{"thread":{"id":"16217","subject":"How it was at GitTogether'08 ?","startedAt":"2008-11-08T01:54:52Z","lastAt":"2008-11-11T23:26:24Z","messageCount":29,"participants":["Jakub Narebski","Johan Herland","David Symonds","Jeff King","Kai Blin","Steven Grimm","Robin Rosenberg","Shawn O. Pearce","Jean-Luc Herren","Tim Ansell","Johannes Schindelin","Junio C Hamano","Jonas Fonseca"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"95184","messageId":"200811080254.53202.jnareb@gmail.com","threadId":"16217","inReplyTo":null,"subject":"How it was at GitTogether'08 ?","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2008-11-08T01:54:52Z","receivedAt":"2008-11-08T01:54:52Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"GitTogether 2008, which took place October 27(Mon)-29(Wed), has ended more\nthan week ago.  Therefore I'd like to ask you impressions (while it is still\nfresh) from GitTogether, describe talks which are not described below,\ncorrect wrong information in below, etc.\n\nPlease reply to this email while GitTogether is on front of your mind...\n\n\nTalks at GitTogether 2008\n=========================\n\nMon, Oct 27, 2008\n-----------------\n* Dscho: Contributing with Git\n  http://www.youtube.com/watch?v=j45cs5_nY2k\n\n* Junio: Git Chronicle\n\n  blog: Junio went though a sort of statistical history of the Git project\n  that was fascinating (turns out there are still about 220 lines of code\n  still around from Linus original first commit).\n\n  http://userweb.kernel.org/~junio/200810-Chron.pdf\n\n* Jeff: Helping new contributors join\n\n* Petr: Renames Again and Again and Again\n\n  IRC: detection of wholesame renames of directories (WIP) and '--follow'\n  limitation were mentioned, but outcome is unclear; pasky plans to hack\n  together some patch implementing explicit renames hinting\n\n* Sam: GitTorrent\n\n  IRC: briefly the history if GitTorrent: way underannounced, not completely\n  finished, GTP/0.1 was \"published\" long ago. There was a GSoC project this\n  year, mentored by mugwump. Unfortunately, the student turned out to be a\n  procrastinator or something, and this project was mainly pushed forward by\n  the mentor. Now after the summer, mugwump took a step back and analyzed\n  the protocol with all the experience won from implementing a large part of\n  it. Turns out that a lot of the stuff was not really necessary, but copied\n  because the BitTorrent protocol had it all.  So, plan is to really strip\n  down GTP.  And not have a separate protocol for GitTorrent, but rather\n  have it as part of git-daemon. [...] Turns out that warthog9 had a few\n  really important things to say that should end up (at least in the 2nd\n  version) in having a more robust discovery/exchange.\n\n  http://utsl.gen.nz/talks/gittorrent/start.html\n\n* Tom: GitHub\n\n  IRC: a tour, some history, and insight into how it works; some nice\n  gimmicks, such as \"gist\" (a git backed pastebin), or like network graphs\n  (look graphically at forks of a repository).\n\n  blog: Tom got to demonstrate GitHub and Gist to the group, most of whom\n  are very command line oriented and had not used either before.\n\n* Shawn: JGit\n\n* Scott: Linkable Library\n\n  blog: got to talk about the need for a linkable git library\n\n  http://thread.gmane.org/gmane.comp.version-control.git/99608\n\n* Sam: perl.git\n\n  blog: Sam demonstrated the work he went through to import 20 years of Perl\n  history into the git repository that the Perl team is just now finishing\n  transitioning to from Perforce.\n\n  http://utsl.gen.nz/talks/perl-history/slides/\n\n\nTue, Oct 28, 2008\n-----------------\n* Shawn: Pack v4\n\n  IRC: Shawn gave a pretty good summary of packv4, and he said what are the\n  major blocks: we need a refactoring of the tree walker first, and then it\n  would add a pretty complex heap of code to make use of the packv4 format\n  for tree walking.  Main concern was _not_ the size of pack (for spearce at\n  least), but the speed of rev-list parsing with path delimiters [...] The\n  tree walking needs a different interface. (Path-delimiting must be a part\n  of it, not an afterthought. Mode filtering must be a part of it, not an\n  afterthought.)  Estimations for packv4: 10% decrease in size, 25% increase\n  in speed.  [...] reworking packv4 to do it for OBJ_BLOB too (and the\n  obvious path to speed-optimize git grep -w) has not been touched at all.\n\n* Jeff: What needs refactoring?\n\n* David Brown: Life with Git\n  http://www.davidb.org/git/git-corp.pdf\n\n* Lighting Round Talks\n  - Scott: iGitHub - git daemon and repository browser on the iPhone\n  - Jeff: Portability autobuilders\n  - Petr: Quick TopGit introduction, problems, future plans\n\n  IRC: iGitHub - its a tiny app you run on your iphone to make your iphone\n  act as a git server/client. since apple doesn't let you tether your laptop\n  through your iphone to get wireless internet you can instead push a\n  repository to the iphone via the laptop, then push from there to the\n  world, or fetch from the world to the iphone, then from the iphone to the\n  laptop. it also has a small history browser and stuff built in.\n\n* Jeffrey Altman: Discussion with OpenAFS\n\n  IRC: was about them trying to convert their franken-CVS repo to\n  Git. Basically, they use CVS, but have a layer on top of it that logically\n  combines two related commits into a single \"delta\", and they want to\n  preseve those as single git commits to the extent possible (sometimes it\n  is not possible because of conflicts). They are looking for ways to script\n  the conversion.\n\n  The OpenAFS guys were also interested in helping with libgit and windows\n  TortoiseGit work (separate conversation).\n\n* Tim: Git as a Media Repository\n  http://www.thousandparsec.net/~tim/media+git.pdf\n\n* Petr: Git UI\n\n  IRC: pasky talked a little bit about what he still misses in git.git CLI:\n  commit -m piping through fmt; the \"reset and checkout\" vs \"switch, reset\n  and restore\" paradigm; add vs stage; a bit about submodules [...] and\n  spent most of the time in a bikeshed argument about\n  add/rm/stage/unstage/staged/revert/restore/switch/. [...] Basically, reset\n  and checkout sucks and has chaotic behaviour with various switches and\n  argument combinations (files given vs not given etc.) So in ideal world,\n  reset and checkout might've been deprecated and instead, switch, restore\n  and unstage would be introduced... but that's not gonna happen.\n\n  blog: The other important, highly visual thing that was discussed, and\n  even a few patches are already in for, is for little improvements to the\n  UI. The full planning document is on Gist, but already things like making\n  use of the term 'stage' for things that happen in the index (such as using\n  \"git diff --staged\" instead of \"git diff --cached\") is being worked\n  on. I'm excited that staging files may soon be done via \"git stage\"\n  rather-than/in-addition-to \"git add\". This is nice for new users who often\n  have a hard time seeing why you have to keep \"git add\"ing to stage your\n  changes.\n\n  http://pasky.or.cz/~pasky/cp/gittogether-ui.sxi\n  http://gist.github.com/20818\n\n  http://thread.gmane.org/gmane.comp.version-control.git/99572\n  http://thread.gmane.org/gmane.comp.version-control.git/99541\n\n* Steven: Life with git-svn\n\n\nWed, Oct 29, 2008\n-----------------\n* Shawn: Bundle Project\n\n  IRC: Gerrit is issue tracker / code review tool, which takes full bundles\n  and moves them around, and only when it was reviewed, it is actually\n  applied.\n\n  http://android.git.kernel.org/?p=tools/gerrit.git\n\n* Discussion on notes\n\n* Petr: introduce the secret pickaxe project\n  http://repo.or.cz/w/giddy.git\n\n* Petr: Git in Pharma Corp\n  http://pasky.or.cz/~pasky/cp/gittogether-novartis.sxi\n  http://pasky.or.cz/g/gitweb.cgi?p=templatedemo.git;a=summary\n\n* Tim: Large media in Git (Repeat)\n\n  IRC: Large = 100M and more (e.g. audio or video assets for a game). the\n  idea was that you would have shallow _and_ lazy clones. There was not much\n  discussion about implementation details, though; except that there should\n  be a mode where blobs are _not_ fetched, and that every non-blob has to be\n  local.\n\n  IRC: Tim Ansell (mithro) of Thousand Parsec still uses SVN for this\n  because he can checkout just one file, edit it, and push it\n  back. Basically he wants to clone the metadata (commits and trees and\n  small blobs) and have large blobs (e.g. >25 MB or some limit) kept on the\n  origin, lazily fetched over a git transport on demand. Tim started to work\n  up a prototype using HTTP. folks thought the idea showed promise, but we\n  need to see code and the impacts it has and where it breaks down to really\n  understand if it is worth doing or not.\n\n  blog: Tim talked about something that I think will be one of the next huge\n  (highly visible) changes in Git you\\u2019re likely to see in the next year\n  \\u2013 handling large meda well, and being able to do narrow and sparse\n  clones, (and shallow clones better). This means being able to clone part\n  of a Git repository, such as just the last revision (shallow), just the\n  \\u2018lib\\u2019 directory (narrow) or just a single file\n  (sparse). Importantly, you would be able to see the history of everything\n  still (it would download the commit and tree objects, which are generally\n  small, but not the larger blobs), and you would be able to do pushes back\n  (which shallow clones can\\u2019t currently do).\n\n  http://www.thousandparsec.net/~tim/media+git.pdf  \n\n* JH: $$$ discussion\n\n  IRC: It was about the Google Summer of Code 2008 mentor money. It was\n  decided (given most of the mentors were there) to use it to help a few\n  folks cover their travel expenses to the GitTogether. Most folks actually\n  had their costs covered by their employer (yay for some big companies!)\n  but a few paid out of pocket, so we are helping them defray it a bit. and\n  the rest is being used to order and distribute some git t-shirts.\n\n* Sam: Git as a DB Backend\n\n  IRC: mugwump had this idea of running a sql sort of database in git, using\n  the git logs as a way to implement an ACID complaint database. to be\n  honest i was busy reading email or something and didn't pay much attention\n  to his talk, but folks seemed to think it was an abuse of git\n\n* Petr: GUI Even a Designer Uses\n\n  IRC: it was rather short and ambiguous; mostly arguing whether to and in\n  what language/toolkit to reimplement git-gui\n\n\nSources:\n========\n* http://git.or.cz/gitwiki/GitTogether\n* http://google-opensource.blogspot.com/2008/11/gittogether-08.html\n* http://github.com/blog/196-gittogether-2008\n* Calendar ID: amj5rbmljs66k9rtibs34kc2c8@group.calendar.google.com\n* http://colabti.org/irclogger/irclogger_log_search/git\n  #git channel on irc.freenode.org\n\n\n-- \nJakub Narebski\n"},{"id":"95185","messageId":"200811080441.25796.johan@herland.net","threadId":"16217","inReplyTo":"200811080254.53202.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Johan Herland","fromEmail":"johan@herland.net","sentAt":"2008-11-08T03:41:25Z","receivedAt":"2008-11-08T03:41:25Z","isPatch":false,"sender":{"key":"johan@herland.net","avatar":"https://avatars.githubusercontent.com/u/547031?v=4"},"body":"On Saturday 08 November 2008, Jakub Narebski wrote:\n> GitTogether 2008, which took place October 27(Mon)-29(Wed), has ended\n> more than week ago.  Therefore I'd like to ask you impressions (while it\n> is still fresh) from GitTogether, describe talks which are not described\n> below, correct wrong information in below, etc.\n\nThanks for providing this summary to those of us that didn't attend.\n\n> * Discussion on notes\n\nCan someone elaborate on this? AFAIK, notes have popped up on this list \noften enough that I'm convinced it would be a _really_ useful feature. The \nonly drawback I was aware of, was the lack of an efficient implementation, \nbut then Jeff comes out of the blue and posts some interesting numbers [1] \na week or so ago. Does this mean there are no remaining obstacles?\n\n[1]: http://article.gmane.org/gmane.comp.version-control.git/99415\n\n\nHave fun!\n\n...Johan\n\n-- \nJohan Herland, <johan@herland.net>\nwww.herland.net\n"},{"id":"95192","messageId":"ee77f5c20811072108o21f97c97i8174f4f7ecd67030@mail.gmail.com","threadId":"16217","inReplyTo":"200811080254.53202.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"David Symonds","fromEmail":"dsymonds@gmail.com","sentAt":"2008-11-08T05:08:27Z","receivedAt":"2008-11-08T05:08:27Z","isPatch":false,"sender":{"key":"dsymonds@gmail.com","avatar":"https://gravatar.com/avatar/b22f5051cbfc11836e36cf7a690e6cde4e225d835e13295ff98d15c7a9ee3c0f?d=mp&s=160"},"body":"On Fri, Nov 7, 2008 at 5:54 PM, Jakub Narebski <jnareb@gmail.com> wrote:\n\nHere's my thoughts on some of these talks.\n\n> Mon, Oct 27, 2008\n> -----------------\n> * Dscho: Contributing with Git\n>  http://www.youtube.com/watch?v=j45cs5_nY2k\n\nIt was a good intro, but I was expecting a few more non-GitTogether\npeople. We had quite a large room, but there was only about a dozen\nother people who came along. I don't know whether that was the fault\nof the timing, lack of advertising, or a lack of interest.\n\n> * Junio: Git Chronicle\n>\n>  blog: Junio went though a sort of statistical history of the Git project\n>  that was fascinating (turns out there are still about 220 lines of code\n>  still around from Linus original first commit).\n\nThis was really interesting. It would be great to put this on a\ngeneral web page instead of in a PDF.\n\n> * Petr: Renames Again and Again and Again\n>\n>  IRC: detection of wholesame renames of directories (WIP) and '--follow'\n>  limitation were mentioned, but outcome is unclear; pasky plans to hack\n>  together some patch implementing explicit renames hinting\n\nOne thing I didn't get around to bringing up: one of the benefits of\ndiff-time rename detection that is often touted is that algorithms can\nimprove over time. Do folk here know whether that has actually\nhappened recently, in a general way? Do people actually expect major\nimprovements in the future?\n\n> * Tom: GitHub\n>\n>  IRC: a tour, some history, and insight into how it works; some nice\n>  gimmicks, such as \"gist\" (a git backed pastebin), or like network graphs\n>  (look graphically at forks of a repository).\n>\n>  blog: Tom got to demonstrate GitHub and Gist to the group, most of whom\n>  are very command line oriented and had not used either before.\n\nThe demo of iGitHub (an iPhone app that can act as a clone/push\ntarget) looked really cool, if it can get further development. It\ncould potentially be really handy for travellers who could push to\ntheir iPhone, and then push from there to an internet server.\n\n> * Scott: Linkable Library\n>\n>  blog: got to talk about the need for a linkable git library\n>\n>  http://thread.gmane.org/gmane.comp.version-control.git/99608\n\nIt's good to see this starting to get wider traction. I think we\ndiscussed that there could be benefits to git itself, beyond just\nhelping other programs access git repositories faster than fork/exec.\n\n> * Sam: perl.git\n>\n>  blog: Sam demonstrated the work he went through to import 20 years of Perl\n>  history into the git repository that the Perl team is just now finishing\n>  transitioning to from Perforce.\n>\n>  http://utsl.gen.nz/talks/perl-history/slides/\n\nIt was very cool to see old-school email addresses like <isis!aburt>\nin git, handled just fine.\n\n> * Tim: Git as a Media Repository\n>  http://www.thousandparsec.net/~tim/media+git.pdf\n\nThis has kicked off some mailing list discussion; I think this can be\na major weak point for git, since checking out only a subtree (and\nonly the latest revision) is the common SVN way, which copes with\nmedia repositories and the like just fine.\n\n\n> * JH: $$$ discussion\n>\n>  IRC: It was about the Google Summer of Code 2008 mentor money. It was\n>  decided (given most of the mentors were there) to use it to help a few\n>  folks cover their travel expenses to the GitTogether. Most folks actually\n>  had their costs covered by their employer (yay for some big companies!)\n>  but a few paid out of pocket, so we are helping them defray it a bit. and\n>  the rest is being used to order and distribute some git t-shirts.\n\nYes, how are the t-shirts going? I seem to remember JH had volunteered\nto do the logistics there.\n\n\nDave.\n"},{"id":"95215","messageId":"20081108141727.GA17100@coredump.intra.peff.net","threadId":"16217","inReplyTo":"200811080441.25796.johan@herland.net","subject":"Re: How it was at GitTogether'08 ?","fromName":"Jeff King","fromEmail":"peff@peff.net","sentAt":"2008-11-08T14:17:27Z","receivedAt":"2008-11-08T14:17:27Z","isPatch":false,"sender":{"key":"peff@peff.net","avatar":"https://avatars.githubusercontent.com/u/45925?v=4"},"body":"On Sat, Nov 08, 2008 at 04:41:25AM +0100, Johan Herland wrote:\n\n> > * Discussion on notes\n> \n> Can someone elaborate on this? AFAIK, notes have popped up on this list \n> often enough that I'm convinced it would be a _really_ useful feature. The \n> only drawback I was aware of, was the lack of an efficient implementation, \n> but then Jeff comes out of the blue and posts some interesting numbers [1] \n> a week or so ago. Does this mean there are no remaining obstacles?\n> \n> [1]: http://article.gmane.org/gmane.comp.version-control.git/99415\n\nThe discussion was along the lines of \"here are some more cool things we\ncould do, if we had notes.\" I don't remember the specifics of the cool\nthings, but they were related to annotating patches with review\ninformation. Shawn can probably elaborate more.\n\nThat led to a \"notes as a tree are nice, but too slow because looking up\na tree entry is linear\" (and obviously you do a ton of lookups in the\nnotes tree during \"git log\"). Dscho had posted an implementation with a\npersistent notes cache long ago. Since I failed to actually look at\nthat, I started on a slightly different approach, which is simply doing\nan in-memory hash table to speedup the notes tree. And those are the\nnumbers and patch I posted.\n\nMy eventual plan was to re-work Dscho's patches with this performance\napproach. But it is not at the top of my queue, so if somebody else\nwanted to pick it up, I would be very happy. Everything I have done so\nfar is in the post you referenced.\n\nThe only other thing I remember discussing was notes namespaces. The two\nobvious approaches are:\n\n 1. a separate ref for each notes namespace, with each note ending up a\n    blob in a tree. So you might have refs/notes/acked-by:$SHA1 as a\n    blob.\n\n 2. one notes ref, with the notes tree pointing a sub-tree that has\n    named entries, one for each note type. So you might have\n    refs/notes:$SHA1/acked-by as a blob.\n\nThe advantage of '1' is that it keeps your different note types\nseparate, which means it is easy to distribute one type but not the\nother. The advantage of '2' is that I do one lookup per-commit, and then\nI can see all of the notes, which keeps performance nice when you want\nto annotate with several note types.\n\nAfter some discussion, I think Dscho and I came to the conclusion that\nsupporting both might be desirable. And it should be pretty\nstraightforward. You can just have multiple note refs (but default to a\n\"main\" one), and within each one, either point to a tree or blob (and we\nwill see which and use it appropriately).\n\nAnd then depending on which notes the user wants, they can refer to them\nappropriately. My suggestion for naming (and this wasn't discussed\nearlier, so Dscho has not endorsed this) would be something like\n\"$X:$Y\", which would mean \"to get the notes for $SHA1, look at the tree\nin refs/notes/$X for the file $SHA1/$Y\". If $Y is empty, then expect\n$SHA1 to be a blob (if it's a tree, maybe look at $SHA1/default). If\n\"$X\" is empty, then use \"refs/notes/default\". If there is no colon,\nassume we have \"$Y\".\n\nSo you could have a bunch of notes in some \"main\" namespace just by\ncalling them some name; without a name, you get some \"default\" note. But\nif you wanted a separate database (say, for SVN information), you could\nuse \"svn:\" or \"svn:name\".\n\n-Peff\n"},{"id":"95221","messageId":"200811081631.06229.jnareb@gmail.com","threadId":"16217","inReplyTo":"ee77f5c20811072108o21f97c97i8174f4f7ecd67030@mail.gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2008-11-08T15:31:04Z","receivedAt":"2008-11-08T15:31:04Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"On Sat, 8 Nov 2008, David Symonds wrote:\n> On Fri, Nov 7, 2008 at 5:54 PM, Jakub Narebski <jnareb@gmail.com> wrote:\n> \n> Here's my thoughts on some of these talks.\n> \n> > Mon, Oct 27, 2008\n> > -----------------\n> > * Dscho: Contributing with Git\n> >  http://www.youtube.com/watch?v=j45cs5_nY2k\n> \n> It was a good intro, but I was expecting a few more non-GitTogether\n> people. We had quite a large room, but there was only about a dozen\n> other people who came along. I don't know whether that was the fault\n> of the timing, lack of advertising, or a lack of interest.\n\nBy the way, it would be nice to have transcript for this talk, just\nlike there is for Linus talk:\n  http://git.or.cz/gitwiki/LinusTalk200705Transcript\n(but this would take some doing).\n \nIt would be also nice to have slides for the talk available somewhere,\njust like slides for \"Git Chronicle\".\n\n> > * Junio: Git Chronicle\n> >\n> >  blog: Junio went though a sort of statistical history of the Git project\n> >  that was fascinating (turns out there are still about 220 lines of code\n> >  still around from Linus original first commit).\n> \n> This was really interesting. It would be great to put this on a\n> general web page instead of in a PDF.\n\nSomething like Sam Vilain slides from \"perl.git\" talk?, \n  http://utsl.gen.nz/talks/perl-history/slides/\n\nIt shouldn't be that hard, depending on the original program the slides\nwere made... well, it was made in Impress from OpenOffice.org 2.4; it\nmight have export to (X)HTML + images, and to SWF (Flash presentation).\n\n> > * Petr: Renames Again and Again and Again\n> >\n> >  IRC: detection of wholesame renames of directories (WIP) and '--follow'\n> >  limitation were mentioned, but outcome is unclear; pasky plans to hack\n> >  together some patch implementing explicit renames hinting\n> \n> One thing I didn't get around to bringing up: one of the benefits of\n> diff-time rename detection that is often touted is that algorithms can\n> improve over time. Do folk here know whether that has actually\n> happened recently, in a general way? Do people actually expect major\n> improvements in the future?\n\nIf I remember correctly there was at least one improvement in rename\ndetection, namely better talking into account filename similarity score,\nso for example similar files moved (or copied) didn't get marked as\ncoming from one source (and rest deleted).\n\n> > * Tom: GitHub\n> >\n> >  IRC: a tour, some history, and insight into how it works; some nice\n> >  gimmicks, such as \"gist\" (a git backed pastebin), or like network graphs\n> >  (look graphically at forks of a repository).\n> >\n> >  blog: Tom got to demonstrate GitHub and Gist to the group, most of whom\n> >  are very command line oriented and had not used either before.\n> \n> The demo of iGitHub (an iPhone app that can act as a clone/push\n> target) looked really cool, if it can get further development. It\n> could potentially be really handy for travellers who could push to\n> their iPhone, and then push from there to an internet server.\n\niGitHub has nothing to do with GitHub; I think you put the comment in\na wrong place; the iGitHub (or iGit / iGitRouter) was a separate talk\nin \"Lighting Round Talks\" next day.\n\n> > * Scott: Linkable Library\n> >\n> >  blog: got to talk about the need for a linkable git library\n> >\n> >  http://thread.gmane.org/gmane.comp.version-control.git/99608\n> \n> It's good to see this starting to get wider traction. I think we\n> discussed that there could be benefits to git itself, beyond just\n> helping other programs access git repositories faster than fork/exec.\n\nWhat benefits would be those? Current design of \"fire and forget\",\nwhich stopped libification efforts till now was used for a reason...\n\n> > * Sam: perl.git\n> >\n> >  blog: Sam demonstrated the work he went through to import 20 years of Perl\n> >  history into the git repository that the Perl team is just now finishing\n> >  transitioning to from Perforce.\n> >\n> >  http://utsl.gen.nz/talks/perl-history/slides/\n> \n> It was very cool to see old-school email addresses like <isis!aburt>\n> in git, handled just fine.\n\nThis is not suprising, as Git treats committer and author email data\nas opaque data, not analysing it at all (some commits from early\nversions of git might not have this data at all, IIRC).\n \n> > * Tim: Git as a Media Repository\n> >  http://www.thousandparsec.net/~tim/media+git.pdf\n> \n> This has kicked off some mailing list discussion; I think this can be\n> a major weak point for git, since checking out only a subtree (and\n> only the latest revision) is the common SVN way, which copes with\n> media repositories and the like just fine.\n\nWell, you can workaround this weakness by (ab)using submodules...\n...and one should always remember that casual partial checkouts\ninterfere a bit with whole-tree commits.\n\n-- \nJakub Narebski\nPoland\n"},{"id":"95254","messageId":"200811091249.40735.jnareb@gmail.com","threadId":"16217","inReplyTo":"200811080254.53202.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2008-11-09T11:49:39Z","receivedAt":"2008-11-09T11:49:39Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"On Sat, 8 Nov 2008, Jakub Narębski wrote:\n\n> GitTogether 2008, which took place October 27(Mon)-29(Wed), has ended more\n> than week ago.  Therefore I'd like to ask you impressions (while it is still\n> fresh) from GitTogether, describe talks which are not described below,\n> correct wrong information in below, etc.\n> \n> Please reply to this email while GitTogether is on front of your mind...\n\nStill missing (neither video, nor slides, nor description, nor email)\nare the following talks from GitTogether'08\n\n \n> Talks at GitTogether 2008\n> =========================\n> \n> Mon, Oct 27, 2008\n> -----------------\n> * Jeff: Helping new contributors join\n> * Shawn: JGit\n \n> Tue, Oct 28, 2008\n> -----------------\n> * Jeff: What needs refactoring?\n> * Lighting Round Talks\n>   - Jeff: Portability autobuilders\n>   - Petr: Quick TopGit introduction, problems, future plans\n> * Steven: Life with git-svn\n\n> Wed, Oct 29, 2008\n> -----------------\n+ What was the difference between \"Tim: Large media in Git (Repeat)\"\n  from Wed, and earlier \"Tim: Git as a Media Repository\" from Tue?\n\n-- \nJakub Narebski\nPoland\n"},{"id":"95267","messageId":"200811091636.55343.kai@samba.org","threadId":"16217","inReplyTo":"200811081631.06229.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Kai Blin","fromEmail":"kai@samba.org","sentAt":"2008-11-09T15:36:51Z","receivedAt":"2008-11-09T15:36:51Z","isPatch":false,"sender":{"key":"kai@samba.org","avatar":"https://gravatar.com/avatar/bc7698f10d926406cee9c7bad46d2b345249777dadd9ece4501f6f2384a03e52?d=mp&s=160"},"body":"On Saturday 08 November 2008 16:31:04 Jakub Narebski wrote:\n\n> > > * Tim: Git as a Media Repository\n> > >  http://www.thousandparsec.net/~tim/media+git.pdf\n> >\n> > This has kicked off some mailing list discussion; I think this can be\n> > a major weak point for git, since checking out only a subtree (and\n> > only the latest revision) is the common SVN way, which copes with\n> > media repositories and the like just fine.\n>\n> Well, you can workaround this weakness by (ab)using submodules...\n> ...and one should always remember that casual partial checkouts\n> interfere a bit with whole-tree commits.\n\nInteresting. How would you use submodules to work around the fact that binary \nfile changes diff very bad and produce huge histories with basically no value \nfor the user of the working copy? Can you do this from a GUI, easily? We're \ntalking about media repositories here, so our users are artists.\n\nCheers,\nKai\n\n\n-- \nKai Blin\nWorldForge developer  http://www.worldforge.org/\nWine developer        http://wiki.winehq.org/KaiBlin\nSamba team member     http://www.samba.org/samba/team/\n--\nWill code for cotton.\n"},{"id":"95268","messageId":"200811091731.48694.jnareb@gmail.com","threadId":"16217","inReplyTo":"200811091636.55343.kai@samba.org","subject":"Re: How it was at GitTogether'08 ?","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2008-11-09T16:31:47Z","receivedAt":"2008-11-09T16:31:47Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"On Sun, 9 Nov 2008, Kai Blin wrote:\n> On Saturday 08 November 2008 16:31:04 Jakub Narebski wrote:\n\n>>>> * Tim: Git as a Media Repository\n>>>>  http://www.thousandparsec.net/~tim/media+git.pdf\n>>>\n>>> This has kicked off some mailing list discussion; I think this can be\n>>> a major weak point for git, since checking out only a subtree (and\n>>> only the latest revision) is the common SVN way, which copes with\n>>> media repositories and the like just fine.\n>>\n>> Well, you can workaround this weakness by (ab)using submodules...\n>> ...and one should always remember that casual partial checkouts\n>> interfere a bit with whole-tree commits.\n> \n> Interesting. How would you use submodules to work around the fact that binary \n> file changes diff very bad and produce huge histories with basically no value \n> for the user of the working copy? Can you do this from a GUI, easily? We're \n> talking about media repositories here, so our users are artists.\n\nWhat I meant here (but perhaps was not clear) was (ab)using submodules\nto allow to have full working repository without large [media] files\nboth in object database (repository) and without them checked out.\n\nThe workaround is to put all large files for example in 'media/' folder,\nand make this folder be submodule. Each clone of repository can have\nthis 'media' submodule either present (both in object database, although\nusually separate from main project object database), or not present\n(not cloned and not checked out).\n\nAs to submodules UI and GUI support for submodules... currently it is\nunfortunately lacking.\n\n\nNote that I explicitly mentioned that (ab)using submodules to better\ndeal with large files is _workaround_, and not _solution_. Lazy clone\nin version proposed by Tim is IMHO correct solution.\n\nP.S. Could anybody document at last `delta' gitattribute?\nP.P.S. You can have separate diff driver for binary files, but I don't\n       know anyone who uses for example some such for images...\n-- \nJakub Narebski\nPoland\n"},{"id":"95269","messageId":"F0422D64-2A35-4755-BD28-782BFA634AAC@midwinter.com","threadId":"16217","inReplyTo":"200811091249.40735.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Steven Grimm","fromEmail":"koreth@midwinter.com","sentAt":"2008-11-09T16:52:58Z","receivedAt":"2008-11-09T16:52:58Z","isPatch":false,"sender":{"key":"koreth@midwinter.com","avatar":"https://gravatar.com/avatar/71b4d2e8b62f168bdc9e9205341159e3567003b4f9e2127c617c5fa0a1f5bad2?d=mp&s=160"},"body":"On Nov 9, 2008, at 3:49 AM, Jakub Narebski wrote:\n> Still missing (neither video, nor slides, nor description, nor email)\n> are the following talks from GitTogether'08\n>> * Steven: Life with git-svn\n\nHere are the slides from David's and my talk:\n\nhttp://docs.google.com/Presentation?id=dhhs72s2_1wtzbnsnj&invite=v4t8kr\n\n-Steve\n"},{"id":"95270","messageId":"200811091854.11411.jnareb@gmail.com","threadId":"16217","inReplyTo":"F0422D64-2A35-4755-BD28-782BFA634AAC@midwinter.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2008-11-09T17:54:09Z","receivedAt":"2008-11-09T17:54:09Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"Steven Grimm wrote:\n> On Nov 9, 2008, at 3:49 AM, Jakub Narebski wrote:\n>>\n>> Still missing (neither video, nor slides, nor description, nor email)\n>> are the following talks from GitTogether'08\n>>\n>>> * Steven: Life with git-svn\n> \n> Here are the slides from David's and my talk:\n> \n> http://docs.google.com/Presentation?id=dhhs72s2_1wtzbnsnj&invite=v4t8kr\n\nThanks. Added to http://git.or.cz/gitwiki/GitTogether\n(even if I cannot view them in my old web browser).\n\n-- \nJakub Narebski\nPoland\n"},{"id":"95275","messageId":"200811091955.19304.kai@samba.org","threadId":"16217","inReplyTo":"200811091731.48694.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Kai Blin","fromEmail":"kai@samba.org","sentAt":"2008-11-09T18:55:14Z","receivedAt":"2008-11-09T18:55:14Z","isPatch":false,"sender":{"key":"kai@samba.org","avatar":"https://gravatar.com/avatar/bc7698f10d926406cee9c7bad46d2b345249777dadd9ece4501f6f2384a03e52?d=mp&s=160"},"body":"On Sunday 09 November 2008 17:31:47 Jakub Narebski wrote:\n\n\n> The workaround is to put all large files for example in 'media/' folder,\n> and make this folder be submodule. Each clone of repository can have\n> this 'media' submodule either present (both in object database, although\n> usually separate from main project object database), or not present\n> (not cloned and not checked out).\n\nTim was talking about that media/ folder and managing that in git. If you want \nto work on the media, you might end up getting hundreds of gigabytes of data \nto get that folder, even if you only need to change one single file.\n\nThat's the issue we're running into, and I don't thing submodules solve this \nat all.\n\nCheers,\nKai\n\n-- \nKai Blin\nWorldForge developer  http://www.worldforge.org/\nWine developer        http://wiki.winehq.org/KaiBlin\nSamba team member     http://www.samba.org/samba/team/\n--\nWill code for cotton.\n"},{"id":"95277","messageId":"200811091958.32893.robin.rosenberg@dewire.com","threadId":"16217","inReplyTo":"F0422D64-2A35-4755-BD28-782BFA634AAC@midwinter.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Robin Rosenberg","fromEmail":"robin.rosenberg@dewire.com","sentAt":"2008-11-09T18:58:32Z","receivedAt":"2008-11-09T18:58:32Z","isPatch":false,"sender":{"key":"robin.rosenberg@dewire.com","avatar":"https://avatars.githubusercontent.com/u/46357?v=4"},"body":"söndag 09 november 2008 17:52:58 skrev Steven Grimm:\n> On Nov 9, 2008, at 3:49 AM, Jakub Narebski wrote:\n> > Still missing (neither video, nor slides, nor description, nor email)\n> > are the following talks from GitTogether'08\n> >> * Steven: Life with git-svn\n> \n> Here are the slides from David's and my talk:\n> \n> http://docs.google.com/Presentation?id=dhhs72s2_1wtzbnsnj&invite=v4t8kr\n\nRequiring a google (or facebook for that matter) account is read public documents\nis not nice.\n\n-- robin\n"},{"id":"95278","messageId":"20081109195445.GC2932@spearce.org","threadId":"16217","inReplyTo":"200811091249.40735.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Shawn O. Pearce","fromEmail":"spearce@spearce.org","sentAt":"2008-11-09T19:54:45Z","receivedAt":"2008-11-09T19:54:45Z","isPatch":false,"sender":{"key":"spearce@spearce.org","avatar":"https://avatars.githubusercontent.com/u/34844?v=4"},"body":"Jakub Narebski <jnareb@gmail.com> wrote:\n> \n> Still missing (neither video, nor slides, nor description, nor email)\n> are the following talks from GitTogether'08\n\nJGit slides:\nhttp://www.spearce.org/2008/11/JGit.pdf\n\nPack v4 slides:\nhttp://www.spearce.org/2008/11/Pack_v4.pdf\n\nI also had these as a Google Doc, but its internal on the google.com\ndomain so I can't easily publish it.  Plus PDF is probably more\nportable to older browsers than the Google Docs site is. :-)\n\n-- \nShawn.\n"},{"id":"95279","messageId":"20081109195538.GD2932@spearce.org","threadId":"16217","inReplyTo":"200811091958.32893.robin.rosenberg@dewire.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Shawn O. Pearce","fromEmail":"spearce@spearce.org","sentAt":"2008-11-09T19:55:38Z","receivedAt":"2008-11-09T19:55:38Z","isPatch":false,"sender":{"key":"spearce@spearce.org","avatar":"https://avatars.githubusercontent.com/u/34844?v=4"},"body":"Robin Rosenberg <robin.rosenberg@dewire.com> wrote:\n> söndag 09 november 2008 17:52:58 skrev Steven Grimm:\n> > On Nov 9, 2008, at 3:49 AM, Jakub Narebski wrote:\n> > > Still missing (neither video, nor slides, nor description, nor email)\n> > > are the following talks from GitTogether'08\n> > >> * Steven: Life with git-svn\n> > \n> > Here are the slides from David's and my talk:\n> > \n> > http://docs.google.com/Presentation?id=dhhs72s2_1wtzbnsnj&invite=v4t8kr\n> \n> Requiring a google (or facebook for that matter) account is read public documents\n> is not nice.\n\nNo, its not.  The owner of the document can publish the document,\nmaking it world-readable, *without* needing a login.  I think they\njust forgot to do that on this particular presentation.\n\nSteven, can you publish that doc, so it doesn't require login\nto read?\n\n-- \nShawn.\n"},{"id":"95284","messageId":"2E4FE676-0128-4664-86A1-5EED75876A37@midwinter.com","threadId":"16217","inReplyTo":"20081109195538.GD2932@spearce.org","subject":"Re: How it was at GitTogether'08 ?","fromName":"Steven Grimm","fromEmail":"koreth@midwinter.com","sentAt":"2008-11-09T21:58:15Z","receivedAt":"2008-11-09T21:58:15Z","isPatch":false,"sender":{"key":"koreth@midwinter.com","avatar":"https://gravatar.com/avatar/71b4d2e8b62f168bdc9e9205341159e3567003b4f9e2127c617c5fa0a1f5bad2?d=mp&s=160"},"body":"On Nov 9, 2008, at 11:55 AM, Shawn O. Pearce wrote:\n> No, its not.  The owner of the document can publish the document,\n> making it world-readable, *without* needing a login.  I think they\n> just forgot to do that on this particular presentation.\n\n\nMy bad. Here's the published, no-login-required version:\n\nhttp://docs.google.com/Presentation?id=dhhs72s2_1wtzbnsnj\n\n-Steve\n"},{"id":"95286","messageId":"200811092303.46127.jnareb@gmail.com","threadId":"16217","inReplyTo":"20081109195445.GC2932@spearce.org","subject":"Re: How it was at GitTogether'08 ?","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2008-11-09T22:03:41Z","receivedAt":"2008-11-09T22:03:41Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"Shawn O. Pearce wrote:\n> Jakub Narebski <jnareb@gmail.com> wrote:\n> > \n> > Still missing (neither video, nor slides, nor description, nor email)\n> > are the following talks from GitTogether'08\n> \n> JGit slides:\n> http://www.spearce.org/2008/11/JGit.pdf\n> \n> Pack v4 slides:\n> http://www.spearce.org/2008/11/Pack_v4.pdf\n\nThanks. Added to http://git.or.cz/gitwiki/GitTogether\n\n-- \nJakub Narebski\nPoland\n"},{"id":"95291","messageId":"49177300.1090308@gmx.ch","threadId":"16217","inReplyTo":"200811080254.53202.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Jean-Luc Herren","fromEmail":"jlh@gmx.ch","sentAt":"2008-11-09T23:32:16Z","receivedAt":"2008-11-09T23:32:16Z","isPatch":false,"sender":{"key":"jlh@gmx.ch","avatar":null},"body":"Jakub Narebski wrote:\n> and the rest is being used to order and distribute some git\n> t-shirts.\n\nDo such git shirts exist already and could be ordered somewhere?\nI might be interested in buying one.\n\nCheers,\njlh\n"},{"id":"95292","messageId":"200811100052.16740.jnareb@gmail.com","threadId":"16217","inReplyTo":"2E4FE676-0128-4664-86A1-5EED75876A37@midwinter.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2008-11-09T23:52:12Z","receivedAt":"2008-11-09T23:52:12Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"Steven Grimm wrote:\n> On Nov 9, 2008, at 11:55 AM, Shawn O. Pearce wrote:\n\n> > No, its not.  The owner of the document can publish the document,\n> > making it world-readable, *without* needing a login.  I think they\n> > just forgot to do that on this particular presentation.\n> \n> \n> My bad. Here's the published, no-login-required version:\n> \n> http://docs.google.com/Presentation?id=dhhs72s2_1wtzbnsnj\n\nWould it be possible to publish PDF version sowehere, for example\nas attachement to http://git.or.cz/gitwiki/GitTogether? My old\nweb browser doesn't support Google Docs...\n\n-- \nJakub Narebski\nPoland\n"},{"id":"95299","messageId":"1226287933.21620.10.camel@vaio","threadId":"16217","inReplyTo":"200811091249.40735.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Tim Ansell","fromEmail":"mithro@mithis.com","sentAt":"2008-11-10T03:32:13Z","receivedAt":"2008-11-10T03:32:13Z","isPatch":false,"sender":{"key":"mithro@mithis.com","avatar":"https://gravatar.com/avatar/1df49ebfaff79712e08212ff66228e688973c823b6452c23268f9a8c0e4d341b?d=mp&s=160"},"body":"> > Wed, Oct 29, 2008\n> > -----------------\n> + What was the difference between \"Tim: Large media in Git (Repeat)\"\n>   from Wed, and earlier \"Tim: Git as a Media Repository\" from Tue?\n\nThe first day I got people feed back on my proposal (whiteboard\ndiscussion) and then presented the proposal on Wednesday.\n\nYou can find the slides of the presentation at \nhttp://www.thousandparsec.net/~tim/media+git.pdf\n\nSorry about the duplicate messages, I hit send too soon.\n\nTim 'mithro' Ansell\n"},{"id":"95314","messageId":"200811101030.54070.jnareb@gmail.com","threadId":"16217","inReplyTo":"200811091955.19304.kai@samba.org","subject":"Re: Large media in git (was: How it was at GitTogether'08)?","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2008-11-10T09:30:53Z","receivedAt":"2008-11-10T09:30:53Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"On Sun, 9 Nov 2008, Kai Blin wrote:\n> On Sunday 09 November 2008 17:31:47 Jakub Narebski wrote:\n> \n> > The workaround is to put all large files for example in 'media/' folder,\n> > and make this folder be submodule. Each clone of repository can have\n> > this 'media' submodule either present (both in object database, although\n> > usually separate from main project object database), or not present\n> > (not cloned and not checked out).\n> \n> Tim was talking about that media/ folder and managing that in git. If you want \n> to work on the media, you might end up getting hundreds of gigabytes of data \n> to get that folder, even if you only need to change one single file.\n> \n> That's the issue we're running into, and I don't thing submodules solve this \n> at all.\n\nAh, well... Submodules cannot be workaround for _this_ issue. You can\nhave only all or nothing: either all files in media/ or none of them,\nboth in working directory like in repository object database... well\nunless you subdivide further.\n\nI guess that mentioned work on the media is in remote setting (you\ncannot have main repository on network drive) so Dana How's proposed\nsolution would not work for you, is it?\n\n-- \nJakub Narebski\nPoland\n"},{"id":"95317","messageId":"alpine.DEB.1.00.0811101055020.30769@pacific.mpi-cbg.de","threadId":"16217","inReplyTo":"200811091955.19304.kai@samba.org","subject":"Re: How it was at GitTogether'08 ?","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2008-11-10T09:58:05Z","receivedAt":"2008-11-10T09:58:05Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nOn Sun, 9 Nov 2008, Kai Blin wrote:\n\n> On Sunday 09 November 2008 17:31:47 Jakub Narebski wrote:\n> \n> > The workaround is to put all large files for example in 'media/' \n> > folder, and make this folder be submodule. Each clone of repository \n> > can have this 'media' submodule either present (both in object \n> > database, although usually separate from main project object \n> > database), or not present (not cloned and not checked out).\n> \n> Tim was talking about that media/ folder and managing that in git. If \n> you want to work on the media, you might end up getting hundreds of \n> gigabytes of data to get that folder, even if you only need to change \n> one single file.\n> \n> That's the issue we're running into, and I don't thing submodules solve \n> this at all.\n\nYou'd have to have a single repository for each and every media file, and \nyou'd need to use shallow clones and shallow fetches.\n\nHowever, a push-conflict will probably be beyond any non-programmer \nskillz.\n\nI'd rather propose to have a different interface, like through a web \nserver, where the user can say \"I have some cool new graphics, in this \n.zip file\" together with a commit message.\n\nKind of a git-gui via browser.\n\nCiao,\nDscho\n"},{"id":"95321","messageId":"200811101109.02885.kai@samba.org","threadId":"16217","inReplyTo":"alpine.DEB.1.00.0811101055020.30769@pacific.mpi-cbg.de","subject":"Re: How it was at GitTogether'08 ?","fromName":"Kai Blin","fromEmail":"kai@samba.org","sentAt":"2008-11-10T10:08:51Z","receivedAt":"2008-11-10T10:08:51Z","isPatch":false,"sender":{"key":"kai@samba.org","avatar":"https://gravatar.com/avatar/bc7698f10d926406cee9c7bad46d2b345249777dadd9ece4501f6f2384a03e52?d=mp&s=160"},"body":"On Monday 10 November 2008 10:58:05 Johannes Schindelin wrote:\nHi Dscho,\n\n> > Tim was talking about that media/ folder and managing that in git. If\n> > you want to work on the media, you might end up getting hundreds of\n> > gigabytes of data to get that folder, even if you only need to change\n> > one single file.\n> >\n> > That's the issue we're running into, and I don't thing submodules solve\n> > this at all.\n>\n> You'd have to have a single repository for each and every media file, and\n> you'd need to use shallow clones and shallow fetches.\n>\n> However, a push-conflict will probably be beyond any non-programmer\n> skillz.\n\nOk, I agree. But you could work around that by teaching the artists to \nfetch/rebase/push instead of just pushing, or hiding this in the GUI. If \nthere's a conflict on a binary data file you're screwed anyway. :)\n\n> I'd rather propose to have a different interface, like through a web\n> server, where the user can say \"I have some cool new graphics, in this\n> .zip file\" together with a commit message.\n>\n> Kind of a git-gui via browser.\n\nIncidentally I'm currently working on something like this, just aimed at \nthe \"artist side\", instead of the VCS side. This certainly is a useable \nsolution for artists. But at some point a developer will want to check out \nthe repository to cut a release tarball, and we're back to wanting shallow \nand narrow clones. :)\n\nCheers,\nKai\n\n-- \nKai Blin\nWorldForge developer  http://www.worldforge.org/\nWine developer        http://wiki.winehq.org/KaiBlin\nSamba team member     http://www.samba.org/samba/team/\n--\nWill code for cotton.\n"},{"id":"95322","messageId":"200811101113.17018.kai@samba.org","threadId":"16217","inReplyTo":"200811101030.54070.jnareb@gmail.com","subject":"Re: Large media in git (was: How it was at GitTogether'08)?","fromName":"Kai Blin","fromEmail":"kai@samba.org","sentAt":"2008-11-10T10:13:16Z","receivedAt":"2008-11-10T10:13:16Z","isPatch":false,"sender":{"key":"kai@samba.org","avatar":"https://gravatar.com/avatar/bc7698f10d926406cee9c7bad46d2b345249777dadd9ece4501f6f2384a03e52?d=mp&s=160"},"body":"On Monday 10 November 2008 10:30:53 Jakub Narebski wrote:\n> On Sun, 9 Nov 2008, Kai Blin wrote:\n> > On Sunday 09 November 2008 17:31:47 Jakub Narebski wrote:\n> > > The workaround is to put all large files for example in 'media/'\n> > > folder, and make this folder be submodule. Each clone of repository can\n> > > have this 'media' submodule either present (both in object database,\n> > > although usually separate from main project object database), or not\n> > > present (not cloned and not checked out).\n> >\n> > Tim was talking about that media/ folder and managing that in git. If you\n> > want to work on the media, you might end up getting hundreds of gigabytes\n> > of data to get that folder, even if you only need to change one single\n> > file.\n> >\n> > That's the issue we're running into, and I don't thing submodules solve\n> > this at all.\n>\n> Ah, well... Submodules cannot be workaround for _this_ issue. You can\n> have only all or nothing: either all files in media/ or none of them,\n> both in working directory like in repository object database... well\n> unless you subdivide further.\n>\n> I guess that mentioned work on the media is in remote setting (you\n> cannot have main repository on network drive) so Dana How's proposed\n> solution would not work for you, is it?\n\nIf my google-fu worked, the proposed solution you're talking about involves \nsimply hiding large blobs from pack files, right? In that case it won't work, \nas the users of the repository indeed are remote.\n\nCheers,\nKai\n\n-- \nKai Blin\nWorldForge developer  http://www.worldforge.org/\nWine developer        http://wiki.winehq.org/KaiBlin\nSamba team member     http://www.samba.org/samba/team/\n--\nWill code for cotton.\n"},{"id":"95325","messageId":"alpine.DEB.1.00.0811101125360.30769@pacific.mpi-cbg.de","threadId":"16217","inReplyTo":"200811081631.06229.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2008-11-10T10:38:23Z","receivedAt":"2008-11-10T10:38:23Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nOn Sat, 8 Nov 2008, Jakub Narebski wrote:\n\n> On Sat, 8 Nov 2008, David Symonds wrote:\n> > On Fri, Nov 7, 2008 at 5:54 PM, Jakub Narebski <jnareb@gmail.com> wrote:\n> > \n> > > Mon, Oct 27, 2008\n> > > -----------------\n> > > * Dscho: Contributing with Git\n> > >  http://www.youtube.com/watch?v=j45cs5_nY2k\n> > \n> > It was a good intro, but I was expecting a few more non-GitTogether \n> > people. We had quite a large room, but there was only about a dozen \n> > other people who came along. I don't know whether that was the fault \n> > of the timing, lack of advertising, or a lack of interest.\n> \n> By the way, it would be nice to have transcript for this talk, just\n> like there is for Linus talk:\n>   http://git.or.cz/gitwiki/LinusTalk200705Transcript\n> (but this would take some doing).\n>\n> It would be also nice to have slides for the talk available somewhere, \n> just like slides for \"Git Chronicle\".\n\nVoila:\n\nhttp://repo.or.cz/w/git/dscho.git?a=blob_plain;f=all-your-rebase.pdf;hb=f23c1c9868256cae8591bddafd64422f95d67b02\n\nCiao,\nDscho\n"},{"id":"95328","messageId":"200811101236.47752.jnareb@gmail.com","threadId":"16217","inReplyTo":"alpine.DEB.1.00.0811101125360.30769@pacific.mpi-cbg.de","subject":"Re: How it was at GitTogether'08 ?","fromName":"Jakub Narebski","fromEmail":"jnareb@gmail.com","sentAt":"2008-11-10T11:36:47Z","receivedAt":"2008-11-10T11:36:47Z","isPatch":false,"sender":{"key":"jnareb@gmail.com","avatar":"https://avatars.githubusercontent.com/u/2706?v=4"},"body":"On Mon, 10 Nov 2008, Johannes Schindelin wrote:\n> On Sat, 8 Nov 2008, Jakub Narebski wrote:\n>> On Sat, 8 Nov 2008, David Symonds wrote:\n>>> On Fri, Nov 7, 2008 at 5:54 PM, Jakub Narebski <jnareb@gmail.com> wrote:\n>>> \n>>>> Mon, Oct 27, 2008\n>>>> -----------------\n>>>> * Dscho: Contributing with Git\n>>>>  http://www.youtube.com/watch?v=j45cs5_nY2k\n>>> \n>>> It was a good intro, but I was expecting a few more non-GitTogether \n>>> people. We had quite a large room, but there was only about a dozen \n>>> other people who came along. I don't know whether that was the fault \n>>> of the timing, lack of advertising, or a lack of interest.\n[...]\n\n>> It would be also nice to have slides for the talk available somewhere, \n>> just like slides for \"Git Chronicle\".\n> \n> Voila:\n> \n> http://repo.or.cz/w/git/dscho.git?a=blob_plain;f=all-your-rebase.pdf;hb=f23c1c9868256cae8591bddafd64422f95d67b02\n\nThanks a lot. Added to http://git.or.cz/gitwiki/GitTogether\n\nP.S. I guess GitTogether wiki page should be perhaps rewritten...\n-- \nJakub Narebski\nPoland\n"},{"id":"95332","messageId":"alpine.DEB.1.00.0811101307250.30769@pacific.mpi-cbg.de","threadId":"16217","inReplyTo":"200811101109.02885.kai@samba.org","subject":"Re: How it was at GitTogether'08 ?","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2008-11-10T12:09:19Z","receivedAt":"2008-11-10T12:09:19Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nOn Mon, 10 Nov 2008, Kai Blin wrote:\n\n> On Monday 10 November 2008 10:58:05 Johannes Schindelin wrote:\n> \n> > > Tim was talking about that media/ folder and managing that in git. \n> > > If you want to work on the media, you might end up getting hundreds \n> > > of gigabytes of data to get that folder, even if you only need to \n> > > change one single file.\n> > >\n> > > That's the issue we're running into, and I don't thing submodules \n> > > solve this at all.\n> >\n> > You'd have to have a single repository for each and every media file, \n> > and you'd need to use shallow clones and shallow fetches.\n> >\n> > However, a push-conflict will probably be beyond any non-programmer \n> > skillz.\n> \n> Ok, I agree. But you could work around that by teaching the artists to \n> fetch/rebase/push instead of just pushing, or hiding this in the GUI. If \n> there's a conflict on a binary data file you're screwed anyway. :)\n\nA fetch in that case would make the artist download things she does not \nneed, right?  So maybe that is not the way to go.\n\n> > I'd rather propose to have a different interface, like through a web \n> > server, where the user can say \"I have some cool new graphics, in this \n> > .zip file\" together with a commit message.\n> >\n> > Kind of a git-gui via browser.\n> \n> Incidentally I'm currently working on something like this, just aimed at \n> the \"artist side\", instead of the VCS side. This certainly is a useable \n> solution for artists.\n\nSeems like a nice starting point for a git-gui-in-a-browser.\n\n> But at some point a developer will want to check out the repository to \n> cut a release tarball, and we're back to wanting shallow and narrow \n> clones. :)\n\nShallow clones are not an issue.  They _should_ work.\n\nAnd for releasing you do not want narrow clones.\n\nCiao,\nDscho\n"},{"id":"95493","messageId":"7vy6zqx7y0.fsf@gitster.siamese.dyndns.org","threadId":"16217","inReplyTo":"200811080254.53202.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2008-11-11T21:35:03Z","receivedAt":"2008-11-11T21:35:03Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Jakub Narebski <jnareb@gmail.com> writes:\n\n> * David Brown: Life with Git\n>   http://www.davidb.org/git/git-corp.pdf\n\nI really enjoyed this talk, which was in a weird sense very encouraging.\n\nBy the way, one thing David talked about that is not in the PDF slides but\nI found quite good to stress was the importance of good commit messages.\n\"Write in present tense, it will read much nicer and you'll appreciate it\nafter reading hundreds of them\".\n\nI'd actually say \"imperative mood\" instead of \"present tense\" (but they\nlook almost always the same in English), but in any case, it really gets\non my nerve to read commit message that talks things in past tense and I\noften end up rewriting other people's commit log messages.\n"},{"id":"95498","messageId":"2c6b72b30811111405u52824e2bg246912ec6dae2162@mail.gmail.com","threadId":"16217","inReplyTo":"200811100052.16740.jnareb@gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Jonas Fonseca","fromEmail":"jonas.fonseca@gmail.com","sentAt":"2008-11-11T22:05:34Z","receivedAt":"2008-11-11T22:05:34Z","isPatch":false,"sender":{"key":"jonas.fonseca@gmail.com","avatar":"https://gravatar.com/avatar/9b7fa23cce50269e5d164312b6ac5ae818a180f837b38f28f3bdf689dd7f96cd?d=mp&s=160"},"body":"On Mon, Nov 10, 2008 at 00:52, Jakub Narebski <jnareb@gmail.com> wrote:\n> Would it be possible to publish PDF version sowehere, for example\n> as attachement to http://git.or.cz/gitwiki/GitTogether? My old\n> web browser doesn't support Google Docs...\n\nhttp://docs.google.com/MiscCommands?command=saveasdoc&docID=dhhs72s2_1wtzbnsnj&exportFormat=pdf\n\n-- \nJonas Fonseca\n"},{"id":"95503","messageId":"alpine.DEB.1.00.0811120025241.30769@pacific.mpi-cbg.de","threadId":"16217","inReplyTo":"2c6b72b30811111405u52824e2bg246912ec6dae2162@mail.gmail.com","subject":"Re: How it was at GitTogether'08 ?","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2008-11-11T23:26:24Z","receivedAt":"2008-11-11T23:26:24Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nOn Tue, 11 Nov 2008, Jonas Fonseca wrote:\n\n> On Mon, Nov 10, 2008 at 00:52, Jakub Narebski <jnareb@gmail.com> wrote:\n> > Would it be possible to publish PDF version sowehere, for example\n> > as attachement to http://git.or.cz/gitwiki/GitTogether? My old\n> > web browser doesn't support Google Docs...\n> \n> http://docs.google.com/MiscCommands?command=saveasdoc&docID=dhhs72s2_1wtzbnsnj&exportFormat=pdf\n\nI also included whatever I could grab (did not take care of mugwumps html \npages yet) into my presentations branch.\n\nCiao,\nDscho\n"}]}