\nBasically, what I need is to have ability to create a permanent link to\na given file in the repository, which stays same across revisions (as\nlong as the file stays with the given name, obviously).\n\n E.g. I would like to have something like\n\n\thttp://www.kernel.org/git/gitweb.cgi?p=cogito%2Fcogito.git;n=contrib/ciabot.pl\n\nfor file contrib/ciabot.pl in the latest Cogito tree, and\n\n\thttp://www.kernel.org/git/gitweb.cgi?p=cogito%2Fcogito.git;n=contrib\n\nfor the list of the contrib/ directory in the latest Cogito tree.\n\n I think I would prefer the link from the repository index to go not to\nthe log page, but some \"summary\" page, which would have some short\ninformation about the repository (owner, description, list of branches\nif gitweb supports that, list of tags, link to the latest tree and link\nto the log).\n\n BTW, why are people using ';' in the URIs for separators instead of\ngood old '&'s? Hmm, actually it just occurred to me that it might be to\nworkaround entity string problems?\n\n Thanks,\n\n-- \n\t\t\t\tPetr \"Pasky\" Baudis\nStuff: http://pasky.or.cz/\nC++: an octopus made by nailing extra legs onto a dog. -- Steve Taylor\n"},{"id":"3020","messageId":"20050511.104907.57995840.yoshfuji@linux-ipv6.org","threadId":"565","inReplyTo":"20050511012626.GL26384@pasky.ji.cz","subject":"Re: gitweb wishlist","fromName":"YOSHIFUJI Hideaki / 吉藤英明","fromEmail":"yoshfuji@linux-ipv6.org","sentAt":"2005-05-11T01:49:07Z","receivedAt":"2005-05-11T01:49:07Z","isPatch":false,"sender":{"key":"yoshfuji@linux-ipv6.org","avatar":"https://gravatar.com/avatar/92c288bb8d3e34afd7bb1574cc1079a3d0aa6d36d80955d1ec6a1cea2868d392?d=mp&s=160"},"body":"In article <20050511012626.GL26384@pasky.ji.cz> (at Wed, 11 May 2005 03:26:26 +0200), Petr Baudis says:\n\n> E.g. I would like to have something like\n> \n> \thttp://www.kernel.org/git/gitweb.cgi?p=cogito%2Fcogito.git;n=contrib/ciabot.pl\n ~~~ why not /?\n\n--yoshfuji\n"},{"id":"3022","messageId":"20050511020401.GM26384@pasky.ji.cz","threadId":"565","inReplyTo":"20050511.104907.57995840.yoshfuji@linux-ipv6.org","subject":"Re: gitweb wishlist","fromName":"Petr Baudis","fromEmail":"pasky@ucw.cz","sentAt":"2005-05-11T02:04:02Z","receivedAt":"2005-05-11T02:04:02Z","isPatch":false,"sender":{"key":"pasky@ucw.cz","avatar":"https://avatars.githubusercontent.com/u/18439?v=4"},"body":"Dear diary, on Wed, May 11, 2005 at 03:49:07AM CEST, I got a letter\nwhere \"YOSHIFUJI Hideaki / ?$B5HF#1QL@\" told me that...\n> In article <20050511012626.GL26384@pasky.ji.cz> (at Wed, 11 May 2005 03:26:26 +0200), Petr Baudis says:\n> \n> > E.g. I would like to have something like\n> > \n> > \thttp://www.kernel.org/git/gitweb.cgi?p=cogito%2Fcogito.git;n=contrib/ciabot.pl\n> ~~~ why not /?\n\nThis was a simple cut'n'paste. I don't care either way personally, but I\nthink it's better to be conservative when generating links and escape\nthis. OTOH it should obviously also accept a URL containing a slash\ninstead.\n\n-- \n\t\t\t\tPetr \"Pasky\" Baudis\nStuff: http://pasky.or.cz/\nC++: an octopus made by nailing extra legs onto a dog. -- Steve Taylor\n"},{"id":"3037","messageId":"1115801227.22799.13.camel@localhost.localdomain","threadId":"565","inReplyTo":"20050511012626.GL26384@pasky.ji.cz","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-11T08:47:07Z","receivedAt":"2005-05-11T08:47:07Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Wed, 2005-05-11 at 03:26 +0200, Petr Baudis wrote:\n> Hello,\n> \n> I would be very happy if you could extend the gitweb scripts a little.\n> Basically, what I need is to have ability to create a permanent link to\n> a given file in the repository, which stays same across revisions (as\n> long as the file stays with the given name, obviously).\n> \n> E.g. I would like to have something like\n> \n> \thttp://www.kernel.org/git/gitweb.cgi?p=cogito%2Fcogito.git;n=contrib/ciabot.pl\n> \n> for file contrib/ciabot.pl in the latest Cogito tree, and\n> \n> \thttp://www.kernel.org/git/gitweb.cgi?p=cogito%2Fcogito.git;n=contrib\n> \n> for the list of the contrib/ directory in the latest Cogito tree.\n\nWith the next round I will try to introduce a real browser through the\ntree instead of the current damn simple tree-object list. It would be\nnice if it can show the the current path and provide a \"up to parent\"\nway. Along with that it should be easy to add the stuff you asked for.\n\n> I think I would prefer the link from the repository index to go not to\n> the log page, but some \"summary\" page, which would have some short\n> information about the repository (owner, description, list of branches\n> if gitweb supports that, list of tags, link to the latest tree and link\n> to the log).\n> \n> BTW, why are people using ';' in the URIs for separators instead of\n> good old '&'s? Hmm, actually it just occurred to me that it might be to\n> workaround entity string problems?\n\nYes, but it's \"bad old\" not \"good\". :)\n http://www.w3.org/TR/html4/appendix/notes.html#h-B.2.2\n\nThanks,\nKay\n\n"},{"id":"3038","messageId":"20050511093015.GK8176@lug-owl.de","threadId":"565","inReplyTo":"20050511012626.GL26384@pasky.ji.cz","subject":"Re: gitweb wishlist","fromName":"Jan-Benedict Glaw","fromEmail":"jbglaw@lug-owl.de","sentAt":"2005-05-11T09:30:15Z","receivedAt":"2005-05-11T09:30:15Z","isPatch":false,"sender":{"key":"jbglaw@lug-owl.de","avatar":null},"body":"On Wed, 2005-05-11 03:26:26 +0200, Petr Baudis wrote:\n> I would be very happy if you could extend the gitweb scripts a little.\n> Basically, what I need is to have ability to create a permanent link to\n> a given file in the repository, which stays same across revisions (as\n> long as the file stays with the given name, obviously).\n\nI've got another one. I'd like to see st_mtime on the file lists to see\nwhen a file was last touched...\n\nMfG, JBG\n\n-- \nJan-Benedict Glaw jbglaw@lug-owl.de . +49-172-7608481 _ O _\n\"Eine Freie Meinung in einem Freien Kopf | Gegen Zensur | Gegen Krieg _ _ O\n fuer einen Freien Staat voll Freier Bürger\" | im Internet! | im Irak! O O O\nret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));\n"},{"id":"3186","messageId":"7v3bssfbsj.fsf@assigned-by-dhcp.cox.net","threadId":"565","inReplyTo":"20050511012626.GL26384@pasky.ji.cz","subject":"Re: gitweb wishlist","fromName":"Junio C Hamano","fromEmail":"junkio@cox.net","sentAt":"2005-05-12T20:07:40Z","receivedAt":"2005-05-12T20:07:40Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"* [Previous page] [Next page] would be nice in addition to last\n 10, day, week, etc.\n\n* Putting the commit headline and \"X hour\"s ago in a separate\n div or span next to each other, so that a long commit headline\n wraps properly and does not start the second line just under\n the \"X hours ago\" timestamp would be nicer (you can see what I\n mean easily by narrowing the browser window).\n\n"},{"id":"3197","messageId":"1115931627.8465.6.camel@dhcp-188.off.vrfy.org","threadId":"565","inReplyTo":"7v3bssfbsj.fsf@assigned-by-dhcp.cox.net","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-12T21:00:26Z","receivedAt":"2005-05-12T21:00:26Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Thu, 2005-05-12 at 13:07 -0700, Junio C Hamano wrote:\n> * [Previous page] [Next page] would be nice in addition to last\n> 10, day, week, etc.\n\nThat should be easy to do with the parameters we have now for the\ngit-rev-list. I will first finish the new browser through the\ntrees/files, then the project overview page and after that try the\npager,\n\n> * Putting the commit headline and \"X hour\"s ago in a separate\n> div or span next to each other, so that a long commit headline\n> wraps properly and does not start the second line just under\n> the \"X hours ago\" timestamp would be nicer (you can see what I\n> mean easily by narrowing the browser window).\n\nBlock elements (div) are not allowed inside an a-tag in XHTML/Strict -\ndon't know how to do this, cause the whole headline should be a link\nwithout the use of javascript. :)\n\nThanks,\nKay\n\n"},{"id":"3200","messageId":"7vy8akdtxz.fsf@assigned-by-dhcp.cox.net","threadId":"565","inReplyTo":"1115931627.8465.6.camel@dhcp-188.off.vrfy.org","subject":"Re: gitweb wishlist","fromName":"Junio C Hamano","fromEmail":"junkio@cox.net","sentAt":"2005-05-12T21:18:32Z","receivedAt":"2005-05-12T21:18:32Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":">>>>> \"KS\" == Kay Sievers writes:\n\n>> * Putting the commit headline and \"X hour\"s ago in a separate\n>> div or span next to each other, so that a long commit headline\n>> wraps properly and does not start the second line just under\n>> the \"X hours ago\" timestamp would be nicer (you can see what I\n>> mean easily by narrowing the browser window).\n\nKS> Block elements (div) are not allowed inside an a-tag in XHTML/Strict -\nKS> don't know how to do this,...\n\nWouldn't splitting the commit headline and \"X hours ago\" into\ntwo separate elements and wrap them individually inside an a-tag\neach pointing at the same destination good enough then?\n\n"},{"id":"3245","messageId":"20050513120618.GA26263@diku.dk","threadId":"565","inReplyTo":"20050511012626.GL26384@pasky.ji.cz","subject":"Re: gitweb wishlist","fromName":"Jonas Fonseca","fromEmail":"fonseca@diku.dk","sentAt":"2005-05-13T12:06:18Z","receivedAt":"2005-05-13T12:06:18Z","isPatch":false,"sender":{"key":"fonseca@diku.dk","avatar":"https://gravatar.com/avatar/f82f3ad698717c51873b020c750a92438c820a24056dc39fe4d07baa10a92264?d=mp&s=160"},"body":"I don't know if this is intentional, but it looks like gitweb discards\neverything after the first line starting with 'Signed-off-by:' on the\nlog page. This will in some cases remove valuable log information when\na second author or the committer adds additional comments to a commit:\n\n \n\n Signed-off-by: \n\n \n\n Signed-off-by: \n\nThe commit page gets it right, which is why I suspect it might just be a\nway to trim the amount of text on the log page.\n\nI also noticed that there is a 'faulty' signed-off-by line in commit\n14ebb908e10f068dc1901d35f4b716bc69143d19 in case the above is\nintentional. Dunno if that should be matched by relaxing the regexp a\nlittle.\n\n-- \nJonas Fonseca\n"},{"id":"3277","messageId":"1116038097.9317.11.camel@dhcp-188.off.vrfy.org","threadId":"565","inReplyTo":"20050513120618.GA26263@diku.dk","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-14T02:34:57Z","receivedAt":"2005-05-14T02:34:57Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Fri, 2005-05-13 at 14:06 +0200, Jonas Fonseca wrote:\n> I don't know if this is intentional, but it looks like gitweb discards\n> everything after the first line starting with 'Signed-off-by:' on the\n> log page. This will in some cases remove valuable log information when\n> a second author or the committer adds additional comments to a commit:\n> \n> \n> \n> Signed-off-by: \n> \n> \n> \n> Signed-off-by: \n> \n> The commit page gets it right, which is why I suspect it might just be a\n> way to trim the amount of text on the log page.\n> \n> I also noticed that there is a 'faulty' signed-off-by line in commit\n> 14ebb908e10f068dc1901d35f4b716bc69143d19 in case the above is\n> intentional. Dunno if that should be matched by relaxing the regexp a\n> little.\n\nAll that should work now.\n\nThanks,\nKay\n\n"},{"id":"3278","messageId":"1116038370.9317.17.camel@dhcp-188.off.vrfy.org","threadId":"565","inReplyTo":"20050511093015.GK8176@lug-owl.de","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-14T02:39:30Z","receivedAt":"2005-05-14T02:39:30Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Wed, 2005-05-11 at 11:30 +0200, Jan-Benedict Glaw wrote:\n> On Wed, 2005-05-11 03:26:26 +0200, Petr Baudis wrote:\n> > I would be very happy if you could extend the gitweb scripts a little.\n> > Basically, what I need is to have ability to create a permanent link to\n> > a given file in the repository, which stays same across revisions (as\n> > long as the file stays with the given name, obviously).\n> \n> I've got another one. I'd like to see st_mtime on the file lists to see\n> when a file was last touched...\n\nWe don't have that information in the trees. The date is only available\nin the commits and reading all commits or storing that information in a\nindex or a database is beyond the scope of that simple cgi.\n\nThanks,\nKay\n\n"},{"id":"3279","messageId":"1116038598.9317.22.camel@dhcp-188.off.vrfy.org","threadId":"565","inReplyTo":"20050511012626.GL26384@pasky.ji.cz","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-14T02:43:18Z","receivedAt":"2005-05-14T02:43:18Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Wed, 2005-05-11 at 03:26 +0200, Petr Baudis wrote:\n> I would be very happy if you could extend the gitweb scripts a little.\n> Basically, what I need is to have ability to create a permanent link to\n> a given file in the repository, which stays same across revisions (as\n> long as the file stays with the given name, obviously).\n> \n> E.g. I would like to have something like\n> \n> \thttp://www.kernel.org/git/gitweb.cgi?p=cogito%2Fcogito.git;n=contrib/ciabot.pl\n> \n> for file contrib/ciabot.pl in the latest Cogito tree, and\n\nhttp://www.kernel.org/git/gitweb.cgi?p=cogito/cogito.git;a=blob;f=contrib/ciabot.pl\n\n> \thttp://www.kernel.org/git/gitweb.cgi?p=cogito%2Fcogito.git;n=contrib\n> \n> for the list of the contrib/ directory in the latest Cogito tree.\n\nhttp://www.kernel.org/git/gitweb.cgi?p=cogito/cogito.git;a=tree;f=contrib\n\n> I think I would prefer the link from the repository index to go not to\n> the log page, but some \"summary\" page, which would have some short\n> information about the repository (owner, description, list of branches\n> if gitweb supports that, list of tags, link to the latest tree and link\n> to the log).\n\nSounds reasonable, I will try that with the next round.\n\nThanks,\nKay\n\n"},{"id":"3299","messageId":"20050514105457.GA21705@diku.dk","threadId":"565","inReplyTo":"1116038598.9317.22.camel@dhcp-188.off.vrfy.org","subject":"Re: gitweb wishlist","fromName":"Jonas Fonseca","fromEmail":"fonseca@diku.dk","sentAt":"2005-05-14T10:54:57Z","receivedAt":"2005-05-14T10:54:57Z","isPatch":false,"sender":{"key":"fonseca@diku.dk","avatar":"https://gravatar.com/avatar/f82f3ad698717c51873b020c750a92438c820a24056dc39fe4d07baa10a92264?d=mp&s=160"},"body":"Kay Sievers wrote Sat, May 14, 2005:\n> On Wed, 2005-05-11 at 03:26 +0200, Petr Baudis wrote:\n> > I would be very happy if you could extend the gitweb scripts a little.\n> > Basically, what I need is to have ability to create a permanent link to\n> > a given file in the repository, which stays same across revisions (as\n> > long as the file stays with the given name, obviously).\n> > \n> > E.g. I would like to have something like\n> > \n> > \thttp://www.kernel.org/git/gitweb.cgi?p=cogito%2Fcogito.git;n=contrib/ciabot.pl\n> > \n> > for file contrib/ciabot.pl in the latest Cogito tree, and\n> \n> http://www.kernel.org/git/gitweb.cgi?p=cogito/cogito.git;a=blob;f=contrib/ciabot.pl\n\nHow about support for getting the 'raw' files and diffs without all the\nHTML markup?\n\n-- \nJonas Fonseca\n"},{"id":"3486","messageId":"1116384951.5094.83.camel@dhcp-188.off.vrfy.org","threadId":"565","inReplyTo":"20050511012626.GL26384@pasky.ji.cz","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-18T02:55:51Z","receivedAt":"2005-05-18T02:55:51Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Wed, 2005-05-11 at 03:26 +0200, Petr Baudis wrote:\n\n> I think I would prefer the link from the repository index to go not to\n> the log page, but some \"summary\" page, which would have some short\n> information about the repository (owner, description, list of branches\n> if gitweb supports that, list of tags, link to the latest tree and link\n> to the log).\n\nI did this now. The top-link shows now the repository listing with a\nnice \"last change\" field. The default link points to an overview page\nwhich also list the tags.\n(The owner filed in that list is not correct until now, cause the\ncron-job needs to be adapted.)\n\nThanks,\nKay\n\n"},{"id":"3492","messageId":"20050518094545.GC25416@pasky.ji.cz","threadId":"565","inReplyTo":"1116384951.5094.83.camel@dhcp-188.off.vrfy.org","subject":"Re: gitweb wishlist","fromName":"Petr Baudis","fromEmail":"pasky@ucw.cz","sentAt":"2005-05-18T09:45:46Z","receivedAt":"2005-05-18T09:45:46Z","isPatch":false,"sender":{"key":"pasky@ucw.cz","avatar":"https://avatars.githubusercontent.com/u/18439?v=4"},"body":"Dear diary, on Wed, May 18, 2005 at 04:55:51AM CEST, I got a letter\nwhere Kay Sievers told me that...\n> On Wed, 2005-05-11 at 03:26 +0200, Petr Baudis wrote:\n> \n> > I think I would prefer the link from the repository index to go not to\n> > the log page, but some \"summary\" page, which would have some short\n> > information about the repository (owner, description, list of branches\n> > if gitweb supports that, list of tags, link to the latest tree and link\n> > to the log).\n> \n> I did this now. The top-link shows now the repository listing with a\n> nice \"last change\" field. The default link points to an overview page\n> which also list the tags.\n> (The owner filed in that list is not correct until now, cause the\n> cron-job needs to be adapted.)\n\nThanks, this is exactly how I envisioned it. :-)\n\n-- \n\t\t\t\tPetr \"Pasky\" Baudis\nStuff: http://pasky.or.cz/\nC++: an octopus made by nailing extra legs onto a dog. -- Steve Taylor\n"},{"id":"3622","messageId":"Pine.LNX.4.58.0505200948150.2206@ppc970.osdl.org","threadId":"565","inReplyTo":"1116384951.5094.83.camel@dhcp-188.off.vrfy.org","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-20T16:54:24Z","receivedAt":"2005-05-20T16:54:24Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Wed, 18 May 2005, Kay Sievers wrote:\n> \n> I did this now. The top-link shows now the repository listing with a\n> nice \"last change\" field. The default link points to an overview page\n> which also list the tags.\n\nIn the summary page, could we get authorship information too? Right now it \nlooks like\n\n\trecent commits\n\n\t15 minutes ago\t[PATCH] Add tests for diff-tree\n\t31 minutes ago\tdiff-tree: use new base_name_compare() helper function\n\t34 minutes ago\tIntroduce \"base_name_compare()\" helper function \n\t...\n\nand wouldn't it be nice if it told you who had written these things, like\n\n\trecent commits\n\n\t15 minutes ago\tJunio C Hamano\t[PATCH] Add tests for diff-tree\n\t31 minutes ago\tLinus Torvalds\tdiff-tree: use new base_name_compare() helper function\n\t34 minutes ago\tLinus Torvalds\tIntroduce \"base_name_compare()\" helper function \n\t...\n\n(limit the name to the first 20 characters or something to make things \nline up).\n\nIf the lines get too long, you could changes \"minutes\" to \"min\", and maybe \nlimit the name to the first 10 characters (or even first name, but there's \na lot of David's around, while 10-12 characters tends to get enough of \nthe last name to be reasonable unique).\n\n\t\tLinus\n"},{"id":"3624","messageId":"7vbr75u8vw.fsf@assigned-by-dhcp.cox.net","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505200948150.2206@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Junio C Hamano","fromEmail":"junkio@cox.net","sentAt":"2005-05-20T17:04:03Z","receivedAt":"2005-05-20T17:04:03Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":">>>>> \"LT\" == Linus Torvalds writes:\n\nLT> If the lines get too long, you could changes \"minutes\" to \"min\", and maybe \nLT> limit the name to the first 10 characters (or even first name, but there's \nLT> a lot of David's around, while 10-12 characters tends to get enough of \nLT> the last name to be reasonable unique).\n\nOr abbreviate the first names. L Torvalds, JC Hamano, etc.\n\n"},{"id":"3626","messageId":"Pine.LNX.4.58.0505201019460.2206@ppc970.osdl.org","threadId":"565","inReplyTo":"7vbr75u8vw.fsf@assigned-by-dhcp.cox.net","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-20T17:21:23Z","receivedAt":"2005-05-20T17:21:23Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Fri, 20 May 2005, Junio C Hamano wrote:\n> \n> Or abbreviate the first names. L Torvalds, JC Hamano, etc.\n\nTechnically, yes. Except I end up at least personally going by either\nfirst names (or _possibly_ email addresses), not by last name. Maybe Linux\nis fairly unique in that, but dammit, I'm not \"Torvalds\", I'm \"Linus\". And\nlike it or not, you're either Junio or junkio at least to me ;)\n\n\t\tLinus\n"},{"id":"3628","messageId":"1116611932.12975.22.camel@dhcp-188","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505200948150.2206@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-20T17:58:52Z","receivedAt":"2005-05-20T17:58:52Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Fri, 2005-05-20 at 09:54 -0700, Linus Torvalds wrote:\n> \n> On Wed, 18 May 2005, Kay Sievers wrote:\n> > \n> > I did this now. The top-link shows now the repository listing with a\n> > nice \"last change\" field. The default link points to an overview page\n> > which also list the tags.\n> \n> In the summary page, could we get authorship information too? Right now it \n> looks like\n> \n> \trecent commits\n> \n> \t15 minutes ago\t[PATCH] Add tests for diff-tree\n> \t31 minutes ago\tdiff-tree: use new base_name_compare() helper function\n> \t34 minutes ago\tIntroduce \"base_name_compare()\" helper function \n> \t...\n> \n> and wouldn't it be nice if it told you who had written these things, like\n> \n> \trecent commits\n> \n> \t15 minutes ago\tJunio C Hamano\t[PATCH] Add tests for diff-tree\n> \t31 minutes ago\tLinus Torvalds\tdiff-tree: use new base_name_compare() helper function\n> \t34 minutes ago\tLinus Torvalds\tIntroduce \"base_name_compare()\" helper function \n> \t...\n> \n> (limit the name to the first 20 characters or something to make things \n> line up).\n\nSomething like that: :)\n http://www.kernel.org/git/?p=cogito/cogito.git;a=summary\n\nKay\n\n"},{"id":"3629","messageId":"Pine.LNX.4.58.0505201111090.2206@ppc970.osdl.org","threadId":"565","inReplyTo":"1116611932.12975.22.camel@dhcp-188","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-20T18:16:00Z","receivedAt":"2005-05-20T18:16:00Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Fri, 20 May 2005, Kay Sievers wrote:\n> \n> Something like that: :)\n> http://www.kernel.org/git/?p=cogito/cogito.git;a=summary\n\nLooking good.\n\nI still think \"minutes\" is an awfully long word to waste vertical space\nwith, especially since it's also the only one that will have up to three\ndigits associated with it. It seems kind of silly to give more (or even\nequal) space to the simplified time field than to the person who did\nsomething.\n\n(To see minutes, you need to use \"git/git.git\" instead of cogito.\n\nSo I still think you should change \"minutes\" to \"min\", which is the common \nabbreviation.\n\n(That \"ago\" is also very repetitive, but removing it would seem to make it \nunreadable, so I guess it's needed).\n\n\t\t\t\tLinus\n"},{"id":"3630","messageId":"Pine.LNX.4.58.0505201123550.2206@ppc970.osdl.org","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505201111090.2206@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-20T18:28:37Z","receivedAt":"2005-05-20T18:28:37Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Fri, 20 May 2005, Linus Torvalds wrote:\n>\n> Looking good.\n\nOh, dang, while I'm at it, why not ask for the \"commmitdiff\" thing to have \nthe commit message in it too, ie basically look like a prettified version \nof \"git-diff-tree -v -M \"\n\nYou already do the first line of it, so there's not much missing. Right\nnow there is no place where everything important from one commit is\n\"brought together\" (ie no single place where you can see both what the\ndiff is all about, and what it actually does).\n\n\t\tLinus\n"},{"id":"3633","messageId":"1116615539.12975.31.camel@dhcp-188","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505201111090.2206@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-20T18:58:59Z","receivedAt":"2005-05-20T18:58:59Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Fri, 2005-05-20 at 11:16 -0700, Linus Torvalds wrote:\n> \n> On Fri, 20 May 2005, Kay Sievers wrote:\n> > \n> > Something like that: :)\n> > http://www.kernel.org/git/?p=cogito/cogito.git;a=summary\n> \n> Looking good.\n> \n> I still think \"minutes\" is an awfully long word to waste vertical space\n> with, especially since it's also the only one that will have up to three\n> digits associated with it.\n\nChanged to \"min\".\n\nKay\n\n"},{"id":"3634","messageId":"1116615600.12975.33.camel@dhcp-188","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505201123550.2206@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-20T19:00:00Z","receivedAt":"2005-05-20T19:00:00Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Fri, 2005-05-20 at 11:28 -0700, Linus Torvalds wrote:\n> \n> On Fri, 20 May 2005, Linus Torvalds wrote:\n> >\n> > Looking good.\n> \n> Oh, dang, while I'm at it, why not ask for the \"commmitdiff\" thing to have \n> the commit message in it too, ie basically look like a prettified version \n> of \"git-diff-tree -v -M \"\n> \n> You already do the first line of it, so there's not much missing. Right\n> now there is no place where everything important from one commit is\n> \"brought together\" (ie no single place where you can see both what the\n> diff is all about, and what it actually does).\n\nSomehting like this?:\n http://kernel.org/git/?p=git/git.git;a=commitdiff;h=de809dbbce497e0d107562615c1d85ff35b4e0c5\n\nKay\n\n"},{"id":"3637","messageId":"20050520191308.GU8105@cip.informatik.uni-erlangen.de","threadId":"565","inReplyTo":"1116615600.12975.33.camel@dhcp-188","subject":"Re: gitweb wishlist","fromName":"Thomas Glanzmann","fromEmail":"sithglan@stud.uni-erlangen.de","sentAt":"2005-05-20T19:13:08Z","receivedAt":"2005-05-20T19:13:08Z","isPatch":false,"sender":{"key":"sithglan@stud.uni-erlangen.de","avatar":null},"body":"Hello Kay,\nI would like to see that I can klick on the file instead of the seperate\n'blob' link in a directory view, becasue that is more intuitive and you can\nalready an klick on directories:\n\nhttp://www.kernel.org/git/?p=git/git.git;a=tree;h=665a48af9e192ed84d2707c95d4c0d9c45eb45ad;hb=411746940f02f6fb90c4b6b97c6f07cee599c2e1\n\nThanks for this great tool!\n\nSincerely,\n\tThomas\n"},{"id":"3636","messageId":"Pine.LNX.4.58.0505201211390.2206@ppc970.osdl.org","threadId":"565","inReplyTo":"1116615600.12975.33.camel@dhcp-188","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-20T19:13:26Z","receivedAt":"2005-05-20T19:13:26Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Fri, 20 May 2005, Kay Sievers wrote:\n> \n> Somehting like this?:\n> http://kernel.org/git/?p=git/git.git;a=commitdiff;h=de809dbbce497e0d107562615c1d85ff35b4e0c5\n\nYes, except I don't think you should repeat the part of the commit message\nthat you used as a header (so in your example, since you already used \"Fix\nup previous commit\" in the header, don't show it below).\n\nBtw, any chance to see gitweb itself under gitweb on kernel.org?\n\n\t\tLinus\n"},{"id":"3638","messageId":"Pine.LNX.4.58.0505201219420.2206@ppc970.osdl.org","threadId":"565","inReplyTo":"1116615600.12975.33.camel@dhcp-188","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-20T19:22:39Z","receivedAt":"2005-05-20T19:22:39Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Fri, 20 May 2005, Kay Sievers wrote:\n> \n> Somehting like this?:\n> http://kernel.org/git/?p=git/git.git;a=commitdiff;h=de809dbbce497e0d107562615c1d85ff35b4e0c5\n\nBtw, at least for me, this looks much more interesting than the \"commit\" \nthing, and maybe it would make sense to make the summary links be to the \n\"commitdiff\" instead of the \"commit\"?\n\nOr is it just so much more expensive to generate, that we want to not have\npeople go there normally? (hpa cc'd, since he may have some insight into\nwhether this is likely to be an issue or not? It's not like git-diff-tree\nis that expensive, but it _does_ end up doing a \"diff\" against each\nchanged file, of course, modulo any caching of results).\n\n\t\tLinus\n"},{"id":"3641","messageId":"428E49DD.406@zytor.com","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505201219420.2206@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"H. Peter Anvin","fromEmail":"hpa@zytor.com","sentAt":"2005-05-20T20:34:37Z","receivedAt":"2005-05-20T20:34:37Z","isPatch":false,"sender":{"key":"hpa@zytor.com","avatar":null},"body":"Linus Torvalds wrote:\n> \n> On Fri, 20 May 2005, Kay Sievers wrote:\n> \n>>Somehting like this?:\n>> http://kernel.org/git/?p=git/git.git;a=commitdiff;h=de809dbbce497e0d107562615c1d85ff35b4e0c5\n> \n> \n> Btw, at least for me, this looks much more interesting than the \"commit\" \n> thing, and maybe it would make sense to make the summary links be to the \n> \"commitdiff\" instead of the \"commit\"?\n> \n> Or is it just so much more expensive to generate, that we want to not have\n> people go there normally? (hpa cc'd, since he may have some insight into\n> whether this is likely to be an issue or not? It's not like git-diff-tree\n> is that expensive, but it _does_ end up doing a \"diff\" against each\n> changed file, of course, modulo any caching of results).\n> \n\nWhat I ended up doing for the diff viewer on kernel.org is that every \npage that's generated gets stuffed in a cache (locklessly indexed by a \nSHA-1 of a canonicalized form of the query); the pages people actually \nsee are then simply pulled from the cache. This caching was a just \nenormous win. In the case of the diff viewer, the header is generated \neach time, since I allow the user to select a custom style sheet (and \ndon't want to cache versions for each style sheet), but that's a trivial \ndetail.\n\n\t-hpa\n\n"},{"id":"3642","messageId":"Pine.LNX.4.58.0505201346330.2206@ppc970.osdl.org","threadId":"565","inReplyTo":"428E49DD.406@zytor.com","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-20T20:49:28Z","receivedAt":"2005-05-20T20:49:28Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Fri, 20 May 2005, H. Peter Anvin wrote:\n\n> \n> What I ended up doing for the diff viewer on kernel.org is that every \n> page that's generated gets stuffed in a cache (locklessly indexed by a \n> SHA-1 of a canonicalized form of the query); the pages people actually \n> see are then simply pulled from the cache. This caching was a just \n> enormous win.\n\nOk. That still leaves the bandwidth issue (the full diffs are bigger than \nthe commit object), but usually the diffs in individual commits aren't \n_that_ large, so maybe it's a non-issue.\n\nOh, btw, I notice that you moved klibc over to git - care to share your\ncvs->git script (I assume you scripted it ;)? That would seem to be an \nobvious addition to the core stuff..\n\n\t\tLinus\n"},{"id":"3643","messageId":"428E4D8C.3020606@zytor.com","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505201346330.2206@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"H. Peter Anvin","fromEmail":"hpa@zytor.com","sentAt":"2005-05-20T20:50:20Z","receivedAt":"2005-05-20T20:50:20Z","isPatch":false,"sender":{"key":"hpa@zytor.com","avatar":null},"body":"Linus Torvalds wrote:\n> \n> Oh, btw, I notice that you moved klibc over to git - care to share your\n> cvs->git script (I assume you scripted it ;)? That would seem to be an \n> obvious addition to the core stuff..\n> \n\nActually, Kay did the conversion... the scripts are clearly very \ncantankerous, because if *I* run them -- I tried -- they don't work! \nSince it's Kay's work, I'll leave them to him, but I would definitely \nlove to move more of my CVS repos over to git, especially syslinux.\n\n\t-hpa\n\n"},{"id":"3648","messageId":"20050520211634.GA20003@cip.informatik.uni-erlangen.de","threadId":"565","inReplyTo":"428E4D8C.3020606@zytor.com","subject":"Re: gitweb wishlist","fromName":"Thomas Glanzmann","fromEmail":"sithglan@stud.uni-erlangen.de","sentAt":"2005-05-20T21:16:34Z","receivedAt":"2005-05-20T21:16:34Z","isPatch":false,"sender":{"key":"sithglan@stud.uni-erlangen.de","avatar":null},"body":"Hello,\nI imported the mutt-cvs for the 1.5 branch into GIT using the following\nscript. But it is a hack. I also think that I will use something like\nthat to build a CVS->GIT vendortracking.\n\ncvsps -x -z 10 -b HEAD -g -p ../../patches/\n\nAnd using the attached script to import the patches in GIT. It works\nquiet well.\n\nSee also msgid: 1115080139.21105.18.camel@localhost.localdomain there\nare the scripts which he used to convert the CVS to GIT for HPA. My\nscripts are based on his work.\n\n\tThomas\n\n\n#!/usr/bin/perl\n\nuse strict;\nuse warnings;\nuse File::Temp qw/ tempfile tempdir /;\n\n# ---------------------\n# PatchSet 1 \n# Date: 2002/07/23 07:41:30\n# Author: hpa\n# Branch: HEAD\n# Tag: (none) \n# Log:\n# Initial revision\n# \n# Members: \n# \tklibc.cvsroot/snprintf.c:INITIAL->1.1 \n# \tklibc.cvsroot/vsnprintf.c:INITIAL->1.1 \n# \tklibc.cvsroot/klibc/Makefile:INITIAL->1.1 \n# \tklibc.cvsroot/klibc/snprintf.c:INITIAL->1.1 \n# \tklibc.cvsroot/klibc/vsnprintf.c:INITIAL->1.1 \n# \n# --- /dev/null\t2005-04-30 18:00:24.840397008 +0200\n# +++ klibc/klibc.cvsroot/snprintf.c\t2005-05-02 19:57:42.879913000 +0200\n# @@ -0,0 +1,19 @@\n# +/*\n\nmy $patch = $ARGV[0];\n\nmy %committer = (\n\tbrendan => [ 'Brendan Cully', 'brendan@kublai.com' ],\n\tme => [ 'Michael Elkins', 'me@sigpipe.org' ],\n\troessler => [ 'Thomas Roessler', 'roessler@does-not-exist.org' ]\n);\n\nmy @log = ();\n\n$ENV{GIT_AUTHOR_EMAIL} = \"\";\n$ENV{GIT_COMMITTER_EMAIL} = \"\";\n\nopen (my $fd, $patch);\nwhile (my $line = <$fd>) {\n\tif ($line =~ m/^Date: (.*)/) {\n\t\t$ENV{GIT_AUTHOR_DATE} = $1;\n\n\t} elsif ($line =~ m/^Author: (.*)/) {\n\t\tif (defined($committer{$1})) {\n\t\t\t$ENV{GIT_COMMITTER_NAME} = @{$committer{$1}}[0];\n\t\t\t$ENV{GIT_COMMITTER_EMAIL} = @{$committer{$1}}[1];\n\t\t\t$ENV{GIT_AUTHOR_NAME} = @{$committer{$1}}[0];\n\t\t\t$ENV{GIT_AUTHOR_EMAIL} = @{$committer{$1}}[1];\n\t\t} else {\n\t\t\t$ENV{GIT_COMMITTER_NAME} = $1;\n\t\t\t$ENV{GIT_AUTHOR_NAME} = $1;\n\t\t}\n\n\t} elsif ($line =~ m/^Log:/) {\n\t\twhile (my $line = <$fd>) {\n\t\t\tif ($line =~ m/^Members: $/) {\n\t\t\t\tpop(@log);\n\t\t\t\tlast;\n\t\t\t} elsif ($line =~ /^From: (.+) <([^>]+@[^>]+)>$/) {\n\t\t\t\t$ENV{GIT_AUTHOR_NAME} = $1;\n\t\t\t\t$ENV{GIT_AUTHOR_EMAIL} = $2;\n\t\t\t}\n\t\t\tpush @log, $line;\n\t\t}\n\t}\n}\n\nclose($fd);\n\nmy ($fh, $logfile) = tempfile(CLEANUP => 1);\nprint $fh @log;\nsystem(\"git patch $patch < $logfile\");\nclose($fh);\n"},{"id":"3649","messageId":"1116625268.12975.103.camel@dhcp-188","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505201219420.2206@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-20T21:41:07Z","receivedAt":"2005-05-20T21:41:07Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Fri, 2005-05-20 at 12:22 -0700, Linus Torvalds wrote:\n> \n> On Fri, 20 May 2005, Kay Sievers wrote:\n> > \n> > Somehting like this?:\n> > http://kernel.org/git/?p=git/git.git;a=commitdiff;h=de809dbbce497e0d107562615c1d85ff35b4e0c5\n> \n> Btw, at least for me, this looks much more interesting than the \"commit\" \n> thing, and maybe it would make sense to make the summary links be to the \n> \"commitdiff\" instead of the \"commit\"?\n\nHow about this:\n http://www.kernel.org/git/?p=git/git.git;a=summary\n\nThe default link is still the same, but you can use the link at the end.\n\nKay\n\n"},{"id":"3650","messageId":"1116626652.12975.118.camel@dhcp-188","threadId":"565","inReplyTo":"428E4D8C.3020606@zytor.com","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-20T22:04:12Z","receivedAt":"2005-05-20T22:04:12Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Fri, 2005-05-20 at 13:50 -0700, H. Peter Anvin wrote:\n> Linus Torvalds wrote:\n> > \n> > Oh, btw, I notice that you moved klibc over to git - care to share your\n> > cvs->git script (I assume you scripted it ;)? That would seem to be an \n> > obvious addition to the core stuff..\n> > \n> \n> Actually, Kay did the conversion... the scripts are clearly very \n> cantankerous, because if *I* run them -- I tried -- they don't work! \n> Since it's Kay's work, I'll leave them to him, but I would definitely \n> love to move more of my CVS repos over to git, especially syslinux.\n\nHere we go;\n\nThese scripts are just a quick hack, I just wanted to know how nice the\nstupid cvs file history can be converted to git-committs.\n\nIt exports the CVS repo with the help of the nice cvsps to individual\npatches. (Every patch contains something like a \"ChangeSet\" by searching\nfor file revisions with the same checkin-date)\n\nThen the patches with the header are split into individual files for\ncommitting it into git (similar to Linus' git-mbox-tools).\n\nIf we reach a CVS tag with a patch during sequential patching, the\nscript throws away the whole current working tree and checks the\nrevision out of CVS. This way we make sure, that the git-tag matches\ntree CVS has tagged. (I've encountered two mismatches in the\n\"patch-chain\" with the CVS revision-tag. These corrections are hardcoded\ninto the script. :)\n\nFor every CVS revision-tag a git-tag without any content except the name\nis created.\n\nAnd the klibc-repo was created with a patched git-commit to fake the\ncommit date with the author date. :)\n\nGood luck with it,\nKay\n"},{"id":"3651","messageId":"428E6123.8050907@zytor.com","threadId":"565","inReplyTo":"1116626652.12975.118.camel@dhcp-188","subject":"Re: gitweb wishlist","fromName":"H. Peter Anvin","fromEmail":"hpa@zytor.com","sentAt":"2005-05-20T22:13:55Z","receivedAt":"2005-05-20T22:13:55Z","isPatch":false,"sender":{"key":"hpa@zytor.com","avatar":null},"body":"Kay Sievers wrote:\n> \n> And the klibc-repo was created with a patched git-commit to fake the\n> commit date with the author date. :)\n> \n\nIn fact, I kind of wish we'd also made committer == author.\n\nSince this whole thing is an import from another revision control \nsystem, one really wants that. It's one of those very rare situations \nin which fudging the commit date is not only fully legitimate, but darn \nnear required.\n\n\t-hpa\n\n"},{"id":"3653","messageId":"Pine.LNX.4.58.0505201604460.2206@ppc970.osdl.org","threadId":"565","inReplyTo":"1116626652.12975.118.camel@dhcp-188","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-20T23:25:10Z","receivedAt":"2005-05-20T23:25:10Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Sat, 21 May 2005, Kay Sievers wrote:\n> \n> These scripts are just a quick hack, I just wanted to know how nice the\n> stupid cvs file history can be converted to git-committs.\n\nUgh, indeed.\n\nIs it a cvsps bug or what that causes you to have to re-order the patches? \nOr is it that you don't handle branches or something in CVS? The fact that\nyou also remove one of the tags \"suppress ash-branch\" in that same number\nsequence that you had to fix up by re-ordering seems to imply that the\nbreakage has something to do with branching.\n\nDoes anybody have any suggestions for a nice and smallish CVS project that\nhas branches that I should look at?\n\n\t\tLinus\n"},{"id":"3661","messageId":"Pine.LNX.4.58.0505201702170.2206@ppc970.osdl.org","threadId":"565","inReplyTo":"428E745C.30304@zytor.com","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-21T00:50:25Z","receivedAt":"2005-05-21T00:50:25Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n[ Thomas added to cc, since he seems to have also worked on this ]\n\nOn Fri, 20 May 2005, H. Peter Anvin wrote:\n> \n> Here is my \"main\" OSS CVS repository; look at the syslinux module. It \n> has at least some minor branching.\n\nOk, \"cvsps\" output scares me. I wonder what\n\n\tWARNING: Invalid PatchSet 775, Tag syslinux-2_12-pre7:\n\t memdisk/init32.asm:1.3=after, memdisk/Makefile:1.26=before. Treated as 'before'\n\tWARNING: Invalid PatchSet 775, Tag syslinux-2_12-pre7:\n\t memdisk/init32.asm:1.3=after, memdisk/e820test.c:1.7=before. Treated as 'before'\n\t...\n\nmeans..\n\nAlso, your syslinux repo is interesting and shows another thing: doing a\n\n\tcvsps -g -p separate\n\nends badly with\n\n\tDirecting PatchSet 938 to file separate/938.patch\n\tcvs rdiff: failed to read diff file header /tmp/cvso8PswZ for mdiskchk.com,v: end of file\n\tsystem command returned non-zero exit status: 1: aborting\n\nwhich doesn't look very promising and causes an empty diff for\nmdiskck.com. Trying with --cvs-direct shows the reason:\n\n\tIndex: syslinux/sample/mdiskchk.com\n\t===================================================================\n\tRCS file: \n\t/home/torvalds/src/osscvs/cvsroot/syslinux/sample/mdiskchk.com,v\n\tretrieving revision 1.1\n\tretrieving revision 1.2\n\tdiff -u -r1.1 -r1.2\n\tBinary files /tmp/cvsU6MGU0 and /tmp/cvsiskFVR differ\n\nwhich shows that anything that bases itself of diffs (ie uses \"-g\" with\ncvsps) is just doomed to failure, since there's no good way to handle\nbinary data. Both Kay's and Thomas' scripts try to do the \"-g\" thing, \nthat's just not right.\n\nSo the cvs->git thing would need to be based on the actual objects, which \nobviously fits git quite well, but I was really hoping to have cvsps give \nsome nice intermediate format..\n\nSo it looks like we should avoid the diff format, and instead use\n\n\tcvsps -p separate\n\nand then just parse the \"Members\" thing and turning each of them either\ninto a \"delete\" (for ->.*DEAD) or \"cvs checkout -rxxx\" (for \".*->xxx\").\n\nHandling branches by literally treating them as different heads in git \nsounds quite simple, and indeed it looks like the basic logic for cvs->git \ntranslation would be\n\n\tfor-each-patch-from-cvsps\n\tdo\n\t\tgit-read-tree -m branchname-from-patch\n\t\tgit-update-cache -f -u -q -a\n\t\tfor-each-member-in-patch\n\t\tdo\n\t\t\tif [ DEAD ]; then\n\t\t\t\trm member\n\t\t\t\tgit-update-cache --remove member\n\t\t\telse\n\t\t\t\tcvs co -rREV member\n\t\t\t\tgit-update-cache --add member\n\t\t\tfi\n\t\t\tcat commit-message-from-patch | \n\t\t\t\tgit-commit-tree $(git-write-tree) -p branchname-from-patch > .git/revs/heads/branchname-from-patch\n\t\tdone\n\tdone\n\nwhich looks like it should work, and handle binary files right.\n\nThere seems to be two questions:\n\n - what to do about branch creation (ie a branch name we haven't seen\n before): it looks like cvsps doesn't tell you what the _originating_\n branch was for a new branch (that may be my confusion - maybe you can't\n create branches off branches in CVS?)\n\n For syslinux, it looks like you can always base it on HEAD, or possibly \n just the previous patch (which looks like it is always HEAD). The above \n pseudo-script will actually do that automatically, simply by virtue of\n the \"git-read-tree -m\" at the top of the loop failing when the\n branchname doesn't exist yet.\n\n - whether to bother to create merge entries for when somebody tried to \n merge a branch back or forth in CVS. \n\n CVS fundamentally doesn't have the notion of such a thing, and cvsps \n can't either. But we could try to guess, based on the commit message, \n perhaps.\n\n NOTE! Such a \"merge\" would not have any real GIT merge functionality \n what-so-ever. It would just introduce a second parent into the commit, \n nothing more.\n\nBah. What crud.\n\n\t\tLinus\n"},{"id":"3684","messageId":"pan.2005.05.21.07.29.14.58375@smurf.noris.de","threadId":"565","inReplyTo":"1116611932.12975.22.camel@dhcp-188","subject":"Re: gitweb wishlist","fromName":"Matthias Urlichs","fromEmail":"smurf@smurf.noris.de","sentAt":"2005-05-21T07:29:16Z","receivedAt":"2005-05-21T07:29:16Z","isPatch":false,"sender":{"key":"matthias@urlichs.de","avatar":"https://gravatar.com/avatar/2708905af227313eba6f2b2ae0f7d0259b5ac5d71baef58fe5a13c699ce0bbf0?d=mp&s=160"},"body":"Hi, Kay Sievers wrote:\n\n> Something like that: :)\n\nCool.\n\nMore feature requests: ;-)\n- Alternate white and almost-white backgrounds in the lists (all of them ;-)\n so that wide-screened people like me don't lose context when their eyes\n travel the long road from left to right edge of the screen. ;-)\n\n- Merges currently don't have diff links. It'd be nice to have one for\n each parent.\n\n- File diffs have the \"diff\" link on the *parent*, not on the child.\n That's counter-intuitive -- if I want to see what the Foo patch changes,\n I should be able to click on the \"diff\" link on _that_ line, not the one\n below it. Example:\nhttp://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=history;h=9636273dae265b9354b861b373cd43cd76a6d0fe;f=MAINTAINERS\n\n-- \nMatthias Urlichs | {M:U} IT Design @ m-u-it.de | smurf@smurf.noris.de\n\n\n"},{"id":"3685","messageId":"pan.2005.05.21.07.35.29.758035@smurf.noris.de","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505201702170.2206@ppc970.osdl.org","subject":"cvs->git (was Re: gitweb wishlist)","fromName":"Matthias Urlichs","fromEmail":"smurf@smurf.noris.de","sentAt":"2005-05-21T07:35:30Z","receivedAt":"2005-05-21T07:35:30Z","isPatch":false,"sender":{"key":"matthias@urlichs.de","avatar":"https://gravatar.com/avatar/2708905af227313eba6f2b2ae0f7d0259b5ac5d71baef58fe5a13c699ce0bbf0?d=mp&s=160"},"body":"Hi, Linus Torvalds wrote:\n\n> Bah. What crud.\n\nI have an old CVS->BK merge script lying around which does all of this\nreasonably correctly. If somebody wants to git-ify it, be my guest\n(I won't have time in the foreseeable futurre :-( ); it's in the\nrsync://server.smurf.noris.de/sourcemgr.git/#main repository as\nbin/b.cvs{,.pl}.\n\n-- \nMatthias Urlichs | {M:U} IT Design @ m-u-it.de | smurf@smurf.noris.de\n\n\n"},{"id":"3699","messageId":"1116695650.19541.12.camel@dhcp-188.off.vrfy.org","threadId":"565","inReplyTo":"pan.2005.05.21.07.29.14.58375@smurf.noris.de","subject":"Re: gitweb wishlist","fromName":"Kay Sievers","fromEmail":"kay.sievers@vrfy.org","sentAt":"2005-05-21T17:14:09Z","receivedAt":"2005-05-21T17:14:09Z","isPatch":false,"sender":{"key":"kay.sievers@vrfy.org","avatar":null},"body":"On Sat, 2005-05-21 at 09:29 +0200, Matthias Urlichs wrote:\n> Hi, Kay Sievers wrote:\n> \n> > Something like that: :)\n> \n> Cool.\n> \n> More feature requests: ;-)\n> - Alternate white and almost-white backgrounds in the lists (all of them ;-)\n> so that wide-screened people like me don't lose context when their eyes\n> travel the long road from left to right edge of the screen. ;-)\n\nDone!\n\n> - Merges currently don't have diff links. It'd be nice to have one for\n> each parent.\n\nDone! But I'm not sure if that is really useful. It may create a very\nvery big diff. :)\n\n> - File diffs have the \"diff\" link on the *parent*, not on the child.\n> That's counter-intuitive -- if I want to see what the Foo patch changes,\n> I should be able to click on the \"diff\" link on _that_ line, not the one\n> below it. Example:\n> http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=history;h=9636273dae265b9354b861b373cd43cd76a6d0fe;f=MAINTAINERS\n\nNo, the \"diff\" link is a diff against the current commit not an\nincremental one from revision to revision. Me may change that, I'm not\nsure what's the best here.\n\nKay\n\n"},{"id":"3839","messageId":"4292A08A.5050108@cobite.com","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505201702170.2206@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"David Mansfield","fromEmail":"david@cobite.com","sentAt":"2005-05-24T03:33:30Z","receivedAt":"2005-05-24T03:33:30Z","isPatch":false,"sender":{"key":"david@cobite.com","avatar":null},"body":"Hi Linus,\n\nLinus Torvalds wrote:\n> [ Thomas added to cc, since he seems to have also worked on this ]\n> \n> On Fri, 20 May 2005, H. Peter Anvin wrote:\n> \n>>Here is my \"main\" OSS CVS repository; look at the syslinux module. It \n>>has at least some minor branching.\n> \n> \n> Ok, \"cvsps\" output scares me. I wonder what\n> \n> \tWARNING: Invalid PatchSet 775, Tag syslinux-2_12-pre7:\n> \t memdisk/init32.asm:1.3=after, memdisk/Makefile:1.26=before. Treated as 'before'\n> \tWARNING: Invalid PatchSet 775, Tag syslinux-2_12-pre7:\n> \t memdisk/init32.asm:1.3=after, memdisk/e820test.c:1.7=before. Treated as 'before'\n> \t...\n> \n> means..\n> \n\nOk. I'll tell you. It means that the committer uses bad practices in \ntagging ;-) It generally means that force tag (cvs tag -F ) was \nused on a specific file. Here's the scenario:\n\ncvsps is trying to associate a tag to a specific commit. But in the cvs \nworld this is not always at all possible. If, for example, a commit \nmade and all files are tagged. Now some random file is modified and \ncommitted. Then, a bug is found in a file from the previously tagged \nset, say the file 'memdisk/init32.asm'. The bug is fixed, committed and \nthe tag is MOVED for _just that file_ forward to the new version. Now \nthere is no commit that can be associated with the tag. In this case, \ncvsps believes this to be a 'FUNKY' tag. There is a more pathological \ncase having to do with 'INVALID' tags... It's enough to make a grown \nman cry.\n\n> Also, your syslinux repo is interesting and shows another thing: doing a\n> \n> \tcvsps -g -p separate\n> \n> ends badly with\n> \n> \tDirecting PatchSet 938 to file separate/938.patch\n> \tcvs rdiff: failed to read diff file header /tmp/cvso8PswZ for mdiskchk.com,v: end of file\n> \tsystem command returned non-zero exit status: 1: aborting\n> \n> which doesn't look very promising and causes an empty diff for\n> mdiskck.com. Trying with --cvs-direct shows the reason:\n> \n> \tIndex: syslinux/sample/mdiskchk.com\n> \t===================================================================\n> \tRCS file: \n> \t/home/torvalds/src/osscvs/cvsroot/syslinux/sample/mdiskchk.com,v\n> \tretrieving revision 1.1\n> \tretrieving revision 1.2\n> \tdiff -u -r1.1 -r1.2\n> \tBinary files /tmp/cvsU6MGU0 and /tmp/cvsiskFVR differ\n> \n> which shows that anything that bases itself of diffs (ie uses \"-g\" with\n> cvsps) is just doomed to failure, since there's no good way to handle\n> binary data. Both Kay's and Thomas' scripts try to do the \"-g\" thing, \n> that's just not right.\n> \n\nI accept patches ;-) Honestly, handling binary data should be trivial I \njust haven't had the interest, and surprisingly noone else on the \ninternet ever has. The only binary file in the kernel appears to be the \nlogo.gif, according to Ingo.\n\n[ discussion on working around broken handling of binary files in cvsps]\n> \n> There seems to be two questions:\n> \n> - what to do about branch creation (ie a branch name we haven't seen\n> before): it looks like cvsps doesn't tell you what the _originating_\n> branch was for a new branch (that may be my confusion - maybe you can't\n> create branches off branches in CVS?)\n> \n> For syslinux, it looks like you can always base it on HEAD, or possibly \n> just the previous patch (which looks like it is always HEAD). The above \n> pseudo-script will actually do that automatically, simply by virtue of\n> the \"git-read-tree -m\" at the top of the loop failing when the\n> branchname doesn't exist yet.\n> \n\nSee attached patch to cvsps.c which displays 'Ancestor branch' when this \ndiffers from Branch.\n\n> - whether to bother to create merge entries for when somebody tried to \n> merge a branch back or forth in CVS. \n> \n> CVS fundamentally doesn't have the notion of such a thing, and cvsps \n> can't either. But we could try to guess, based on the commit message, \n> perhaps.\n> \n> NOTE! Such a \"merge\" would not have any real GIT merge functionality \n> what-so-ever. It would just introduce a second parent into the commit, \n> nothing more.\n> \n> Bah. What crud.\n> \n\nHey, a polished turd is only so shiny... cvsps is a 99% solution [to \nthe problem of extracting metatdata from cvs] only and cvs makes the \nother 1% impossible.\n\nDavid\n\n\n--- cvsps.c~\t2003-04-11 10:06:01.000000000 -0400\n+++ cvsps.c\t2005-05-23 23:26:12.110231536 -0400\n@@ -1402,6 +1402,16 @@\n \t tm->tm_hour, tm->tm_min, tm->tm_sec);\n printf(\"Author: %s\\n\", ps->author);\n printf(\"Branch: %s\\n\", ps->branch);\n+ \n+ /* check if ancestor was different branch */\n+ if (!list_empty(&ps->members)) \n+ {\n+\t PatchSetMember * psm = list_entry(ps->members.next, PatchSetMember, link);\n+\t const char * abr = psm->pre_rev ? psm->pre_rev->branch : NULL;\n+\t if (abr && strcmp(ps->branch, abr) != 0)\n+\t\t printf(\"Ancestor branch: %s\\n\", abr);\n+ }\n+\n printf(\"Tag: %s %s\\n\", ps->tag ? ps->tag : \"(none)\", tag_flag_descr[ps->tag_flags]);\n printf(\"Log:\\n%s\\n\", ps->descr);\n printf(\"Members: \\n\");\n"},{"id":"3840","messageId":"4292A1F2.7020606@zytor.com","threadId":"565","inReplyTo":"4292A08A.5050108@cobite.com","subject":"Re: gitweb wishlist","fromName":"H. Peter Anvin","fromEmail":"hpa@zytor.com","sentAt":"2005-05-24T03:39:30Z","receivedAt":"2005-05-24T03:39:30Z","isPatch":false,"sender":{"key":"hpa@zytor.com","avatar":null},"body":"David Mansfield wrote:\n> \n> Ok. I'll tell you. It means that the committer uses bad practices in \n> tagging ;-) It generally means that force tag (cvs tag -F ) was \n> used on a specific file. Here's the scenario:\n> \n> cvsps is trying to associate a tag to a specific commit. But in the cvs \n> world this is not always at all possible. If, for example, a commit \n> made and all files are tagged. Now some random file is modified and \n> committed. Then, a bug is found in a file from the previously tagged \n> set, say the file 'memdisk/init32.asm'. The bug is fixed, committed and \n> the tag is MOVED for _just that file_ forward to the new version. Now \n> there is no commit that can be associated with the tag. In this case, \n> cvsps believes this to be a 'FUNKY' tag. There is a more pathological \n> case having to do with 'INVALID' tags... It's enough to make a grown \n> man cry.\n> \n\nThis is only pathological if the tag now represents a state that never \nactually existed in the history of the repository. I don't believe \nthere are any such cases in the syslinux repository; I could be wrong, \nbut I am *highly* sceptical.\n\n> \n> I accept patches ;-) Honestly, handling binary data should be trivial I \n> just haven't had the interest, and surprisingly noone else on the \n> internet ever has. The only binary file in the kernel appears to be the \n> logo.gif, according to Ingo.\n> \n> [ discussion on working around broken handling of binary files in cvsps]\n> \n\nActually, as long as we can create the tree that exists between each \nchangeset, we should be OK.\n\n> \n> Hey, a polished turd is only so shiny... cvsps is a 99% solution [to \n> the problem of extracting metatdata from cvs] only and cvs makes the \n> other 1% impossible.\n> \n\nNo sh*t...\n\n\t-hpa\n"},{"id":"3841","messageId":"Pine.LNX.4.58.0505232048190.2307@ppc970.osdl.org","threadId":"565","inReplyTo":"4292A08A.5050108@cobite.com","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-24T03:52:26Z","receivedAt":"2005-05-24T03:52:26Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Mon, 23 May 2005, David Mansfield wrote:\n> > \n> > Bah. What crud.\n> > \n> \n> Hey, a polished turd is only so shiny... cvsps is a 99% solution [to \n> the problem of extracting metatdata from cvs] only and cvs makes the \n> other 1% impossible.\n\nThe \"what crud\" refers to cvs. cvsps seems to be a great way to make a\ntool to migrate away from CVS (or if forced to use CVS, at least show it\nin a sane manner). So don't take it the wrong way.\n\nI've gotten side-tracked with purely git issues, and since I don't \nactually have any CVS archives, the cvs->git translation will be on the \nback-burner for a while, but your \"Ancestor branch\" patch seems to at \nleast solve the problem that cvsps didn't show all the information that \nwas there. So now I know how to do branches, even if I don't think I'd \never _really_ merge them back (which is as much info as CVS contains). \n\nThey'd just be dangling references, ie you could get to them if you wanted \nto, for historical reasons, and they could be merged merged by hand, of \ncourse. Some day..\n\n\t\tLinus\n"},{"id":"3842","messageId":"4292AD5E.3000106@cobite.com","threadId":"565","inReplyTo":"4292A1F2.7020606@zytor.com","subject":"Re: gitweb wishlist","fromName":"David Mansfield","fromEmail":"david@cobite.com","sentAt":"2005-05-24T04:28:14Z","receivedAt":"2005-05-24T04:28:14Z","isPatch":false,"sender":{"key":"david@cobite.com","avatar":null},"body":"H. Peter Anvin wrote:\n> David Mansfield wrote:\n> \n>>\n>> Ok. I'll tell you. It means that the committer uses bad practices in \n>> tagging ;-) It generally means that force tag (cvs tag -F ) was \n>> used on a specific file. Here's the scenario:\n>>\n>> cvsps is trying to associate a tag to a specific commit. But in the \n>> cvs world this is not always at all possible. If, for example, a \n>> commit made and all files are tagged. Now some random file is \n>> modified and committed. Then, a bug is found in a file from the \n>> previously tagged set, say the file 'memdisk/init32.asm'. The bug is \n>> fixed, committed and the tag is MOVED for _just that file_ forward to \n>> the new version. Now there is no commit that can be associated with \n>> the tag. In this case, cvsps believes this to be a 'FUNKY' tag. \n>> There is a more pathological case having to do with 'INVALID' tags... \n>> It's enough to make a grown man cry.\n>>\n> \n> This is only pathological if the tag now represents a state that never \n> actually existed in the history of the repository. I don't believe \n> there are any such cases in the syslinux repository; I could be wrong, \n> but I am *highly* sceptical.\n> \n\nI didn't mean that YOUR repository had more pathological stuff in it, \njust that SOME do. 'FUNKY' tags are not really that bad, it's just that \nthere is not a single commit to assign them to (i.e. at no point were \nall of the objects in the repository at that state simultaneously), \nwhich makes the import of such a tag difficult into a more commit \noriented system.\n\nAnother way to reach 'funky'ness is to modify a file, commit and tag, \nwithout having done a 'cvs update' first (and a colleague has done a \ncommit since your last 'cvs update')\n\nDavid\n\n\n\n"},{"id":"3843","messageId":"20050524045840.GI12141@cip.informatik.uni-erlangen.de","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505201702170.2206@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Thomas Glanzmann","fromEmail":"sithglan@stud.uni-erlangen.de","sentAt":"2005-05-24T04:58:40Z","receivedAt":"2005-05-24T04:58:40Z","isPatch":false,"sender":{"key":"sithglan@stud.uni-erlangen.de","avatar":null},"body":"Hello,\n\n> \tWARNING: Invalid PatchSet 775, Tag syslinux-2_12-pre7:\n> \t memdisk/init32.asm:1.3=after, memdisk/Makefile:1.26=before. Treated as 'before'\n> \tWARNING: Invalid PatchSet 775, Tag syslinux-2_12-pre7:\n> \t memdisk/init32.asm:1.3=after, memdisk/e820test.c:1.7=before. Treated as 'before'\n> \t...\n\nactually I think this is the broken upstream version. It can't parse\ndates right. Just look at the exported patches and see if them all from\n1970. However the debian package has a patch in which solves it:\n\nmaybe you should try with the attached patch or with the version that\ncomes with debian sarge. I also reported this problem a while back to\nthe original author.\n\n\tThomas\n\n\n--- cvsps-2.0rc1.orig/util.c\n+++ cvsps-2.0rc1/util.c\n@@ -13,6 +13,7 @@\n #include \n #include \n #include \n+#include \n #include \n #include \n #include \n@@ -140,24 +141,51 @@\n return *res;\n }\n \n+static int get_int_substr(const char * str, const regmatch_t * p)\n+{\n+ char buff[256];\n+ memcpy(buff, str + p->rm_so, p->rm_eo - p->rm_so);\n+ buff[p->rm_eo - p->rm_so] = 0;\n+ return atoi(buff);\n+}\n+\n void convert_date(time_t * t, const char * dte)\n {\n- /* HACK: this routine parses two formats,\n- * 1) 'cvslog' format YYYY/MM/DD HH:MM:SS\n- * 2) time_t formatted as %d\n- */\n- \n- if (strchr(dte, '/'))\n+ static regex_t date_re;\n+ static int init_re;\n+\n+#define MAX_MATCH 16\n+ size_t nmatch = MAX_MATCH;\n+ regmatch_t match[MAX_MATCH];\n+\n+ if (!init_re) \n+ {\n+\tif (regcomp(&date_re, \"([0-9]{4})[-/]([0-9]{2})[-/]([0-9]{2}) ([0-9]{2}):([0-9]{2}):([0-9]{2})\", REG_EXTENDED)) \n+\t{\n+\t fprintf(stderr, \"FATAL: date regex compilation error\\n\");\n+\t exit(1);\n+\t}\n+\tinit_re = 1;\n+ }\n+ \n+ if (regexec(&date_re, dte, nmatch, match, 0) == 0)\n {\n+\tregmatch_t * pm = match;\n \tstruct tm tm;\n+\n+\t/* first regmatch_t is match location of entire re */\n+\tpm++;\n \t\n-\tmemset(&tm, 0, sizeof(tm));\n-\tsscanf(dte, \"%d/%d/%d %d:%d:%d\", \n-\t &tm.tm_year, &tm.tm_mon, &tm.tm_mday, \n-\t &tm.tm_hour, &tm.tm_min, &tm.tm_sec);\n-\t\n+\ttm.tm_year = get_int_substr(dte, pm++);\n+\ttm.tm_mon = get_int_substr(dte, pm++);\n+\ttm.tm_mday = get_int_substr(dte, pm++);\n+\ttm.tm_hour = get_int_substr(dte, pm++);\n+\ttm.tm_min = get_int_substr(dte, pm++);\n+\ttm.tm_sec = get_int_substr(dte, pm++);\n+\n \ttm.tm_year -= 1900;\n \ttm.tm_mon--;\n+\ttm.tm_isdst = 0;\n \t\n \t*t = mktime(&tm);\n }\n"},{"id":"3844","messageId":"4292B5E3.2080505@zytor.com","threadId":"565","inReplyTo":"4292AD5E.3000106@cobite.com","subject":"Re: gitweb wishlist","fromName":"H. Peter Anvin","fromEmail":"hpa@zytor.com","sentAt":"2005-05-24T05:04:35Z","receivedAt":"2005-05-24T05:04:35Z","isPatch":false,"sender":{"key":"hpa@zytor.com","avatar":null},"body":"David Mansfield wrote:\n>>\n>> This is only pathological if the tag now represents a state that never \n>> actually existed in the history of the repository. I don't believe \n>> there are any such cases in the syslinux repository; I could be wrong, \n>> but I am *highly* sceptical.\n> \n> I didn't mean that YOUR repository had more pathological stuff in it, \n> just that SOME do. 'FUNKY' tags are not really that bad, it's just that \n> there is not a single commit to assign them to (i.e. at no point were \n> all of the objects in the repository at that state simultaneously), \n> which makes the import of such a tag difficult into a more commit \n> oriented system.\n> \n> Another way to reach 'funky'ness is to modify a file, commit and tag, \n> without having done a 'cvs update' first (and a colleague has done a \n> commit since your last 'cvs update')\n> \n\nNot sure, sounds more likely.\n\nEither which way, I guess there are two ways to deal with them in 'git'; \neither as standalone trees (tags pointing to tree objects), or probably \nmore sensical, as impromptu branches if one can find a sane origin object.\n\n\t-hpa\n"},{"id":"3850","messageId":"Pine.LNX.4.58.0505240110580.2307@ppc970.osdl.org","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505232048190.2307@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-24T08:25:19Z","receivedAt":"2005-05-24T08:25:19Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Mon, 23 May 2005, Linus Torvalds wrote:\n> \n> I've gotten side-tracked with purely git issues, and since I don't \n> actually have any CVS archives, the cvs->git translation will be on the \n> back-burner for a while, but your \"Ancestor branch\" patch seems to at \n> least solve the problem that cvsps didn't show all the information that \n> was there.\n\nNaff.\n\nI just checked in a \"cvs2git.c\" file in the \"tools\" project (which has my \npatch application stuff).\n\nIt's still buggy, and it's hacky as hell, but you can basically do \nsomething like this:\n\n\tcvsps | cvs2git > script\n\nwith the normal setup for \"cvsps\", and \"cvs2git\" needs one additional\nstage, namely it wants to know the RCSDIR where to find the RCS files\n(that should be basically \"$CVSROOT/module\").\n\nThat _script_ then creates a git archive. Very hacky. So after you've \nsuccessfully created the conversion script, check it to see that it looks \nsane, and then do\n\n\tsh script\n\nand the end result is a git'ified version of your CVS repo (and a \ncorrupted working directory, btw, so look out. It _shouldn't_ corrupt \nyour old CVS repo, though, so it should be ok).\n\nIt has the logic for branches, but it doesn't work, and I'm fed up enough\nwith CVS and RCS for the moment that I'm not going to work on it any more\ntonight. I don't know what stupid bug I have (I've had about a million of\nthem on this silly program), but it's at a point where I think others\nmight find it interesting, and it's probably/hopefully some really\nembarrassing typo or something and easily fixed.\n\nIt converted Peter's \"syslinux\" repository in a couple of minutes, \nresulting in 1038 commits (it _should_ have resulted in 1046 commits, \nthat's the branch thing afaik) and most of it looks sane:\n\n\tdiff-tree cfb715c827e19226a446d47c98a7460fd94633ff (from a809559323f1b370717e475dd252b24686f97727)\n\tAuthor: hpa \n\tDate: Thu May 19 22:30:50 2005 -0700\n\t \n\t gcc4 compilation fix\n\t\n\t\n\tdiff-tree a809559323f1b370717e475dd252b24686f97727 (from 4d65331b50a7b5ce858bb55a58f37b17ebc26c72)\n\tAuthor: hpa \n\tDate: Sun May 8 22:47:03 2005 -0700\n\t \n\t New Multiboot module; increase command line limit to 1023\n\t\n\t\n\tdiff-tree 4d65331b50a7b5ce858bb55a58f37b17ebc26c72 (from e88244753d528f695790adc96f0542d20dc33882)\n\tAuthor: hpa \n\tDate: Fri Apr 29 07:08:03 2005 -0700\n\t \n\t Don't clobber live registers, it's not nice\n\t\n\t\n\tdiff-tree e88244753d528f695790adc96f0542d20dc33882 (from a49e189e35d208648a0d0b52ff652a5f3f8a707e)\n\tAuthor: hpa \n\tDate: Fri Apr 29 07:05:52 2005 -0700\n\t \n\t Use the correct register\n\n\t...\n\t...\n\t...\n\n\tdiff-tree 350772d45425a85dae86ec721d6bd3fde5595d50 (from 47ee894e7821f50cb83ea14b08132337577b2a1e)\n\tAuthor: hpa \n\tDate: Sat Jan 31 13:24:35 1998 -0800\n\t \n\t Slightly less ugly Id tag.\n\t\n\t\n\tdiff-tree 47ee894e7821f50cb83ea14b08132337577b2a1e (from a8b52f1c31055049b276d14c67436d06dd7757aa)\n\tAuthor: hpa \n\tDate: Sat Jan 31 13:22:38 1998 -0800\n\t \n\t Added Id tags.\n\t\n\t\n\tdiff-tree b924672aadb2c3b7f3cac1aaf52fbb4a1ed86b8d (from root)\n\tAuthor: hpa \n\tDate: Sat Jan 31 13:16:05 1998 -0800\n\t \n\t Initial revision\n\nAnd btw, it's definitely cvsps that does all the heavy lifting here. \n\"cvs2git\" itself is 255 lines of horrid crud, and should have been \nwritten in perl, except I only do C..\n\n\t\tLinus\n"},{"id":"3854","messageId":"Pine.LNX.4.58.0505240849050.2307@ppc970.osdl.org","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505240110580.2307@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-24T16:00:57Z","receivedAt":"2005-05-24T16:00:57Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Tue, 24 May 2005, Linus Torvalds wrote:\n> \n> It has the logic for branches, but it doesn't work, and I'm fed up enough\n> with CVS and RCS for the moment that I'm not going to work on it any more\n> tonight.\n\nI'm back, and yes, it was a really stupid thing.\n\nHowever, David, I need more help deciphering \"cvsps\" output..\n\nFixing the branch handling shows that cvsps does some really strange\nthings with the newly added \"Ancestor grpah\". Here's one example:\n\n\t---------------------\n\tPatchSet 372 \n\tDate: 2002/02/03 21:37:50\n\tAuthor: hpa\n\tBranch: syslinux-1_6x-1\n\tAncestor branch: HEAD\n\tTag: syslinux-1_67 \n\tLog:\n\tNew mailing list information\n\t\n\tMembers: \n\t syslinux.doc:1.48->1.48.2.1 \n\t\n\t---------------------\n\tPatchSet 373 \n\tDate: 2002/02/11 23:08:47\n\tAuthor: hpa\n\tBranch: HEAD\n\tTag: (none) \n\tLog:\n\ttftpd32 needs version 2.11 or later.\n\t\n\tMembers: \n\t pxelinux.doc:1.28->1.29 \n\t\n\t---------------------\n\tPatchSet 374 \n\tDate: 2002/02/18 23:43:43\n\tAuthor: hpa\n\tBranch: syslinux-1_6x-1\n\tAncestor branch: HEAD\n\tTag: syslinux-1_6x-merge-2 \n\tLog:\n\tActually make the -o option work properly.\n\t\n\tMembers: \n\t syslinux.c:1.13->1.13.2.1 \n\t\n\t---------------------\n\nnote how both 372 _and_ 374 claim to have HEAD as their ancestor, and are \non the \"syslinux-1_6x-1\" branch. What's up with that? Right now this \ncauses my git archive to first create 372 as a branch off HEAD, and then \noverwrite that with 374, resulting in a dangling branch for 372 that \n_exists_, but it's not reachable any more, because the branch name that it \nused has been overwritten by the _new_ branch off HEAD.\n\nSide note: cvs2git is pretty robust since it doesn't rely on patches\nanywhere, so the head of the branch likely ends up being correct, if that\n\"syslinux.doc\" file has been modified anywhere else in the branch. So this\n_usually_ just results in (a) git-fsck-cache complaining about unreachable\ncommits and (b) possible history being hard to find.\n\nMaybe this cvs2git behaviour is the right thing to do, and what really\nhappened was that the changes described by PatchSet 372 aren't really\navailable any more even in CVS, unless you go back by date or something \nlike that.\n\nHowever, I suspect it's a cvsps bug in the \"ancestor branch\" thing. I\ncould work around it by just saying \"if I have already seen this branch,\nI'll ignore the ancestor information\".\n\nSo I'd like to know whether this is a cvsps issue or whether I actually\nended up doing the right thing and it really should be a dangling\nbranch-name that got re-used...\n\n(And if it's a cvsps issue, I'd obviously prefer to get a cvsps patch \ninstead of having a questionable workaround in cvs2git).\n\n\"Davi-Mansfieldobi, you're our only hope..\"\n\n\t\tLinus\n"},{"id":"3856","messageId":"42935322.6070902@cobite.com","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505240110580.2307@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"David Mansfield","fromEmail":"david@cobite.com","sentAt":"2005-05-24T16:15:30Z","receivedAt":"2005-05-24T16:15:30Z","isPatch":false,"sender":{"key":"david@cobite.com","avatar":null},"body":"Linus Torvalds wrote:\n> \n> On Mon, 23 May 2005, Linus Torvalds wrote:\n> \n>>I've gotten side-tracked with purely git issues, and since I don't \n>>actually have any CVS archives, the cvs->git translation will be on the \n>>back-burner for a while, but your \"Ancestor branch\" patch seems to at \n>>least solve the problem that cvsps didn't show all the information that \n>>was there.\n> \n> \n> Naff.\n> \n> I just checked in a \"cvs2git.c\" file in the \"tools\" project (which has my \n> patch application stuff).\n> \n> It's still buggy, and it's hacky as hell, but you can basically do \n> something like this:\n> \n> \tcvsps | cvs2git > script\n> \n> with the normal setup for \"cvsps\", and \"cvs2git\" needs one additional\n> stage, namely it wants to know the RCSDIR where to find the RCS files\n> (that should be basically \"$CVSROOT/module\").\n> \n> That _script_ then creates a git archive. Very hacky. So after you've \n> successfully created the conversion script, check it to see that it looks \n> sane, and then do\n> \n> \tsh script\n> \n> and the end result is a git'ified version of your CVS repo (and a \n> corrupted working directory, btw, so look out. It _shouldn't_ corrupt \n> your old CVS repo, though, so it should be ok).\n\nI'll take a look. One problem is that many folks use non-local cvs... \nnot sure if that will be an issue. I'll look to cleaning this up if \nnecessary.\n\n> \n> It has the logic for branches, but it doesn't work, and I'm fed up enough\n> with CVS and RCS for the moment that I'm not going to work on it any more\n> tonight. I don't know what stupid bug I have (I've had about a million of\n> them on this silly program), but it's at a point where I think others\n> might find it interesting, and it's probably/hopefully some really\n> embarrassing typo or something and easily fixed.\n> \n\nI actually found an issue with the 30-second ancestor branch patch I \nsent and I'm doing that one properly now. Once that's done I can look \nat the branch capture logic in cvs2git and see if anything pops out.\n\n\n> It converted Peter's \"syslinux\" repository in a couple of minutes, \n> resulting in 1038 commits (it _should_ have resulted in 1046 commits, \n> that's the branch thing afaik) and most of it looks sane:\n\nReally cool. What's 8 commits between friends?\n\nDavid\n"},{"id":"3855","messageId":"Pine.LNX.4.58.0505240911050.2307@ppc970.osdl.org","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505240849050.2307@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-24T16:16:25Z","receivedAt":"2005-05-24T16:16:25Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Tue, 24 May 2005, Linus Torvalds wrote:\n> \n> Fixing the branch handling shows that cvsps does some really strange\n> things with the newly added \"Ancestor grpah\". Here's one example:\n\nAhh, looking at cvsps source, I think I see what's going on. \n\nIt's deciding the \"previous branch\" by looking at what the previous branch \nfor the first individual file in the PatchSet was, which fails because in \nthis case, PatchSet 372 was changing \"syslinux.doc\", and Patchset 374 was \nchanging \"syslinux.c\", and thus the previous version of the individual \n_files_ were both in the HEAD branch.\n\nSo it does look like I should just ignore the \"Ancestor branch\" \ninformation if the new branch already existed.\n\nOf course, some semantics will never be translatable when trying to treat \nCVS as a sane system (ie treating CVS as if it was changeset-based is \nalways going to cause strange corner cases since it really is file-based), \nbut that should most likely give the best approximation of what a \nconversion should do.\n\n\t\tLinus\n"},{"id":"3857","messageId":"20050524161745.GA9537@cip.informatik.uni-erlangen.de","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505240110580.2307@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Thomas Glanzmann","fromEmail":"sithglan@stud.uni-erlangen.de","sentAt":"2005-05-24T16:17:45Z","receivedAt":"2005-05-24T16:17:45Z","isPatch":false,"sender":{"key":"sithglan@stud.uni-erlangen.de","avatar":null},"body":"Hello Linus,\nI tried cvs2git and have the following problem:\n\n\t---------------------\n\tPatchSet 26\n\tDate: 1998/06/20 03:53:44\n\tAuthor: roessler\n\tBranch: mutt-0-93\n\tAncestor branch: HEAD\n\tTag: (none)\n\tLog:\n\tdocumenting alias-path\n\n\tMembers:\n\t\tdoc/manual.sgml:1.2->1.2.2.1\n\n\t---------------------\n\nAnd your script does that:\n\n\texport GIT_COMMITTER_NAME=roessler\n\texport GIT_COMMITTER_EMAIL=roessler\n\texport GIT_AUTHOR_NAME=roessler\n\texport GIT_AUTHOR_EMAIL=roessler\n\texport GIT_AUTHOR_DATE='1998/06/20 03:53:44'\n\tln -sf refs/heads/'master' .git/HEAD\n\tgit-read-tree -m HEAD\n\tgit-checkout-cache -f -u -a\n\tmkdir -p doc\n\tco -p -r1.2.2.1 '/home/cip/adm/sithglan/work/mutt/cvsrepository/doc/Attic/manual.sgml,v' > 'doc/manual.sgml'\n\tgit-update-cache --add -- 'doc/manual.sgml'\n\ttree=$(git-write-tree)\n\tcat > .cmitmsg < .git/HEAD\n\nThe problem might be that this is the first commit in the branch. But I thought\nit should end up in refs/heads/mutt-0-93. The problem is that this ends\nup a empty file and next time the script is working on it, it fails\nbecause the branch is empty:\n\n\t+ export GIT_COMMITTER_NAME=roessler\n\t+ GIT_COMMITTER_NAME=roessler\n\t+ export GIT_COMMITTER_EMAIL=roessler\n\t+ GIT_COMMITTER_EMAIL=roessler\n\t+ export GIT_AUTHOR_NAME=roessler\n\t+ GIT_AUTHOR_NAME=roessler\n\t+ export GIT_AUTHOR_EMAIL=roessler\n\t+ GIT_AUTHOR_EMAIL=roessler\n\t+ export 'GIT_AUTHOR_DATE=1998/06/20 07:12:32'\n\t+ GIT_AUTHOR_DATE=1998/06/20 07:12:32\n\t+ ln -sf refs/heads/mutt-0-93 .git/HEAD\n\t+ git-read-tree -m HEAD\n\tusage: git-read-tree ( | -m [ ])\n\t+ git-checkout-cache -f -u -a\n\t+ co -p -r1.1.1.1.2.2 /home/cip/adm/sithglan/work/mutt/cvsrepository/handler.c,v\n\t/home/cip/adm/sithglan/work/mutt/cvsrepository/handler.c,v --> standard output\n\trevision 1.1.1.1.2.2\n\t+ git-update-cache --add -- handler.c\n\t++ git-write-tree\n\t+ tree=9e4d085838e4e62a8c4236a6713a7dd8d7b07b4e\n\t+ cat\n\t++ cat .cmitmsg\n\t++ git-commit-tree 9e4d085838e4e62a8c4236a6713a7dd8d7b07b4e -p HEAD\n\tusage: git-commit-tree [-p ]* < changelog\n\t+ commit=\n\t+ echo\n\n\tThomas\n"},{"id":"3858","messageId":"Pine.LNX.4.58.0505240929051.2307@ppc970.osdl.org","threadId":"565","inReplyTo":"20050524161745.GA9537@cip.informatik.uni-erlangen.de","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-24T16:31:49Z","receivedAt":"2005-05-24T16:31:49Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Tue, 24 May 2005, Thomas Glanzmann wrote:\n> \n> And your script does that:\n> \n> \texport GIT_COMMITTER_NAME=roessler\n> \texport GIT_COMMITTER_EMAIL=roessler\n> \texport GIT_AUTHOR_NAME=roessler\n> \texport GIT_AUTHOR_EMAIL=roessler\n> \texport GIT_AUTHOR_DATE='1998/06/20 03:53:44'\n> \tln -sf refs/heads/'master' .git/HEAD\n> \tgit-read-tree -m HEAD\n> \tgit-checkout-cache -f -u -a\n> \tmkdir -p doc\n> \tco -p -r1.2.2.1 '/home/cip/adm/sithglan/work/mutt/cvsrepository/doc/Attic/manual.sgml,v' > 'doc/manual.sgml'\n> \tgit-update-cache --add -- 'doc/manual.sgml'\n> \ttree=$(git-write-tree)\n> \tcat > .cmitmsg < \tdocumenting alias-path\n> \tEOFMSG\n> \tcommit=$(cat .cmitmsg | git-commit-tree $tree -p HEAD)\n> \techo $commit > .git/HEAD\n> \n> The problem might be that this is the first commit in the branch. But I thought\n> it should end up in refs/heads/mutt-0-93.\n\nYes, you're using the cvs2git from yesterday, which didn't write the new\ncommit to the right branch. This is part of the branch fixing I've done.\n\nWait another few minutes and I'll commit my fix the problem with cvsps\nbranch handling (and I need to escape '$' in < \n> Wait another few minutes and I'll commit my fix the problem with cvsps\n> branch handling (and I need to escape '$' in < ll .git/refs/heads/\n\ttotal 16K\n\t-rw-rw-r-- 1 torvalds torvalds 41 May 24 09:36 branch-1_xx\n\t-rw-rw-r-- 1 torvalds torvalds 41 May 24 09:37 master\n\t-rw-rw-r-- 1 torvalds torvalds 41 May 24 09:36 syslinux\n\t-rw-rw-r-- 1 torvalds torvalds 41 May 24 09:36 syslinux-1_6x-1\n\nand doing a \n\n\tgit-rev-tree branch-1_xx master syslinux syslinux-1_6x-1 | wc -l\n\nreports 1046 total revisions (which also matches cvsps exactly).\n\nSo things look ok, but I haven't actually checked the _contents_ of the\ntree, except to look that the pathces that \"git-whatchanged -p\" reports\nlook sane.\n\nThere's two remaining bad things:\n\n - name translation doesn't exist (so all of Peters changesets get \n reported as author \"hpa \")\n\n - the commit time will be the conversion time, not the original commit \n time (but the _author_ time will be correct). I suspect that for a \n conversion like this, we really should add support for GIT_COMMIT_DATE. \n\n That would also make the archive conversion 100% reproducible, ie\n everybody should get the exact same objects (and thus the exact same\n SHA1 values) which is good.\n\nI'll add the GIT_COMMIT_DATE thing, but the name translation is for \nsomebody else.\n\n\t\tLinus\n"},{"id":"3860","messageId":"42935F96.8030205@cobite.com","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505240849050.2307@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"David Mansfield","fromEmail":"david@cobite.com","sentAt":"2005-05-24T17:08:38Z","receivedAt":"2005-05-24T17:08:38Z","isPatch":false,"sender":{"key":"david@cobite.com","avatar":null},"body":"Linus Torvalds wrote:\n> \n> On Tue, 24 May 2005, Linus Torvalds wrote:\n> \n>>It has the logic for branches, but it doesn't work, and I'm fed up enough\n>>with CVS and RCS for the moment that I'm not going to work on it any more\n>>tonight.\n> \n> \n> I'm back, and yes, it was a really stupid thing.\n> \n> However, David, I need more help deciphering \"cvsps\" output..\n> \n> Fixing the branch handling shows that cvsps does some really strange\n> things with the newly added \"Ancestor grpah\". Here's one example:\n> \n\nYes. While not falling asleep last night I realized that the \nquick-and-dirty approach was bogus. I need to track what the ancestor \nis as I'm building up the data structure, not while outputting it. So \nI'm working on a correct version which puts ancestor_branch into the \nPatchSet structure itself.\n\nIt's completely done now except for that it segfaults instantly.\n\nBTW where did you get the cvsroot for syslinux? Could I get a copy \nsomewhere?\n\nDavid\n\n\n"},{"id":"3861","messageId":"Pine.LNX.4.58.0505241017510.2307@ppc970.osdl.org","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505240943080.2307@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-24T17:23:48Z","receivedAt":"2005-05-24T17:23:48Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Tue, 24 May 2005, Linus Torvalds wrote:\n> \n> - the commit time will be the conversion time, not the original commit \n> time (but the _author_ time will be correct). I suspect that for a \n> conversion like this, we really should add support for GIT_COMMIT_DATE. \n> \n> That would also make the archive conversion 100% reproducible, ie\n> everybody should get the exact same objects (and thus the exact same\n> SHA1 values) which is good.\n> \n> I'll add the GIT_COMMIT_DATE thing, but the name translation is for \n> somebody else.\n\nDone. I've also fixed the timezone to \"+0000\", so that it doesn't matter \nwhere you do the conversion, you should always get the same results \n(again, I just pushed that out, it might not have hit the public mirrors \nyet).\n\nTo get GIT_COMMITTER_DATE (note: COMMITTER, not COMMIT, to illogically \nmatch the name/email ones) you obviously also need a new git. So to have \nit all working right, you should have the top commits in git-tools and git \nbe\n\n cvs2git: set timezone info to UTC, the way CVS does\n\nand\n\n git-commit-tree: allow overriding of commit date\n\nrespectively.\n\nAnd if this doesn't work for you, point me to the CVS archive that causes \nyou trouble. \n\n\t\tLinus\n"},{"id":"3862","messageId":"Pine.LNX.4.58.0505241024450.2307@ppc970.osdl.org","threadId":"565","inReplyTo":"42935F96.8030205@cobite.com","subject":"Re: gitweb wishlist","fromName":"Linus Torvalds","fromEmail":"torvalds@osdl.org","sentAt":"2005-05-24T17:28:44Z","receivedAt":"2005-05-24T17:28:44Z","isPatch":false,"sender":{"key":"torvalds@linux-foundation.org","avatar":"https://avatars.githubusercontent.com/u/1024025?v=4"},"body":"\n\nOn Tue, 24 May 2005, David Mansfield wrote:\n> \n> Yes. While not falling asleep last night I realized that the \n> quick-and-dirty approach was bogus. I need to track what the ancestor \n> is as I'm building up the data structure, not while outputting it. \n\nYes.\n\n> It's completely done now except for that it segfaults instantly.\n\nVery good. Are you going to also make a new release at some point, so that \nwe don't have strange random patches floating around?\n\n> BTW where did you get the cvsroot for syslinux? Could I get a copy \n> somewhere?\n\nPeter sent it in private email, I don't know how public that is (it\nprobably is perfectly public and he just didn't want to spam the mailing\nlist or run afoul of size limits, but I just don't know for sure, so..),\nbut I bet he'll happily send it to you too.\n\n\t\tLinus\n"},{"id":"3863","messageId":"42937290.8030507@zytor.com","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505241024450.2307@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"H. Peter Anvin","fromEmail":"hpa@zytor.com","sentAt":"2005-05-24T18:29:36Z","receivedAt":"2005-05-24T18:29:36Z","isPatch":false,"sender":{"key":"hpa@zytor.com","avatar":null},"body":"Linus Torvalds wrote:\n> \n> Peter sent it in private email, I don't know how public that is (it\n> probably is perfectly public and he just didn't want to spam the mailing\n> list or run afoul of size limits, but I just don't know for sure, so..),\n> but I bet he'll happily send it to you too.\n> \n\nAlready sent... I haven't looked it over to make sure there isn't \nanything that shouldn't be in there yet, so if you need to distribute it \nplease give me a warning so I can look it over first.\n\n\t-hpa\n\n"},{"id":"3864","messageId":"20050524182951.GB9537@cip.informatik.uni-erlangen.de","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505240943080.2307@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Thomas Glanzmann","fromEmail":"sithglan@stud.uni-erlangen.de","sentAt":"2005-05-24T18:29:51Z","receivedAt":"2005-05-24T18:29:51Z","isPatch":false,"sender":{"key":"sithglan@stud.uni-erlangen.de","avatar":null},"body":"Hello,\n\n> - name translation doesn't exist (so all of Peters changesets get \n> reported as author \"hpa \")\n\nI pick that up.\n\n> - the commit time will be the conversion time, not the original commit \n> time (but the _author_ time will be correct). I suspect that for a \n> conversion like this, we really should add support for GIT_COMMIT_DATE. \n\nThanks. I wanted to send patches or at least ask you this for ages, but\nnever did. :-)\n\n\tThomas\n"},{"id":"3865","messageId":"20050524184612.GA23637@cip.informatik.uni-erlangen.de","threadId":"565","inReplyTo":"Pine.LNX.4.58.0505241017510.2307@ppc970.osdl.org","subject":"Re: gitweb wishlist","fromName":"Thomas Glanzmann","fromEmail":"sithglan@stud.uni-erlangen.de","sentAt":"2005-05-24T18:46:12Z","receivedAt":"2005-05-24T18:46:12Z","isPatch":false,"sender":{"key":"sithglan@stud.uni-erlangen.de","avatar":null},"body":"Hello,\n\n> And if this doesn't work for you, point me to the CVS archive that causes \n> you trouble.\n\nyou should try the mutt cvs repository[1].\n\nI have the following issues all seem easy to fix:\n\n\t- PatchSet 1 depends on PatchSet 2 (but cvsps gets the ordering wrong;\n\t should be easy fixable) (I just swichted the two before\n\t running cvs2git)\n\n\t- Some Shell escapes (I didn't looked into them yet)\n\n\t\t(faui02new) [/var/tmp/sithglan/mutt-cvs] bash ~/work/cvsps/sane\n\t\tdefaulting to local storage area\n\t\tCommitting initial tree 7e68fd9a5104b61192a7da7357549d95b3a0620c\n\t\tIgnoring path .cvsignore\n\t\t...\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: line 1: ...: command not found\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: command substitution: line 1: unexpected EOF while looking for matching `''\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: command substitution: line 4: syntax error: unexpected end of file\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: command substitution: line 1: unexpected EOF while looking for matching `''\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: command substitution: line 5: syntax error: unexpected end of file\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: command substitution: line 1: unexpected EOF while looking for matching `''\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: command substitution: line 2: syntax error: unexpected end of file\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: command substitution: line 1: unexpected EOF while looking for matching `''\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: command substitution: line 3: syntax error: unexpected end of file\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: command substitution: line 1: unexpected EOF while looking for matching `''\n\t\t/home/cip/adm/sithglan/work/cvsps/sane: command substitution: line 26: syntax error: unexpected end of file\n\nBut hey this looks really good: :-))))))\n\n(faui02new) [/var/tmp/sithglan/mutt-cvs] git parent ~/work/mutt/git/mutt-cvs\n(faui02new) [/var/tmp/sithglan/mutt-cvs] git parentdiff\n(faui02new) [/var/tmp/sithglan/mutt-cvs]\n\nI think I will run my 'import patch by patch script again' and check the\nchangesets against the cvs2git tree, but it looks fine for me.\n\n\tThomas\n\n[1] To make it reproducable for you:\n\nI used the attached patch against cvsps-2.0rc1 which fixes date\ncovnersion problems and of course includes the ancestor thing.\n\nrsync -r rsync://cvs.gnupg.org/mutt-cvs-rep mutt-cvs-rep\n\n\ndiff --git a/cvs_direct.c b/cvs_direct.c\n--- a/cvs_direct.c\n+++ b/cvs_direct.c\n@@ -126,7 +126,7 @@ CvsServerCtx * open_cvs_server(char * p_\n \tsend_string(ctx, \"Root %s\\n\", ctx->root);\n \n \t/* this is taken from 1.11.1p1 trace - but with Mbinary removed. we can't handle it (yet!) */\n-\tsend_string(ctx, \"Valid-responses ok error Valid-requests Checked-in New-entry Checksum Copy-file Updated Created Update-existing Merged Patched Rcs-diff Mode Mod-time Removed Remove-entry Set-static-directory Clear-static-directory Set-sticky Clear-sticky Template Set-checkin-prog Set-update-prog Notified Module-expansion Wrapper-rcsOption M E F MT\\n\", ctx->root);\n+\tsend_string(ctx, \"Valid-responses ok error Valid-requests Checked-in New-entry Checksum Copy-file Updated Created Update-existing Merged Patched Rcs-diff Mode Mod-time Removed Remove-entry Set-static-directory Clear-static-directory Set-sticky Clear-sticky Template Set-checkin-prog Set-update-prog Notified Module-expansion Wrapper-rcsOption M E F\\n\", ctx->root);\n \n \tsend_string(ctx, \"valid-requests\\n\");\n \n@@ -894,6 +894,7 @@ char * cvs_rlog_fgets(char * buff, int b\n }\n else if (strcmp(lbuff, \"ok\") == 0 ||strcmp(lbuff, \"error\") == 0)\n {\n+\tdebug(DEBUG_TCP, \"cvs_direct: rlog: got command completion\");\n \treturn NULL;\n }\n \ndiff --git a/cvsps.1 b/cvsps.1\n--- a/cvsps.1\n+++ b/cvsps.1\n@@ -3,7 +3,7 @@\n CVSps \\- create patchset information from CVS\n .SH SYNOPSIS\n .B cvsps\n-[-h] [-x] [-u] [-z ] [-g] [-s ] [-a ] [-f ] [-d [-d ]] [-l ] [-b ] [-r [-r ]] [-p ] [-v] [-t] [--norc] [--summary-first] [--test-log ] [--bkcvs] [--no-rlog] [--diff-opts