Re: [git-users] Permission denied ONLY after pulling bundles

[Thomas Ferris Nicolaisen]

On Monday, June 8, 2015 at 10:53:33 AM UTC+2, Rossella Barletta wrote:
>
> Yes i tryed on Windows and it works perfectly, it does not have the same 
> behaviour. The permission problems comes out only when i push from Linux 
> (after pulling a bundle).
>
> Basically having the git repo on Windows and cloning from it on Linux 
> gives me the permission problems after pushing the modifications from the 
> bundle file. If the clone is created on Windows, pushing after pulling a 
> bundle does not give any problem.
>
> Even thinking about a user permission problem, consider that the same user 
> could perfectly push before (i mean before pulling the bundle) and to avoid 
> any doubt i gave the chmod -R 777 permission to all the files after pulling 
> the bundle.
>
>>
>>
This thread has been cross-posted to the Git mailing list, and as it has 
received more traction there: 

 http://comments.gmane.org/gmane.comp.version-control.git/270771

I'd advise to abandon this thread.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Subversion --> GIT (Externals)

[Thomas Ferris Nicolaisen]

On Wednesday, June 3, 2015 at 9:31:40 AM UTC+2, reinhar...@yahoo.de wrote:
>
> Hi!
>
> how can SVN Externals be mapped to GIT?
>
> We're heavily using externals. We reference folders and files with 
> externals at a certain revision.
>

If your goal is to pick single files from shared projects, perhaps using 
something like https://git-annex.branchable.com/ could be an option. I was 
hoping that git-annex supports SVN as a backend, but unfortunately it does 
not (yet - you can write your own extensions though).

If you're willing to place these assets that you have through svn-externals 
on some other kind of server, you could use annex out of the box to sync 
single files into your repository. Just a half thought-through idea :)

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Query on git submodules

[Thomas Ferris Nicolaisen]

On Thursday, May 28, 2015 at 1:45:33 PM UTC+2, Mattias Vannergård wrote:
>
> Hi!
>
>>
Note that this topic continued on the Git mailing list here: 

http://comments.gmane.org/gmane.comp.version-control.git/269716

Mattias, perhaps you should post your reply there as well. 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Query on git submodules

[Thomas Ferris Nicolaisen]

On Wednesday, May 20, 2015 at 10:11:43 AM UTC+2, Sarah Frawley wrote:
>
> Hi there
>
> I am a design automation engineer supporting 200+ designers who use git 
> for hardware design.  We also use the submodule feature where we can have 
> quite complex hierarchy’s with 10+ layers.  We have experience issues with 
> re-use of design projects was we move from one project derivative to 
> another due to the complexity of the hierarchy along with lack of 
> discipline (using absolute paths in .gitmodule files). To enforce more 
> discipline I am currently working on a pre-commit hook to check the 
> integrity of .gitmodule files.  I would be interested in seeing how other 
> users in the community find submodules for large scale projects and if 
> there are any best known methods for using them.
>
>
I think submodules are still considered advanced for many of the users on 
this list, and many downright avoid them.

You may be able to gather some more feedback from sharing your plans and 
ideas with the Git development mailing list [1], as at least it is 
frequented by the submodule developers. 

I would also recommend investing a good few hours in trawling Google for 
"real" deep articles on experiences and recommendations on using 
submodules, as I know they are out there, but unfortunately buried in a mix 
of articles about "submodules for newbies" and "submodules sucks".

[1] https://gist.github.com/tfnico/4441562 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: post-receive hook does not get run

[Thomas Ferris Nicolaisen]

On Thursday, May 21, 2015 at 3:07:05 PM UTC+2, Mattias Vannergård wrote:
>
> Sorry! 
> Using git 1.9.5 and gerrit 2.11 
> BR 
> /Mattias


What do your push commands and URLs look like? Please post the entire 
command you use for pushing, and include the output of "git remote -vv" 
(anonymize company names in hostnames, etc).

If you are pushing to Gerrit, the hooks that you have on some 
filesystem-level repositories are probably not involved at all... 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: post-receive hook does not get run

[Thomas Ferris Nicolaisen]

On Thursday, May 21, 2015 at 2:35:08 PM UTC+2, Mattias Vannergård wrote:
>
>
> What am I missing here?
>

Oh, you forgot to mention your Git version :) 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: post-receive hook does not get run

[Thomas Ferris Nicolaisen]

On Thursday, May 21, 2015 at 2:35:08 PM UTC+2, Mattias Vannergård wrote:
>
> I have a server with a lot of bare repos, but none of these projects will 
> run the post-receive hook.
>
> For example:
> Path is /gitRepos/repositories/MyNewProject.git
>
> The content of the post-receive hook is
> !#/bin/sh
> echo "Hello"
>
> Running /gitRepos/repositories/MyNewProject.git/hooks/post-receive locally 
> gives the result
> Hello
>
> so obviously it is runnable. The post-receive file is listed as:
> -rwxr-xr-x 1 gitadmin users  109 May 21 13:57 post-receive
>
> Also, calling the script via ssh @server 
> "/gitRepos/repositories/MyNewProject.git/hooks/post-receive" also gives the 
> result
> Hello
>
> All users are in the users group.
>
> But pushing to MyNewProject yields nothing else but
> remote: Processing changes: refs: 1, done
>

That's odd. I don't see any text like that when I push to the remote. Are 
you running any special kind of Git-server in between? What is your git 
remote url? 

Keep in mind that the hook won't trigger if there is nothing new to push.

Trying it out locally (over ssh to the same machine), it works as expected:


[master][~/temp/foo]$ git push tfnico@localhost:temp/foo.git master 
 

Counting objects: 3, done.

Delta compression using up to 4 threads.

Compressing objects: 100% (2/2), done.

Writing objects: 100% (3/3), 306 bytes | 0 bytes/s, done.

Total 3 (delta 1), reused 0 (delta 0)

remote: hello

To tfnico@localhost:temp/foo.git

   934bafb..23aeb19  master -> master

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: to git developers : please provide a recent 2.x official 64 bit windows client

[Thomas Ferris Nicolaisen]

On Wednesday, May 20, 2015 at 10:55:28 AM UTC+2, mae.l...@web.de wrote:
>
> hi,
>
> thank you for the greatest vcs ever ...
>
> first the fact : 32 bit is virtually dead , it began 11 years ago with the 
> arrival of the first 32/64 bit hybrid processors.
>
> even the cheapest desktop pc and laptops featuring now cpu with 64bit 
> instructions.
>
> a 64 bit version of git in windows would speed up a lot of things (i/o 
> transactions, checksumming, etc) ... its not only a clever marketing slogan 
> :-)
>
> SO PLEASE PLEASE PLEASE PROVIDE AN OFFICIAL 64 BIT WINDOWS VERSION OF GIT. 
> EVEN BETTER IF IT IS A RECENT GIT 2.X :-)
>
> if for some strange reason you must provide a 32bit windows version for 
> compatibility reasons take an example on other professional project and 
> provide both : let the user decide which one he needs and of course his 
> computer supports ...
>
> thank you for considering
>


This mailing list is for volunteer Git users helping other Git users with 
their problems. There's not much Git development or releasing going on here 
:)

There is another list that focuses on developing Git for Windows: 
http://groups.google.com/group/msysgit

Their most recent efforts include a rewrite that brings the latest 2.x 
versions of Git to Windows, both 32 and 64 
bit: https://git-for-windows.github.io/

It's still a "pre-release" though, so be aware that it may be unstable.

More background is available on their 
wiki: https://github.com/msysgit/msysgit/wiki

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: [ANN] multigit: overlaid git repositories

[Thomas Ferris Nicolaisen]

On Thursday, May 14, 2015 at 1:47:11 PM UTC+2, Cosmin Apreutesei wrote:
>
> Hello everyone,
>
> Hope this is the right place to announce a new git tool.
>
> It's called multigit, and it allows you to work with multiple repositories 
> overlaid over a single directory. 
>
> URL: https://github.com/capr/multigit
>
> It's used extensively as the package manager for luapower[1], and it's 
> similar to vcsh[2] for those who know that project.
>
> It is useful for projects which are made of different components that are 
> developed separately, but which need to deploy files in different parts of 
> the directory structure of the project, like for instance:
>
>   * manage customizations made to a web app in a separate repository.
>   * putting your home directory under source control.
>   * package and/or config manager for a Linux distro.
>
> I hope someone finds it useful.
>
>
That does look very interesting. I haven't got a use-case for it at the 
moment, but I'm sure it could come in handy.

I've marked it for inclusion in the next edition of Git Rev News [1]

[1] http://git.github.io/rev_news/ 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: determine a linux kernel version of a patch file

[Thomas Ferris Nicolaisen]

On Thursday, May 14, 2015 at 7:07:06 PM UTC+2, Pawel Por wrote:
>
> Hi,
>
> At the beginning I'm sorry if this post is completely unrelated to this 
> mailing list.
> I'm trying to find the base linux kernel version from which a patch has 
> been created and posted to LKML.
> Most patches contain the index line. Is it the well known git index 
> (staging area) ?
> Can I use the index to find the exact kernel version from which a patch 
> was created.
> Please look at an example patch sent to LKML:
> https://lkml.org/lkml/2015/5/14/497
>
> It contains the following index line:
> index 8dd1e55..9b0c81e 100644
>
> I tried to find it using "git log" but I couldn't.
>
>
I'm guessing those are object blobs, which don't really appear in the 
normal git log.

Try doing a git show 9b0c81e to see if that blob exists in your repository 
at all. If it does, you can see which commit(s) point to it. More hints 
here: http://stackoverflow.com/questions/223678/which-commit-has-this-blob

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Trouble with the git-http-backend; Documentation unprecise

[Thomas Ferris Nicolaisen]

On Tuesday, May 12, 2015 at 9:40:13 AM UTC+2, Matthias Lantsch wrote:
>
> *Hello Git Community,*
> I usually always try to solve problems myself, but I really have gotten 
> onto my limits with this.
> Situation:
> I am trying to create a php "glue" script between apache and the git http 
> backend in order to allow myself to hook into all git processes via my php 
> framework
> (So auth and automatic building and version number handling can be handled 
> from my php git web frontend).
>
>
I'm afraid I don't understand the use-case. But if you have low-level 
concrete questions that don't get answered on this mailing list, you could 
try asking on the Git developers' mailing 
list: https://gist.github.com/tfnico/4441562

I think if I was writing a framework that would operate on Git 
repositories, I would use some library like libgit2 to interoperate with 
the repositories, instead of using the rather limited HTTP protocol. There 
is a dormant project that implements libgit2 bindings in PHP 
here: https://github.com/libgit2/php-git

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: migrate history from svn to git on already created git repo

[Thomas Ferris Nicolaisen]

On Tuesday, May 12, 2015 at 12:30:22 PM UTC+2, Yogesh Tyagi wrote:
>
> Hi,
>
> Is it possible to migrate svn repo history to git when we have already 
> created git repo. Can somebody please share the commands?
>
>
I once demoed how to connect the history of a freshly converted repository 
from SVN with an existing Git repo (which is an older export from the same 
SVN repo): 
 http://blog.tfnico.com/2010/10/gitsvn-6-grafting-together-svn-history.html

Hope that helps!

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Rebase onto history from another repository

[Thomas Ferris Nicolaisen]

On Tuesday, May 12, 2015 at 1:39:32 PM UTC+2, J66st wrote:
>
> I have two repositories A and B. A is my working repository. B is a bare 
> repository containing a more detailed history than A. This is illustrated 
> in the following simplified graphs.
>
> *Repository A:*
>
>   
>  /  / 
>
> A1D1-E--F--G [master]
>
>   \  \
>
>H--I--J [topic]
>
> *Repository B:*
>
>   
>  //
>
> A2---B2--C2--D2
>
>
> What I want is in repository A to replace history A1 thru D1 by history A2 
> thru D2:
>
> *New repository A:*
>
>   
>  //
>
> A2---B2--C2--D2--E--F--G [master]
>
>   \  \
>
>H--I--J [topic]
>
>
>
>
>
> Notice that both histories use identical tag names for tagging versions!
> My first idea is to set repository B as a remote to repository A, then 
> fetch from B, rebase E onto D2, then prune A1 thru D1. But wouldn't 
> fetching from repository B cause the tag names to clash?
>
>
>
As far as I can remember, incoming tags being fetch will overwrite current 
tags. Just try it in a copy of the repository and see what happens.

What I usually do when binding together history from two distinct graphs, 
is grafting. See for example:


   - https://git.wiki.kernel.org/index.php/GraftPoint
   - http://ben.straubnet.net/post/939181602/git-grafting-repositories
   - blog.tfnico.com/2010/10/gitsvn-6-grafting-together-svn-history.html

You could also consider rebase --onto: 
http://stackoverflow.com/questions/3810348/setting-git-parent-pointer-to-a-different-parent

 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Remote End Hung Up Error

[Thomas Ferris Nicolaisen]

On Monday, May 11, 2015 at 10:50:00 PM UTC+2, Stephen Scott wrote:
>
> Greetings;
>
>  
>
> I am getting this error when trying to retrieve a repo from git. Has 
> anyone seen this issue before, and if you have how did you resolve it.
>
>  
>
> Cloning into 'x'...
>
> remote: Counting objects: 187683, done
>
> remote: Finding sources: 100% (187683/187683)
>
> remote: Getting sizes: 100% (98684/98684)
>
> remote: Compressing objects: 89% (8667/9643)
>
> efrror: RPC failed; result=56, HTTP code = 200 MiB | 42.00 KiB/s
>

According to the googles, the above error message seems to appear when 
there is some issue on the network disturbing the HTTPS protocol, like a 
firewall, VPN or anti-virus. Try some of the workarounds offered on 
StackOverflow:

https://www.google.com/webhp?q=efrror%3A+RPC+failed%3B+result%3D56%2C+HTTP+code+%3D+200

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Remote End Hung Up Error

[Thomas Ferris Nicolaisen]

On Monday, May 11, 2015 at 10:50:00 PM UTC+2, Stephen Scott wrote:
>
> Greetings;
>
>  
>
> I am getting this error when trying to retrieve a repo from git. Has 
> anyone seen this issue before, and if you have how did you resolve it.
>

Please provide some information about your environment (both client and 
server):

   - Operating system (specifically which version)
   - Git version (git --version)
   - Git configuration (system, home, repository)
   - Other relevant environment details which you think might be interesting
   
And show us what kind of command/url you use for cloning. You can anonymize 
the domain/url, but we need to know if it starts with git:// or git@..., 
https://... or ssh://...

Can you clone from one directory to the other on the server?

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Extend .gitignore to support setting of a file size limit so that all files over a certain size will by automatically ignored

[Thomas Ferris Nicolaisen]

On Friday, May 8, 2015 at 11:53:02 AM UTC+2, Thomas Ferris Nicolaisen wrote:
>
> If you still can justify why this would be a valuable feature for a 
> noteworthy amount of Git users, the place to suggest new features is the 
> main Git mailing list: https://gist.github.com/tfnico/4441562
>

Please note that OP continued the discussion on the Git dev list here:

http://git.661346.n2.nabble.com/Please-consider-extending-gitignore-to-support-setting-of-a-file-size-limit-td7630203.html

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: newbie downloading git

[Thomas Ferris Nicolaisen]

 On Sunday, May 10, 2015 at 2:14:36 AM UTC+2, Tricia Finkle wrote:
>
> Hi,
>
> I'm just getting started with github and git.  I have tried to download 
> git several times from the git-smc site, source forge and mirror sites.  I 
> install it and it says "install successful", but then it is nowhere to be 
> found.  When I try to configure in Terminal it gives me an "illegal 
> instruction" so it seems like it is not really installed, despite the 
> installer copying files and saying the install is successful.  Am I doing 
> something wrong?  Can you install git on OS 10.6.8?  I don't get any error 
> messages so I don't even know where it is failing.  Advice would be greatly 
> appreciated.
>

Can you copy and paste the output from the following commands in Terminal?

   - git --version
   - which git
   - echo $PATH
   
In any case, it seems others have had this 
problem: 
http://stackoverflow.com/questions/22892890/git-for-mac-fails-to-launch-illegal-instruction

Perhaps it's a combination of using Snow Leopard with the OSX installer. 
There are a couple of other ways to install Git on Mac, mostly by using 
these providers/software managers programs:

* Homebrew, my personal favorite. Install it and run `brew install git`
* Fink, if you come from a Debian background, this should be suitable: 
http://www.finkproject.org/
* MacPorts, maybe the most mature/oldest/stable alternative: 
https://www.macports.org/

The key is that you should only have one of the above installed on your 
system, as they can step on each others feet.

Further more, make sure that the Homebrew/Fink/MacPorts installer you 
download is suitable for Snow Leopard.

If you don't want to install any of the above package managers, the last 
option I know about is installing XCode developer tools, as they bundle Git 
within.

In the end, be aware that you might end up with a really old version of 
Git, which, although Git has been super-stable and fully functional since 
1.7'ish, the finer details of Git documentation online might have outrun 
your Git version.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Extend .gitignore to support setting of a file size limit so that all files over a certain size will by automatically ignored

[Thomas Ferris Nicolaisen]

On Thursday, May 7, 2015 at 10:45:29 PM UTC+2, Roger Mendes wrote:
>
> Allow .gitignore to support setting a file size limit so that all files 
> over a certain size will by automatically ignored when performing git add, 
> commit -a...; 
>

Small note, there is no parameter to git-commit that will automatically add 
all untracked files.
 

> Exclusions to this size limit for files can be allowed by specifying 
> exceptions '!' syntax that already exists. It probably should be considered 
> to have a default limit already built into git of say 100MB (used by 
> github) which can be changed in the .gitignore file.
>
> This would avoid accidental commits of large files which are known to be 
> poorly supported.
>
> This seems like a fairly simple feature to add, I'm not sure why no one 
> has done this or thought of it given the number of complaints, questions... 
> with large file handling.
>
> We can ignore by file name attribute, why not by file size?
>
> Yes, I know I can add pre-commit scripts... to workaround this issue, but 
> there should be a better way (or maybe I'm missing something?).
>


The use-case is somewhat exotic. I think most people do some fair share of 
thinking before tracking new files, and the mental cost of manually leaving 
out or ignoring "large" files is not that high. If you do set some 
arbitrary limit on what constitutes a "large file which should not be 
tracked", you're trying to automate something which is not so easy to 
automate. 

I mean, 5 years from now, storage (SSD), network-speed and memory will be a 
lot cheaper, and Git will make many performance enhancements, so maybe by 
then it will perfectly OK to check in 100 MB files, or even GB files. You'd 
have to review and maintain these ".gitignore" size entries regularly.

If you still can justify why this would be a valuable feature for a 
noteworthy amount of Git users, the place to suggest new features is the 
main Git mailing list: https://gist.github.com/tfnico/4441562

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [git-users] git push does not update remote workdir

[Thomas Ferris Nicolaisen]

On Friday, May 1, 2015 at 11:08:43 PM UTC+2, Thomas Ferris Nicolaisen wrote:
>
> On Wednesday, April 29, 2015 at 1:53:42 PM UTC+2, Konrád Lőrinczi wrote:
>>
>> Unfortunately push-to-checkout did not result files in workdir after 
>> pushing to server, so it was not usable for me.
>>
>>>
>>> <https://github.com/git/git/commit/0855331941b723b227e93b33955bbe0b45025659>
>>>
>>
> Not sure if I'm missing out on something obvious here, but 
> push-to-checkout works as expected when trying it out:
>

Correction: the push-to-checkout hook only works since Git 2.4.

Git 2.3 introduced this updateInstead configuration, which is what I showed 
off in my example. The push-to-checkout hook is more a fix for those cases 
where the target repository gets manipulated between pushes or something 
like that. Depending on your use-case you may not need that.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [git-users] git push does not update remote workdir

[Thomas Ferris Nicolaisen]

On Wednesday, April 29, 2015 at 1:53:42 PM UTC+2, Konrád Lőrinczi wrote:
>
> Unfortunately push-to-checkout did not result files in workdir after 
> pushing to server, so it was not usable for me.
>
>>
>> 
>>
>
Not sure if I'm missing out on something obvious here, but push-to-checkout 
works as expected when trying it out:

# Note that Git version must be >= 2.3
[master][~/temp/foo-web]$ git --version
git version 2.3.6 

# Take some random repo and clone it to a "remote" location:
[~/temp]$ git clone foo foo-web
Cloning into 'foo-web'...
done.


# So, let's pretend that this non-bare repository is on our web-host:
[~/temp]$ cd foo-web

# Configure allowing incoming pushes
[master][~/temp/foo-web]$ git config receive.denyCurrentBranch updateInstead

# Create the basic hook, as examplified in the test in 
 https://github.com/git/git/commit/0855331941b723b227e93b33955bbe0b45025659
[master][~/temp/foo-web]$ vim .git/hooks/push-to-checkout

### contents of the push-to-checkout hook:

/bin/sh
echo >&2 updating from $(git rev-parse HEAD)
echo >&2 updating to "$1"

git update-index -q --refresh && git read-tree -u -m HEAD "$1" || {
status=$?
echo >&2 read-tree failed
exit $status
}

## EOF


# Make the hook executable
[master][~/temp/foo-web]$ chmod +x .git/hooks/push-to-checkout

#Now let's take it for a spin. Go back to the original repo:
[master][~/temp/foo-web]$ cd ..
[~/temp]$ cd foo
[master][~/temp/foo]$ git remote add web ../foo-web
[master][~/temp/foo]$ git push web master
Everything up-to-date

# OK, that wasn't very impressive. Make some changes first, a new file for 
example:
[master][~/temp/foo]$ echo `random_word` >> heya.txt; git add .;git commit 
-m `random_word`
[master b6765e2] overpaint
 1 file changed, 1 insertion(+)
 create mode 100644 heya.txt

# Away we go:
[master][~/temp/foo]$ git push web master
Counting objects: 3, done.
Delta compression using up to 4 threads.
Compressing objects: 100% (2/2), done.
Writing objects: 100% (3/3), 322 bytes | 0 bytes/s, done.
Total 3 (delta 0), reused 0 (delta 0)
To ../foo-web
   eb2711a..b6765e2  master -> master
[master][~/temp/foo]$ cd ..
[~/temp]$ cd foo-web

# TADA:
[master][~/temp/foo-web]$ ls
README   foo  heya.txt

Does that work for you?

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: My recent advances ;-)

[Thomas Ferris Nicolaisen]

On Sunday, April 12, 2015 at 12:10:17 PM UTC+2, Dryden wrote:
>
> That was a type: my first cause is to MAKE changes in a different 
> directory. I want to stay clear of changing master at all times, If I can, 
> and if it is even in the slightest way necessary (ie. stuff that can break 
> things).
>
> But merging BACK into Master is still the most difficult thing there is. 
> It is unclear to me. The Atlassian tutorial spends very little time on 
> merging.
>
> I don't know how to do it, or how to do it well. I am not sure how to 
> rebase onto the master in such a way that I can see what is going on. Maybe 
> it is just rebase  --onto master and I can try... and it doesn't 
> work.
>
>
Generally, if you have more concrete cases and examples of what you are 
trying to do, you'll get replies here quicker.

Conflicts are a natural thing in any version control system, Git included. 
If you change the same lines in the same files in two parallel branches, 
you are going to get conflicts whether you merge or rebase. That's just 
fact.

Now, if you are getting conflicts in places where you weren't expecting 
them at all, i.e. it doesn't make sense for these conflicts to appear, you 
may be doing something wrong. 

So please show us some concrete examples of what you are doing. Find the 
minimal example you can do to recreate your problem, and then copy/paste 
the terminal input/output in a mail back here.
 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [git-users] git push does not update remote workdir

[Thomas Ferris Nicolaisen]

I didn't get into the details of this discussion, but I just wanted to 
shoot in that there is a new hook since Git 2.3 called "push-to-checkout". 
Perhaps that would be the right thing here. Nice to know about anyhow:

https://github.com/git/git/commit/0855331941b723b227e93b33955bbe0b45025659

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git-svn clone. Ignore tags by mask/regex.

[Thomas Ferris Nicolaisen]

On Wednesday, April 22, 2015 at 7:37:32 PM UTC+2, Андрей Турбанов wrote:
>
> Hello.
>
> I am trying to clone svn repository.
> This repo has standart svn layout: trunks/branches/tags. 
> But there are some *non-project* related tags in tags directory.
>
> my-super-tool
> trunk
> branches
> tags
> my-super-tool-1.0
> my-super-tool-2.0
> my-super-tool-3.0
> PC_sec-1.0
> PC_sec-1.1
> PC_sec-1.2
> PC_trd-1.2
> PC_trd-1.2
> ...
>
> I want to clone with only *my-super-tool-** tags
> What I tried:
> git svn clone --trunk trunk --branches branches --tags tags/my-super-tool-*
> git svn clone -s --ignore-paths=tags/PC_*
>
>
> First doesn't work: git-svn is unable to accept mask for tags
> Seconed doesn't work too: git-svn clones this tags anyway
>

First of all, you're not using regular expressions to express the ignore 
paths. Quoting the git-svn man page:

   --ignore-paths=
>This allows one to specify a Perl regular expression that 
> will cause skipping of all matching paths from checkout from
>SVN. The --ignore-paths option should match for every fetch 
> (including automatic fetches due to clone, dcommit, rebase,
>etc) on a given repository.
>config key: svn-remote..ignore-paths
>If the ignore-paths configuration key is set, and the 
> command-line option is also given, both regular expressions will
>be used.
>Examples:
>Skip "doc*" directory for every fetch
>--ignore-paths="^doc"
>
>Skip "branches" and "tags" of first level directories
>--ignore-paths="^[^/]+/(?:branches|tags)"


Try using finding some examples of regular expressions used with git-svn, 
and try that out. Example:

 --ignore-paths='^(?!directory-a|directory-b)'

If there are too many things to exclude, also consider using the 
--include-paths option instead.

Is it possible to clone this type of svn repo?
> May be with some hack/workaround (I tried to hack git svn sources, but I'm 
> not an expert with perl)?
> Where can I add feature request for ability to use mask/regex for branches 
> and tags?
>

It's been a while since I worked with git-svn, so I can only confirm that 
it is non-trivial to ignore paths from being included in the clone like you 
want. I believe your case should be possible, but there's always the chance 
of bugs, depending on your version of Git, platform, etc etc.

You can also consider just doing a full clone without ignoring, and after 
that manually tidy up and delete branches that you don't want to keep.

If no-one here can answer you directly (alas, I lack the time to research 
myself), you should ask on the Git developer's mailing list, and please 
include mention of similar questions in the past:

* http://comments.gmane.org/gmane.comp.version-control.git/197219
* http://comments.gmane.org/gmane.comp.version-control.git/166255

Here's are some instructions recommended for those writing to the Git list 
for the first time: https://gist.github.com/tfnico/4441562

If all else fails, there are other tools than git-svn that can convert 
complicated layouts. See for 
example http://www.catb.org/esr/reposurgeon/reposurgeon.html
 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [git-users] g....@vger.kernel.org Fails

[Thomas Ferris Nicolaisen]

On Wednesday, April 22, 2015 at 2:58:13 AM UTC+2, Evan Flechsig wrote:
>
> There is a "plain text" mode in Gmail that might suffice. Click the grey 
> down arrow and select "Plain text mode". I'm not sure if it will meet the 
> requirements, but it might be worth a shot before you start installing and 
> configuring an additional email client. 
>

Yup, this is what I do when sending mail to the vger list.  The arrow 
button is in the lower right of the compose window.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: There may be a mistake in book Pro Git about rebase

[Thomas Ferris Nicolaisen]

On Tuesday, April 21, 2015 at 2:04:40 PM UTC+2, Deew Lew wrote:
>
> chapter 3.6 :Git Branching - Rebasing 
>

For the record, the URL 
is http://git-scm.com/book/en/v2/Git-Branching-Rebasing
 

>
> the book says:
>
> This basically says, “Check out the client branch, figure out the patches 
> from the common ancestor of the client and server branches, and then 
> replay them onto master.”
>
> I think "then replay them onto master " should be "then replay them onto 
> client start from the common ancestor patch of the client server and 
> master" ?
>
> I test like this: 
> three branches: master server client 
> master:c0
> server:c0< - c1 <- c4 <- c5
> client:c0 <- c1 <- c2 <- c3
> then i did : 
> git rebase --onto master server client 
> the three branches were:
> master:c0
> server:c0< - c1 <- c4 <- c5
> client:c0 <- c2 <- c3
>
> See it ,the client branch changed , not the master !
>
> And i'm weak at English, maybe original is right . And the problem is my 
> understanding :) 
>
>
Your understanding is correct, but I don't think your sentence reads so 
well. I agree with you that the sentence in the book does not sound optimal 
either. Here's what I would write:

Original:

> This basically says, “Check out the client branch, figure out the patches 
> from the common ancestor of the client and server branches, and then replay 
> them onto master.” 


Suggestion:

> This basically says, “Take the client branch, figure out the patches since 
> it diverged from the master branch, and replay these patches in the client 
> branch as if it was based directly off the master branch instead". In other 
> words: transplant the client branch to be based off the master branch 
> instead of server. 


If you want to suggest concrete changes to the book, the right way to go 
about is by submitting pull-requests on its GitHub repository. The file in 
question is located here, so you can just use the Edit button to suggest 
the change:

https://github.com/progit/progit2/blob/master/book/03-git-branching/sections/rebasing.asc

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Extended hook commands

[Thomas Ferris Nicolaisen]

On Friday, March 20, 2015 at 1:48:30 AM UTC+1, Todd Doughty wrote:
>
> Hello all,
>
> I'm part of a research project trying to transition to git.  We have a lot 
> of binary files in our old version and to deal with this, we are thinking 
> of using git-fat .  While this is 
> fairly simple to accomplish, and easy to run with git fat commands, we 
> would like (if possible to use git hooks to automate the process). 
>  However, it seems like this requires hooks attached to git push/fetch.  Is 
> there any development ongoing to extend the commands that can implement a 
> git hook?  Or is there any way to extend this ourselves?  Thanks.
>

Perhaps you could achieve what you want using smudge/clean filters? There's 
a lot to learn from the recent discussions regarding Git handling big 
files: https://news.ycombinator.com/item?id=9343021  - perhaps you'll even 
find some other big-file option that works better for you.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Sync fails due to local and remote commits

[Thomas Ferris Nicolaisen]

On Friday, March 27, 2015 at 1:35:08 AM UTC+1, Jerry Franklin wrote:
>
> Greetings, all.
>
> Sometimes I get the following error message when I try to commit a file:
>
> Sync Failed.  There are both local and remote commits.  Please commit all 
> your changes and then sync again.
>
> I only get this message occasionally.  I can discern no pattern.  In fact, 
> if I click Cancel, then make another change to the same file and re-commit, 
> I do NOT get this error message.  Even more vexing, as near as I can tell, 
> my changes are committed even when I click Cancel, as my co-workers are 
> able to access them.
>
> I don't know whether this is related, but a couple days ago, when I tried 
> to commit a file, Git displayed instead a list of about 10 file conflicts 
> -- on files that not only had I not touched, but that I had not even 
> opened.  They were all files one of my colleagues had just worked on.
>
> This is both vexing and worrisome.  I'm afraid I'm going to lose work.  I 
> can't find anything about this in the documentation, and as near as I can 
> determine my repositories are set up correctly.  
>
> I'm using Git with the online Help authoring tool Flare, running on 
> Windows, which in turn is running on a Mac via Parallels.
>

I think you'd better ask in a forum concerning itself with Flare. "Sync" is 
not a term in the Git vocabulary, and we have no idea what Flare means with 
this error.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: unable to reset working copy changes

[Thomas Ferris Nicolaisen]

On Saturday, April 18, 2015 at 1:04:19 AM UTC+2, todd freed wrote:
>
> Hi, I am unable to reset my working copy changes.
>
> I have two files in this project which differ only in their case:
>  listwise/liblistwise/op/c/c.c
>  listwise/lblistwise/op/C/C.c
>
> It appears that git is being confused by the similar file names.
>
> Is this a bug?
>
> todd@HOSSBEAST /q/Code/fab (master)
> $ git version
> git version 1.9.5.msysgit.1
>
>
>
I'm assuming you're on Windows since  you use a Git built with msysgit.

Windows does not have a case-sensitive file-system. This means that once a 
file comes out of your repository with name 'c', it will overwrite any file 
with the name 'C'.

A quick googling turned out that there are ways to make this work on 
Windows as long as you're inside cygwin at least: 

http://superuser.com/questions/266110/how-do-you-make-windows-7-fully-case-sensitive-with-respect-to-the-filesystem

But it seems non-trivial, and the solution does not work outside the shell. 
I guess you're better off using another OS, or at least a virtual machine 
to work on this particular repository.

If you are in control of this repository, I would heavily recommend 
changing some naming conventions. Having multiple files with the same name 
is asking for trouble and confusion.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Why does "git log -G" works with "regexp-ignore-case" but not with other regexp-related options?

[Thomas Ferris Nicolaisen]

On Friday, April 17, 2015 at 12:13:54 PM UTC+2, Tim Friske wrote:
>
> Hi Thomas,
>
> I followed your suggestion and posted my idea to the Git developer mailing 
> list g...@vger.kernel.org . Let's see what they say.
>
>
Great. If anyone else comes across this post, follow it over at 
http://permalink.gmane.org/gmane.comp.version-control.git/267350 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: gitignore problem

[Thomas Ferris Nicolaisen]

On Tuesday, March 24, 2015 at 8:20:01 AM UTC+1, Philipp Kraus wrote:
>
> but Git reports that the files will be ignored, so there seems to be an 
> error in my configuration. Can anybody help me to create a working 
> gitignore configuration and can explain me my mistake?
>

Perhaps git-check-ignore could help? 
http://git-scm.com/docs/git-check-ignore 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Why does "git log -G" works with "regexp-ignore-case" but not with other regexp-related options?

[Thomas Ferris Nicolaisen]

On Friday, April 17, 2015 at 12:01:10 AM UTC+2, Tim Friske wrote:
>
> Hi,
>
> I wonder why  "git log -G" works with the "regexp-ignore-case" 
> option but not with the other regexp-related options? Wouldn't it be useful 
> to make the "G" option support the following options?
>
>- basic-regexp
>- extended-regexp
>- fixed-strings
>- perl-regexp
>
> Similarly I think it is not very consistent that one cannot combine any of 
> the above options with the "S" but instead have yet another option 
> called "pickaxe-regex" to toggle between "fixed-string" and 
> "extended-regexp" semantics for the argument passed to option "S".
>
> The description of the above options in the git-log(1) manpage of Git 
> version 2.1 do not explicitly say that they *do not* support the 
> "G" and "S" option.
>
> Wouldn't it be nice to have all of the above options collaborate with each 
> other?
>

I've never used much regexp with Git, so I can't comment much. You can send 
the idea to the Git developers by following these 
instructions: https://gist.github.com/tfnico/4441562

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Mercurial

[Thomas Ferris Nicolaisen]

On Tuesday, April 14, 2015 at 2:05:10 PM UTC+2, rusi wrote:
>
> Not exactly a git question... 
> If someone (who prefers git) has to work with a mercurial repo, are 
> one of the git-mercurial bridges preferred or should one just stick to 
> mercurial? 
> If so which bridge is recommended? 
>

A friend of mine who was in a similar position ended up writing his own 
bridge-wrapper which I believe uses the hg -> git bridge that comes with 
Mercurial: https://gist.github.com/rjeschke/868a8cc071bb9f31c29b

Based on his experience, you're probably better off just learning Mercurial 
and using it first-class :)

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: How to uninstall on Mac?

[Thomas Ferris Nicolaisen]

On Sunday, April 12, 2015 at 8:52:27 PM UTC+2, Liz wrote:
>
> Hi - hope someone can advise me how to uninstall Git from a Mac.
>
> I have only just installed Git from here: http://git-scm.com/download/mac
>
> The install instructions contain things that I do not understand, eg.
>
>
>
> *"Step 2 - Restart bash sessions--This include 
> GNU screen sessions, TMUX sessions, etc. If you wish to preserve your 
> precious screen session, just `source /etc/profile` and yo"*
>
> When I went to find clarification from the documentation, I discovered 
> that there is a Mac package that might be geared at the less-technically 
> minded: https://mac.github.com/
>
> So I would like to uninstall the original Git and install the mac.github 
> one.
>
> The original installation came with a file uninstall.sh 
>
> Do I copy and paste the contents of uninstall.sh into the terminal window?
>
>
An .sh file is usually a script you can execute simply by entering its name 
(using Terminal):

cd /path/to/script-folder
./uninstall.sh


-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: GIT Babysteps Timeline chart

[Thomas Ferris Nicolaisen]

On Friday, March 27, 2015 at 1:35:06 AM UTC+1, Ed Schiffer wrote:
>
> Hi y'all,
>
> I'm a designer from Rio de Janeiro, new user of Git, and I was having a 
> hard time understanding all the data transfer of it. To resolve it, I've 
> started making a chart showing exactly what each command does, and it 
> surely cleared my mind. I so present it to you, in a way it may help 
> others, and asking for any advice or suggestions.
>
>
>
This is really nice. Well done! You should post it on a website so we can 
share it on the Twitters and Google+ :) 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Enforcing best practices in GIT Server inherited

[Thomas Ferris Nicolaisen]

On Tuesday, March 10, 2015 at 1:44:09 AM UTC+1, William Lasiewicz wrote:
>
> 3 questions.  Thanks in advance for reading this and to whoever answers my 
> 3 questions.
>
> I am the release manager and we have perforce but we inherited a GIT 
> repository.
> So far about 20 developers access this which is done through creating a 
> ssh key and then adding that key on the machine where we have git.
> Everyone logs in through the same user "git".  Everything is at Git\home, 
> which contains about 25 .git directories.
> Telling who checked in files is on the honor system.
> As far as I can tell,  setting up a git user is done by adding a user to 
> the machine.
> Is this really how this is done for this?
> The first thing I want to do is give everyone their own user name so code 
> access could be restricted to the areas that are needed.  Is this something 
> not done in GIT?
>
>
I think you're better off looking at Git as a protocol or format, rather 
than a hosting tool. There are many hosting tools that support, or are 
built around Git, someone already mentioned Gitolite to you in another 
thread, but there are also more "enterprisey" solutions like GitLab, GitHub 
Enterprise, Atlassian Stash and so on, if you want to pay for more features 
and support.
 

> So far I can't come up with a reason why you would use this instead of a 
> real tool,  but we don't have perforce licenses to add for all these people 
> and the bosses want me to look in to taking our perforce which has 
> substantial automation in builds and deploys and converting it to this tool 
> which looks to be as sophisticated as the old file based stuff that was 
> used 15 years ago.
> What could be benefited from using this?
>

If you're asking about the benefits of Git, well, it depends on what you 
need from your SCM. Git's main strengths are performance (for small/medium 
sized repositories at least), it is distributed, it's very adaptable and 
versatile (can fit all kinds of branching strategies and release 
processes), and it has a huge user-base in the programming industry, being 
undoubtably the most used VCS in open source.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Model SVN file externals in GIT

[Thomas Ferris Nicolaisen]

On Friday, March 6, 2015 at 5:37:41 PM UTC+1, reinhar...@yahoo.de wrote:
>
> Hi,
>
> I've been investigating on how to model SVN Externals - especially file 
> externals - with GIT. We'd like to use GIT if externals are supported.
>
> Possible options are:
> - subtree merges
> - submodules
>
> But none of these seem to support file externals (that is reference a file 
> in your repository with a specified version and copy that file into a 
> subfolder upon checkout).
>

What are you trying to achieve with these file externals? Can you elaborate 
on your use case? Perhaps someone will come up with some ideas then.

If you're trying to keep single large files out of the repository, perhaps 
git-annex  would be something to look at.
 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Slow git pushes, 1 min in pack-objects

[Thomas Ferris Nicolaisen]

Also, have a look 
at https://www.kernel.org/pub/software/scm/git/docs/git-config.html and 
consider tweaking the options that mention memory. I note that you have 
written "pack.windowmemory=100m" wrong (should be uppercase M).

You may also want to try disabling delta compression for some of the binary 
file patterns. 
See https://www.kernel.org/pub/software/scm/git/docs/gitattributes.html for 
more.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Slow git pushes, 1 min in pack-objects

[Thomas Ferris Nicolaisen]

On Thursday, March 5, 2015 at 6:27:29 PM UTC+1, Stephen Morton wrote:
>
>
>> I'm very linux knowledgeable, and becoming more git knowledgeable, so 
> feel free to ask me questions as technical as you want.
>
>
Can you monitor swap/memory usage during the compression? Git can grow 
quite memory hungry if there are big files or big repos involved.

Regarding your filesystem test, I think you'll be better off 
reading/writing many many small/medium files than one big one. I'm no 
expert on this, but I saw in the benchmark I posted earlier that 
performance varies wildly depending on file size.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: release notes 1.9.2

[Thomas Ferris Nicolaisen]

On Friday, February 27, 2015 at 9:00:47 PM UTC+1, saurabh arora wrote:
>
> Release notes for 1.9.2 has a comment for the fixes made from ver 1.9.1  “git 
> mv" that moves a submodule forgot to adjust the array that uses to keep 
> track of which submodules were to be moved to update its configuration.”
>
>  Want to know what the actual issue was?
>

Duplicate thread, already answered here: 
https://groups.google.com/d/msg/git-users/DeDU8o64ITk/QlHY7Lez70kJ 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Slow git pushes, 1 min in pack-objects

[Thomas Ferris Nicolaisen]

On Wednesday, March 4, 2015 at 9:11:45 PM UTC+1, Stephen Morton wrote:
>
> I'm experiencing very slow git pushes. On the order of 1 minute to push a 
> trivial one-line change. When I set GIT_TRACE=1, I see that it seems to be 
> taking a lot of time in the pack-objects phase. 
>
> Others are not seeing this with the same repo, but I'm the only one 
> working in a VM.
> ...
>
> Any ideas?
>
>
Could have something to do with the filesystem. Is your Git repository 
located in a folder shared from the host system, using some kind of virtual 
fs? This could be a performance bottleneck:

http://mitchellh.com/comparing-filesystem-performance-in-virtual-machines

If you can't figure out the filesystem type, please paste us the output of 
`pwd` while the repo is current directory, along with the output of `mount`.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Why no git remote get-url origin?

[Thomas Ferris Nicolaisen]

On Thursday, February 26, 2015 at 5:10:25 PM UTC+1, Andrew Pennebaker wrote:
>
> git-remote offers a set-url subcommand for setting a remote's URL:
>
> $ git remote set-url origin g...@github.com:mcandre/regenboog.git
>
> But I can't find a get counterpart:
>
> $ git remote get-url origin
> error: Unknown subcommand: get-url
>
> Could we add this getter to make the interface more intuitive?
>

I doubt this is necessary, unless you can provide a really good argument 
for expanding an already large Git user interface. As Philip explained, 
get/set is not intuitive in the existing Git operations.

The command `git remote` will list any remotes. If you want more 
information, like the urls, do it verbose:

git remote -v

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: 1.9.2 Release notes

[Thomas Ferris Nicolaisen]

On Friday, February 27, 2015 at 9:00:50 PM UTC+1, saurabh arora wrote:
>
> Release notes for 1.9.2 has a comment for the fixes made from ver 1.9.1  “git 
> mv" that moves a submodule forgot to adjust the array that uses to keep 
> track of which submodules were to be moved to update its configuration.”
>
>  Want to know what the actual issue was?
>

According to one of the maintainer "What's cooking" mails [1]:

* jk/mv-submodules-fix (2014-03-17) 2 commits
>   (merged to 'next' on 2014-03-17 at 7cae3b1)
>  + mv: prevent mismatched data when ignoring errors.
>  + builtin/mv: fix out of bounds write
>  "git mv" that moves a submodule forgot to adjust the array that
>  uses to keep track of which submodules were to be moved to update
>  its configuration. 


If you want more information, try searching the archives [2] for the above 
commit messages. You'll find plenty of discussion there.

[1] 
http://article.gmane.org/gmane.comp.version-control.git/243907/match=what's+cooking+git+mar+2014+06+tue+25
[2] http://dir.gmane.org/gmane.comp.version-control.git

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Installing Git on Windows tech preview of version 10

[Thomas Ferris Nicolaisen]

On Sunday, March 1, 2015 at 3:37:25 PM UTC+1, rick woolley wrote:
>
> For weeks I have been running Windows 10 tech preview and EVERYTHING loads 
> and runs great ... except maybe Git.  When I click on the installer/loader 
> file in explorer it causes explorer to crash. Anyone out the a Windows 10 
> insider, or can anyone comment on ... oh wait, several minutes passed on my 
> i7 for no apparent reason, and then Git did indeed install, it ought to 
> work fine on 10, but I will let you know.
>

Thanks for the pointer. You may want to send further information or future 
issues to the Git for Windows project as instructed here: 
https://gist.github.com/tfnico/4441562  

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Git for XCODE lost uncommitted changes

[Thomas Ferris Nicolaisen]

On Thursday, February 19, 2015 at 3:07:56 AM UTC+1, AJ Manoulian wrote:
>
> I have a project in xcode and I did a time machine backup of everything 
> before I did a clean install of the new OS. I had some changes in my 
> project that I hadn't committed to git. When I brought all my files back on 
> to the computer, all those changes seemed to be lost. When I open up the 
> project file it is the version from when I performed my last commit. Has 
> this happened to anyone else? Does anyone know where to look for the most 
> current file? I'm pretty new to git and xcode, so sorry if I'm leaving 
> anything out. 
>

Changes which have not yet been committed are not tracked by Git at all. So 
it's the same with these changes as any other files you had on the system 
during data migration. This has nothing to do with neither Git nor Xcode. 

Probably you made the changes after the last time machine sync ran, and 
thus you lost all the later changes. Time Machine has a "Back up now" 
option you should run before you do a clean install.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: problem in tortoisegit commit

[Thomas Ferris Nicolaisen]

On Sunday, February 8, 2015 at 10:53:56 AM UTC+1, Mousa Najafi wrote:
>
>
> 
> hi
>
> when i right click and select git commit during to commit show me message 
> for commit you must set user name and email so i select yes in the git 
> setting i select global and type my name and email when click OK show me 
> error "saving config failed (key: "user.name", value:"Mousa Najafi"). 
> libgit2 returned: Failed to create locked file 
> 'c:user/mousa//.gitconfig.lock': the system connot finde the path specified.
>
> please help me to resolve that
>

I've never seen anything like it, but then again, I don't use TortoiseGit 
actively. Perhaps it's a new bug that snuck in as they're switching to use 
libgit2 under the hood. Interesting. You should ask them about it on their 
forum:

https://groups.google.com/forum/#!forum/tortoisegit-users 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Weird integrity problem?

[Thomas Ferris Nicolaisen]

On Thursday, February 5, 2015 at 4:02:00 PM UTC+1, J66st wrote:
>
> The discussion will continue in the Git for Windows group:
> https://groups.google.com/forum/#!topic/msysgit/6XLoSPH26kc
>
>
Great, I'm glad the issue is getting the eyes of some experts. Discussion 
was redirected to their GH issues: 
https://github.com/msysgit/git/issues/312 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Weird integrity problem?

[Thomas Ferris Nicolaisen]

On Thursday, February 5, 2015 at 2:31:54 PM UTC+1, J66st wrote:
>
> Thomas, thanks for your time!
>
> I prepared a new minimal demo package with a shell script to easily 
> install my test-repository and run the test.
> Please download the attached Demo.zip, unpack it in a clean directory.
> It contains two files: DixiLink2.zip (this is my test-repository) and a 
> shell script "weird-git-demo".
> Run the shell script from that directory. Script logging is enabled, so 
> you will see what happens.
> The script will unpack the repository in a subdirectory DemoRepo, do some 
> log, status and diff commands to prove what seems wrong.
>


OK, just to give some quick feedback on this; I see the same problem on my 
Windows when I run your script. I don't see it on OSX. 

If I unzip the DixiLink2.zip using Git Bash's unzip command, I can recreate 
it.

If I use 7zip or Windows built-in extractor to unzip, I don't get the 
problem.

If I do have the problem, a simple "touch dixilinkerr.h" will solve it.

So I think it is a bit like I suspected. Git status somehow converses with 
a part or API of the filesystem (cache) that is not up to speed with the 
actual contents of the file.

I've recreated an OSX directory to my virtual Windows XP, and whereas I can 
see git status doing the wrong thing on Windows, on OSX it's acting fine in 
the same actual directory. Any modification of the file from either system 
will correct the issue, it seems.


-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Weird integrity problem?

[Thomas Ferris Nicolaisen]

On Thursday, February 5, 2015 at 12:44:37 AM UTC+1, J66st wrote:
>
> Adding another snippet to the previous post.
> The failure occurs in the DixiLink2/ directory.
> Then I clone the directory to DixiLink6/ : there everything works fine.
> A directory of tree compare shows not a single difference, in contents nor 
> in attributes.
> (Well, could there be a difference in some alternative stream, this is 
> NTFS.)
> Sorry for the lengthy posts.
>


I skimmed across your findings, and by the looks of it there is something 
really strange going on.

Am I correct in understanding that zipping and then unzipping the directory 
fixed it in one case?

In another example, you fixed it by renaming the file twice?

How do you effectively reproduce the issue in another folder? Just a simple 
copy of the folder where the problem originated?

Since the problem seems somewhat ephemeral, and goes away when doing some 
simple file operations, I tend to think that this has something to do with 
NTFS. Perhaps Git is getting some kind of cached value from the file system 
when doing diff/status, while you get the actual real contents when 
manually reading the file. Can you recreate the issue on different disk 
(USB stick)? What if it's FAT32 formatted?

Perhaps it would be good to have someone from the Git for Windows mailing 
list see if they recognize this problem.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Weird integrity problem?

[Thomas Ferris Nicolaisen]

On Tuesday, February 3, 2015 at 12:06:10 AM UTC+1, Thomas Ferris Nicolaisen 
wrote:
>
>
>
>> *git status* reports "nothing to commit, working directory clean".
>> So there are no staged files and it sees no difference between the file 
>> in my working directory and the latest version in the repository. Although 
>> they ARE different!
>>
>
>
> I can try again on Windows tomorrow to see if there's any difference 
> there. It would be an entertaining idea that there is something like a 
> SHA-collision going on here, but highly unlikely.
>

Just tried on Windows XP, with Git 1.9.4. I see the diff there too, no 
clean working directory. 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Weird integrity problem?

[Thomas Ferris Nicolaisen]

On Monday, February 2, 2015 at 6:27:18 PM UTC+1, J66st wrote:
>
>
> Did you unpack my attached zipfile and tried it it its current state? You 
> will notice the following (unless your git behaves different from mine):
>

Yes, I did unpack it.
 

>
> *git status* reports "nothing to commit, working directory clean".
> So there are no staged files and it sees no difference between the file in 
> my working directory and the latest version in the repository. Although 
> they ARE different!
>

That's odd. Here's the output I get:

❯ git st
On branch master
Changes not staged for commit:
  (use "git add ..." to update what will be committed)
  (use "git checkout -- ..." to discard changes in working directory)

modified:   dixilinkerr.h

no changes added to commit (use "git add" and/or "git commit -a")

It appears the file has dos line-endings, which makes the whole file appear 
changed on my system, but even if I convert to unix line endings, there's 
still an actual diff:

❯ git diff
diff --git a/dixilinkerr.h b/dixilinkerr.h
index daeca6b..59f4c35 100644
--- a/dixilinkerr.h
+++ b/dixilinkerr.h
@@ -2,8 +2,8 @@
 
 #pragma once
 
-#ifndef CAT_DIXILINK_ERR
-#define CAT_DIXILINK_ERR 2000
+#ifndef EXCPCAT_DIXILINK
+#define EXCPCAT_DIXILINK 2000
 #endif
 
 #ifndef __DixiLinkErr_H_INCLUDED__
@@ -16,7 +16,7 @@
 #ifndef BASE_DIXILINK_ERR
#define BASE_DIXILINK_ERR 0
#ifdef USE_DEFAULT_CAT
-   #define BASE_DIXILINK_ERR CAT_DIXILINK_ERR
+   #define BASE_DIXILINK_ERR EXCPCAT_DIXILINK
#endif
 #endif
 
@@ -135,4 +135,4 @@ enum EDixiLinkErr
//!< Marks end of list
 };
 
-#endif
\ No newline at end of file
+#endif
   


I can try again on Windows tomorrow to see if there's any difference there. 
It would be an entertaining idea that there is something like a 
SHA-collision going on here, but highly unlikely.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git corruption 1.9.4 <> 2.2.2

[Thomas Ferris Nicolaisen]

On Tuesday, January 27, 2015 at 10:50:22 AM UTC+1, Zsolt Ero wrote:
>
> Git corrupts all repositories, even by calling simple read-only commands 
> (status, etc.), in the following scenario:
>
> 1. Create or commit to a repo under Git 2.2.2 (system spec: OS X 10.9.5, 
> homebrew latest git)
> 2. Go to the same folder from a Windows computer (Windows 8.1, git version 
> 1.9.4.msysgit.0, HFS+ for Windows 8 Free for accessing the partition)
> 3. By simply reading the branch information in cmder using msysgit, the 
> repo is corrupted.
>
> This is what it says at the bottom when simply running "git":
> error: object file .git/objects/dd/881c8dff481756ddba39034b7ba3c5cbeb1ec3 
> is empty
> error: object file .git/objects/dd/881c8dff481756ddba39034b7ba3c5cbeb1ec3 
> is empty
> fatal: loose object dd881c8dff481756ddba39034b7ba3c5cbeb1ec3 (stored in 
> .git/objects/dd/881c8dff481756ddba39034b7ba3c5cbeb1ec3) is corrupt
>
> "git status" also writes the following lines:
> error: object file .git/objects/e8/45566c06f9bf557d35e8292c37cf05d97a9769 
> is empty
> fatal: loose object e845566c06f9bf557d35e8292c37cf05d97a9769 (stored in 
> .git/objects/e8/45566c06f9bf557d35e8292c37cf05d97a9769) is corrupt
>
> I don't know if it matters that it's a Windows computer, but I can 100% 
> reproduce the issue like this.
>
> I'll be checking if it's related to the file system driver by making a zip 
> from the clean 2.2.2 repo.
>
>
>
Serious Git bugs should go to the respective mailing list [1].

I just tried recreating this on OSX with Git 2.2.1 (from homebrew). I then 
opened the same repository on a Windows virtual machine (Windows XP on 
VMWare Fusion) with Git 1.9.4. Works fine.

I then tried upgrading to Git 2.2.2, repeating the operation. Still no 
problem.

Can you reproduce it when initializing a repo and just committing one file 
in it?

Can you recreate it when moving/copying/cloning the repo to a local NTFS 
disk on Windows?

[1] https://gist.github.com/tfnico/4441562

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git rebase -i HEAD~2 does not work

[Thomas Ferris Nicolaisen]

On Wednesday, January 28, 2015 at 10:37:47 PM UTC+1, akulak...@gmail.com 
wrote:
>
> Hi, I get this error in mac os x 10.10:
>
> git rebase -i HEAD~2
>
> Unknown command: Fixed #24052 -- Added note on ``dependencies`` property
>
That's strange.. Please provide us with the outputs of the following 
commands:

git --version
which git
echo $SHELL

How did you install Git? Are you running any special aliases or wrappers 
around it?

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Namespaces and HTTP

[Thomas Ferris Nicolaisen]

On Saturday, January 24, 2015 at 2:40:39 AM UTC+1, Thomas Schulte wrote:
>
> Hello,
> I'm struggling with a repository that contains some namespaces.
>

Duplicate 
thread: https://groups.google.com/d/topic/git-users/m8Br86WNhng/discussion 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [git-users] Namespaces, git-remote-ext and http

[Thomas Ferris Nicolaisen]

On Monday, January 26, 2015 at 6:24:51 PM UTC+1, Thomas Schulte wrote:
>
> Am 26.01.2015 um 17:49 schrieb Thomas Ferris Nicolaisen  >:
>
> On Saturday, January 24, 2015 at 2:40:39 AM UTC+1, Thomas Schulte wrote:
>>
>> Hello, 
>>   
>> I'm struggling with a repository that contains some namespaces. 
>> The following basically shows what I'm trying: 
>>
>> --8<-- 
>> $ git init --bare myrepo.git 
>> Initialized empty Git repository in d:/test/myrepo.git/ 
>>   
>> $ git clone myrepo.git/ 
>> Cloning into 'myrepo'... 
>> warning: You appear to have cloned an empty repository. 
>> done. 
>>   
>> $ git clone ext::'git --namespace=project-one %s d:/test/myrepo.git' 
>> myrepo-project-one 
>> Cloning into 'myrepo-project-one'... 
>> warning: You appear to have cloned an empty repository. 
>> --8<-- 
>>   
>> As you can see, this is quite fine :-) 
>> It's also no problem to place this repository on a http server: 
>>
>> --8<-- 
>> $ git clone http://server/path/to/myrepo.git tools 
>> Cloning into 'tools'... 
>> Unpacking objects: 100% (6/6), done. 
>> --8<-- 
>>
>> But why does the following go wrong? 
>>
>> --8<-- 
>> $ git clone ext::'git --namespace=project-one %s 
>> http://server/path/to/myrepo.git' mytest 
>> Cloning into 'mytest'... 
>> fatal: 'http://server/path/to/myrepo.git' does not appear to be a git 
>> repository 
>> fatal: Could not read from remote repository. 
>>
>> Please make sure you have the correct access rights 
>> and the repository exists. 
>> --8<-- 
>>
>> URL checked multiple times, it's exactly the same as used in "git clone 
>> http...". 
>> Any hints? 
>>
>>
> What backend is behind that http server?
>
> The docs here [1] state:
>
> The smart HTTP server, git-http-backend[1] 
>> <http://git-scm.com/docs/git-http-backend>, will pass GIT_NAMESPACE 
>> through to the backend programs; see git-http-backend[1] 
>> <http://git-scm.com/docs/git-http-backend> for sample configuration to 
>> expose repository namespaces as repositories.
>
>
> Are you sure your server is passing this parameter onwards to the backend?
>
> [1] http://git-scm.com/docs/gitnamespaces
>
>
> The server is a Microsoft TFS Server 2013. I'm trying to find a way to 
> provide one repository that contains multiple tools for various projects. 
> But each of these projects only uses a subset of these tools and I want to 
> limit the size of the clones. 
>
> Thanks.
>
>
OK, then I think it's safe to assume that TFS Server does not yet support 
namespaces.

Namespaces are primarily for separating repositories that contain much of 
the same objects. Or branches of the same software that you want to keep 
isolated from eachother. Given your usecase, I think you'd be better off 
just using a different repository for each tool.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Namespaces, git-remote-ext and http

[Thomas Ferris Nicolaisen]

On Saturday, January 24, 2015 at 2:40:39 AM UTC+1, Thomas Schulte wrote:
>
> Hello, 
>   
> I'm struggling with a repository that contains some namespaces. 
> The following basically shows what I'm trying: 
>
> --8<-- 
> $ git init --bare myrepo.git 
> Initialized empty Git repository in d:/test/myrepo.git/ 
>   
> $ git clone myrepo.git/ 
> Cloning into 'myrepo'... 
> warning: You appear to have cloned an empty repository. 
> done. 
>   
> $ git clone ext::'git --namespace=project-one %s d:/test/myrepo.git' 
> myrepo-project-one 
> Cloning into 'myrepo-project-one'... 
> warning: You appear to have cloned an empty repository. 
> --8<-- 
>   
> As you can see, this is quite fine :-) 
> It's also no problem to place this repository on a http server: 
>
> --8<-- 
> $ git clone http://server/path/to/myrepo.git tools 
> Cloning into 'tools'... 
> Unpacking objects: 100% (6/6), done. 
> --8<-- 
>
> But why does the following go wrong? 
>
> --8<-- 
> $ git clone ext::'git --namespace=project-one %s 
> http://server/path/to/myrepo.git' mytest 
> Cloning into 'mytest'... 
> fatal: 'http://server/path/to/myrepo.git' does not appear to be a git 
> repository 
> fatal: Could not read from remote repository. 
>
> Please make sure you have the correct access rights 
> and the repository exists. 
> --8<-- 
>
> URL checked multiple times, it's exactly the same as used in "git clone 
> http...". 
> Any hints? 
>
>
What backend is behind that http server?

The docs here [1] state:

The smart HTTP server, git-http-backend[1] 
> , will pass GIT_NAMESPACE 
> through to the backend programs; see git-http-backend[1] 
>  for sample configuration to 
> expose repository namespaces as repositories.


Are you sure your server is passing this parameter onwards to the backend?

[1] http://git-scm.com/docs/gitnamespaces

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: best open source git diff viewer

[Thomas Ferris Nicolaisen]

On Sunday, January 18, 2015 at 9:40:30 PM UTC+1, Jatin Sahu wrote:
>
> Could someone please tell me, which is the best open source diff tool 
> having supporting multiple file formats like beyond compare. How to 
> configure git to use such tools.


It's very hard to say which is the "best" one, but there are a number of 
options that you can sort by license on Wikipedia: 
http://en.wikipedia.org/wiki/Comparison_of_file_comparison_tools

Once you have found a tool you're happy with, google for "[tool-name] 
configure git mergetool difftool". For example:

http://lmgtfy.com/?q=kdiff3+configure+git+mergetool+difftool

 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Weird integrity problem?

[Thomas Ferris Nicolaisen]

On Friday, January 23, 2015 at 6:11:54 PM UTC+1, J66st wrote:
>
> Working with git now for a few weeks, and trying to move my Visual 
> SourceSafe repositories to git.
>
> I am encountering a weird problem. I have set up a small example repo (see 
> the attached zipfile).
>
> Containing a single file dixilinkerr.h. There are three versions of this 
> file.
>
> v1: Contains a #pragma message line, and a #define CAT_DIXILINK_ERR
> v2: Edit: the #pragma message line is removed.
> v3: Edit: the symbol CAT_DIXILINK_ERR is replaced by EXCPCAT_DIXILINK
>
> Watching the repository and working directory in the current state:
>
> *git status*:
> nothing to commit, working directory clean.
>
> *git log* shows 2 commits: v1 and v2.
>
> *gitk* shows 2 commits:
> The first one in the root commit is v1.
> But the second one shows different result depending on the chosen view:
> *gitk Patch* view: the last commit is a patch from v1 to v3! (So 
> bypassing v2!!)
> *gitk Tree* view: the last commit is v2
> THIS IS WEIRD!
>
>
> *git show HEAD:dixilinkerr.h*
> Shows v2.
>
> So the strange thing is:
> v3 is not really in the repository. If you check out the latest version 
> you get v2.
>
> v3 is in my working directory.
> But git does not see any changed files, and the index is empty. Git commit 
> won't add the file.
>
> So the file is kind of halfway checked in...
>
> I ran into this problem several times when checking in thousands of files 
> using a script.
> It seems to happen only when the file size of the two last versions of a 
> file are equal but the contents differ.
> In the above example the v2 and v3 versions of the file have the same size.
>
> I am using msysgit version 1.9.0 on Windows 7.
>
> Is this a bug or am I missing something?
>


As far as I can see, you have only committed this file twice. So there is 
no v3, there is only the uncommitted/unstaged changes in your working 
directory. gitk displays uncommitted changes with the message "Local 
uncommitted changes, not checked in to index".

So there is no v3. You have to commit first. Read 
through http://git-scm.com/book/en/v2/Getting-Started-Git-Basics and things 
should become a bit clearer.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: can git log --graph indicate which branch HEAD is on?

[Thomas Ferris Nicolaisen]

On Friday, January 23, 2015 at 4:00:00 PM UTC+1, Wolfram Volpi wrote:
>
> Is there a way git log --graph could indicate the current branch?
>

I might be side-stepping the question here, but what I personally use is 
some form of git-prompt [1] which extends the PS1/prompt to always show the 
current branch of a repository:

[~/temp]$ cd foo
[master][~/temp/foo]$  
[master][~/temp/foo]$ git checkout dev
Switched to branch 'dev'
[dev][~/temp/foo]$ 

Lots of shell enhancements like oh-my-zsh [2] come with stuff like this 
built in. 

For Windows users there is posh-git [3].

[1] https://github.com/git/git/blob/master/contrib/completion/git-prompt.sh
[2] https://github.com/robbyrussell/oh-my-zsh
[3] https://github.com/dahlbyk/posh-git

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: gitk showing tags n instead of actual tags

[Thomas Ferris Nicolaisen]

On Tuesday, January 20, 2015 at 9:51:19 PM UTC+1, gary lucas wrote:
>
> See attachment, 
>
> $ git version
> git version 1.9.5.msysgit.0
>
> I'd like to be able to see the actual tags, this didn't happen under a 
> previous version.
>
> Any ideas?
>
>
>
Duplicate 
thread? https://groups.google.com/d/msg/git-users/UF-WDrbarmQ/nKmWHmnksssJ 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: gitk recent msysgit build no longer shows full tag text. Instead it shows '2 tags' or 'n tags'

[Thomas Ferris Nicolaisen]

On Tuesday, January 20, 2015 at 10:27:52 AM UTC+1, gary lucas wrote:
>
> See attachment.
>
> I updated recently and this behaviour is driving me crazy, our CI 
> environment tags our builds and it's hard to see at a glance in gitk what 
> happened where.
>
> $ git version
> git version 1.9.5.msysgit.0
>
> Is there a way to set this back to the prior default of not collapsing the 
> tags down into something useless?
>
>
Note that if you click the tags, it will display their contents in the pane 
below.  

If you give more space to the rendering, that is, a smaller font, or 
enlarge the pane (they're draggable on the edges), it will display the tags 
that can fit. A reload (Shift+F5) may be necessary to re-render the view.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: HELP

[Thomas Ferris Nicolaisen]

On Tuesday, January 20, 2015 at 6:21:29 AM UTC+1, Stephanie Gonzalez wrote:
>
> I'm looking for someone in sanbernardino county ca to help me figure out 
> who is using my email and this site to clone my phone and spy on me I'm not 
> computer smart at all and this has been going on for 2 years, phone after 
> phone laptop after laptop ,apple and android and 4 home computers I have 
> tones of files and I need help to read computer codes  and help me catch 
> this person and I will pay whoever can give me answers .I know of email 
> address that have accounts and so much data saved and printed the police 
> won't help me and I don't know we're else to turn .I found this accout 
> today which this email was made less then four days ago with the new phone 
> I just got at same time it has files that track and who knows what else 
> .I'm a stay at home mom and I do not make files or anything on or about 
> anything these files aren't mine please help anyone who's has sometime and 
> wants to earn some cash .I would be so thankful to you


Not sure if this is a legit call for help, but in any case this mailing 
list is not the place to ask. Call the police if you think you've been the 
victim of a crime.

(partly answering this just so my google-group filter won't show this 
thread as being unanswered.)

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git svn fails to dcommit multiple commits

[Thomas Ferris Nicolaisen]

On Thursday, January 22, 2015 at 2:02:26 PM UTC+1, Bicen Zhu wrote:
>
> Hi All,
>
> I have met the exactly same problem as describled at 
> http://stackoverflow.com/questions/4238876/git-svn-fails-to-dcommit-even-after-clean-checkout
>  
> 
>
> But the question was asked 4 years ago, the answer was related http proxy, 
> but i didn't set http_proxy,
>
>  
Also note:  The answer write-through proxying described in the answer has 
nothing to do with setting up http_proxy on your client side. It's about 
clustering several SVN servers together for distributing across networks 
for performance reasons, and I believe it is transparent to the end user. 
Are you using a large corporate SVN server which is spread across 
offices/countries? It could be that the administrators have set 
up write-through proxying without you knowing.

Also note that not much has changed in git-svn the last 4 years, so 
problems that existed back then could still exist today.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git svn fails to dcommit multiple commits

[Thomas Ferris Nicolaisen]

On Thursday, January 22, 2015 at 2:02:26 PM UTC+1, Bicen Zhu wrote:
>
> Hi All,
>
> I have met the exactly same problem as describled at 
> http://stackoverflow.com/questions/4238876/git-svn-fails-to-dcommit-even-after-clean-checkout
>
> But the question was asked 4 years ago, the answer was related http proxy, 
> but i didn't set http_proxy,
>
> i use the latest git version v2.2.2
>
> Anyone comment?
>
>
>
Is your SVN server modifying the incoming commits somehow, as one of the 
commenters ask? 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: submodules or something else?

[Thomas Ferris Nicolaisen]

On Wednesday, January 14, 2015 at 8:23:02 PM UTC+1, Artem Kula wrote:
>
> Hello, I have a few libraries I using in more then one project of 
> Codeigniter. Every project using own git container and own git backup. They 
> are not in branch of one project.
>

This doesn't make much sense in Git terminology. There are no containers, 
projects or backups in Git speak. And all Git repositories have at least 
one or more branches. I'm not familiar with Codeigniter, so you'll have to 
explain it in more portable terms.

>
> Sometimes I need to update library and wants this library automaticaly 
> updated on every of my project in git after commiting.
>
> How can I organise this?
>

Submodules usually mean that you manually upgrade the submodule in the 
containing Git repository whenever you want to update. To the best of my 
knowledge, you cannot have an automatic update (unless this has since 
become a feature in submodules).
 

> I has tried with git submodules, but something goes wrong with location of 
> library files and updating on every project, may be it's (submodules) not 
> working properly.
>

Please post the commands you performed, along with their output, and what 
you expected.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: CVS to GIT Migration

[Thomas Ferris Nicolaisen]

On Wednesday, January 14, 2015 at 7:03:33 AM UTC+1, Akshaybar Kumar Singh 
wrote:
>
> Hi All,
>
> I am new to GIT repository and using CVS. 
> Now i am moving to use GIT, and going to migrate all the stuff at CVS 
> should be at GIT along with history details.
>
>
> Kindly help me for migrating CVS to GIT. suggest me any tool which can do 
> for me. 
>
 
>

I just mentioned this in another thread as well: reposurgeon has a history 
[1] of migrating projects from (among other things) CVS to Git.

[1] http://www.catb.org/~esr/reposurgeon/features.html

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git svn - experience working with large repositories

[Thomas Ferris Nicolaisen]

On Monday, January 19, 2015 at 9:26:43 AM UTC+1, Johannes Mueller wrote:
>
> Hi,
>
> I would be interested in how large your git svn repositories are. I am 
> working with a large svn repository and am interested in comments where the 
> feasibility of working with git stops.
> How many commits does your svn repository have? How large is it GB-wise? 
> How many people author commits?
> I will share my experience once I (successfully?) get it to work. Though I 
> am not sure if it will block access of our limited bandwidth so I will try 
> to clone the repository over night or with limited bandwidth.
>

Some years ago, I was working pretty productively with a 8 GB big repo, 
around 100k revisions, 30 different authors. The initial git-svn clone took 
a couple of days to complete (over LAN connection to SVN server), but after 
that it was fine. I used a Mac, but I know people on Windows quickly run 
into performance problems.

Once you run into the limitations of git-svn, you can consider reposurgeon 
[1], or the commercial offerings from SubGit [2].

[1] http://www.catb.org/esr/reposurgeon/
[2] http://www.subgit.com/

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: folder with same name as a deleted branch blocks git fetch?

[Thomas Ferris Nicolaisen]

On Wednesday, January 14, 2015 at 5:07:04 PM UTC+1, Lars Hartviksen wrote:
>
> A developer made a branch called 2015/January/Feature1234.
> It was later decided that coding should be split into subtasks and 
> branches like this:
>
> 2015/January/Feature1234/master
> 2015/January/Feature1234/Task1
> 2015/January/Feature1234/Task2
> 2015/January/Feature1234/Task3
>
> so we did this:
>
> git branch -d 2015/January/Feature1234
>
> git push origin :2015/January/Feature1234
>
> and we created the new branches from DEV.
>
>
Generally: You cannot have colliding branch names, and you cannot collide 
with branches' "folder" names. Prefixes (like 2015/) are basically like 
folders on the file system. You cannot have a file called foo and a 
directory called foo/ in the same place.
 

> Now what happened is I couldn't fetch from origin, because of an error 
> message: 2015/January/Feature1234/master could not be created. Probably 
> because there seemed to be a branch called 2015/January/Feature1234.
>

Not sure if you had this branch checked out and tracked already in your 
local repository. That would require you too manually delete it first.
 

> So what I did was try deleting the branch called 2015/January/Feature1234 
> from origin in Sourcetree. This gave me a "doesn't exist" error message. 
> But after that it seems sourcetree realized the branch was deleted and I 
> could fetch from origin and everything seemed to be ok.
>

My guess here is that SourceTree was out of date for a small period in 
which it still displayed the branch.
 

>
> What happened here? Any easy / proper way to fix this?(expecting a few 
> mails from other developers who can't fetch from origin)
>

The problem is that the fetch operation is both trying to delete file and 
create a directory with the same name, and this is going down in the wrong 
order. It's easy to reproduce:

First: Take a repo, create a branch with the old pattern:

cd foo #some git repo
git branch 2015/January/Feature1234

Now, clone the repo
cd ..
git clone foo foo2
cd foo2
git branch -a
* master
  remotes/origin/2015/January/Feature1234
  remotes/origin/HEAD -> origin/master
  remotes/origin/master

So far, so good. Now delete the old branch and create the new one in the 
original repo (foo):

git branch -d 2015/January/Feature1234
Deleted branch 2015/January/Feature1234 (was 8deac45).
git branch 2015/January/Feature1234/master 

And then fetch in foo2:
git fetch
error: 'refs/remotes/origin/2015/January/Feature1234' exists; cannot create 
'refs/remotes/origin/2015/January/Feature1234/master'
error: Cannot lock the ref 
'refs/remotes/origin/2015/January/Feature1234/master'.
>From /Users/tfnico/temp/foo
 ! [new branch]  2015/January/Feature1234/master -> 
origin/2015/January/Feature1234/master  (unable to update local ref)
error: some local refs could not be updated; try running
 'git remote prune origin' to remove any old, conflicting branches

So, things are messed up. I try running a prune, as Git advises in the last 
line above:

git remote prune origin
Pruning origin
URL: /Users/tfnico/temp/foo
 * [pruned] origin/2015/January/Feature1234

Now try fetching again. It should work:

[master][~/temp/foo2]$ git fetch
>From /Users/tfnico/temp/foo
 * [new branch]  2015/January/Feature1234/master -> 
origin/2015/January/Feature1234/master

Long story short, you broke the rules of Git, and SourceTree did not show 
you the correct error message, nor did it show the solution. 

Morale of the story is, when running into Git problems, try to recreate it 
in a simple repository using the command line tools.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [git-users] Enforce C# code style with git hooks

[Thomas Ferris Nicolaisen]

On Wednesday, January 14, 2015 at 3:47:17 PM UTC+1, Peter van der Does 
wrote:
>
>
> I don't know uncrustify, how can you tell with uncrustify if code is not 
> in your preferred code style? 
>
>
>
There are a myriad of tools like this: 
http://en.wikipedia.org/wiki/List_of_tools_for_static_code_analysis

They're basically some executable that run through a path of source files, 
parse them, do some analysis, and spit out a report that is either read by 
humans or used by some machine/script to notify the programmer that the 
code does not conform to their configured standards.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Enforce C# code style with git hooks

[Thomas Ferris Nicolaisen]

On Tuesday, January 13, 2015 at 8:35:05 PM UTC+1, Marko Paloski wrote:
>
> Hello,
>
> I want to make i script that check the code style in the commit, so if 
> don't matches, it will not allow the developers to commit the code.
>
> The script will be in pre-commit local or in pre-recieve on the server. I 
> find something about uncrustify :
> http://www.itk.org/pipermail/insight-developers/2010-September/015333.html
> http://uncrustify.sourceforge.net/ 
> 
> https://github.com/bengardner/uncrustify
>
>
> But i don't know how to implement. Yes it can be bash or powershell script.
> Thanks,
>

You may want to have a look server side hooks here: 
http://git-scm.com/book/en/v2/Customizing-Git-Git-Hooks

I would be wary of doing these kinds of checks in a hook the way you plan. 
I like more the approach where code is suggested using pull- or change 
requests in pushed branches - and then automatic build agents run 
stylecheckers against those branches, before they are allowed to be merged 
to the main branch.

Alternatively, I would try motivating developers to run style-checkers 
locally before pushing (many IDEs can do this continuously). Putting 
obstacles in way of developers is not a good motivator. It's better to 
improve tools and culture to make their job easier.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Unable to Clone Repository With Files that Contain ~

[Thomas Ferris Nicolaisen]

On Wednesday, January 7, 2015 at 5:14:41 PM UTC+1, Mike Rawdon wrote:
>
> If I create a repository containing a file with a ~ in it in a 
> subdirectory(note it only fails from within a subdirectory, not the root 
> directory) I am unable to fully clone the repository using Git for Windows 
> 1.95.  If I downgrade to Git for Windows 1.90, I am able to clone it 
> successfully.
>

I'm assuming this was discussed here: 
https://groups.google.com/d/topic/msysgit/fK-OFZvpWUA/discussion

For reference, bugs specific to Git for Windows should go on their mailing 
list.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [git-users] Re: newbie: when, exactly, do I "git init" ???

[Thomas Ferris Nicolaisen]

On Thursday, December 4, 2014 3:08:27 AM UTC+1, derelict wrote:
>
>  Hi Thomas;
> Thank you for your in-depth response!
> So generally, it sounds like under git I would expect to have *many* 
> different repositories, unlike Svn where I had one single repository which 
> contained all the projects that we'd ever worked on.  Do I understand 
> correctly?  
>

Absolutely. In our team of 5 devs, we have about 100 repos. Most of them 
are little libraries or utilities that we use across different projects - 
some of these are only needed for one project, but it made sense for them 
architecturally to have their own repository.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [git-users] Re: newbie: when, exactly, do I "git init" ???

[Thomas Ferris Nicolaisen]

> However, acting on your advice (which I think is excellent in this case),
> I'll go back to basics and forget about the sharing issue for now.
> From my ongoing reading, I *think* the way Git intends to work is that each
> project is a separate "repository"; am I seeing this right??  Given my
> experience with CVS and Svn, I view the "repository" as the container that
> contains the entire history of all projects that I have ever worked on.  I
> am suspecting that this world-view will not work with Git...

Yup, you're on the right track there.

> But if this Git world-view is correct, where do I keep a central collection
> of all of my projects (what I call "the repository"), so I can (for example)
> back it up and have everything secured somewhere???  It seems like Git wants
> to make all of my projects completely unrelated to each other - but they're
> not!!  They are all projects that I'm responsible for, and need to be able
> to collect and control in various ways...

You should use a Git repository manager, like GitBlit, GitLab or
GitHub (there's myriad of different ones, but I won't get into how to
pick the right one here). Many of these offer to organize repositories
into some sort of group, which can help you or your team keep track.

Then when it comes down to efficiently dealing with multiple
repositories, here are some ideas:

* Group repositories locally by bundling them in normal directories
* Use gitslave, or myrepos to check out many adjacent repositories in
one command
* Consider git submodules if you want to add the sources from one repo
into another (library code that does not change often). If submodules
don't work for you, try subtree.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git fsck finds a missing blob but doesn't say why

[Thomas Ferris Nicolaisen]

On Sunday, November 30, 2014 4:34:42 AM UTC+1, Dale Worley wrote:
>
> (I am using git version 1.8.3.1.) 
>
> Some references for git-fsck are like this: 
>
> 
> https://git.help.collab.net/entries/22580428-Recovering-from-repository-corruption
>  
>
> $ git fsck --full 
> broken link fromtree 2d9263c6d23595e7cb2a21e5ebbb53655278dff8 
>   toblob 4b9458b3786228369c63936db65827de3cc06200 
> missing blob 4b9458b3786228369c63936db65827de3cc06200 
>
> that is, when git-fsck reports a missing blob, it tells what object 
> contained a link to the blob. 
>
> However, when I run git-fsck, all I get is the missing blob's hash: 
>
> $ git fsck --full 
> Checking object directories: 100% (256/256), done. 
> Checking objects: 100% (4038/4038), done. 
> missing blob 2322b10fb1cdab6fb38a9ad1c2042285939553b6 
> missing blob 955c487443b7c11202aca66a0fcb0a1b29ce4645 
> missing blob fb67076078ca5f64f8eb6a29481f1ab4ff9bca5d 
> missing blob a7edc7ed7a924b4e2aca62d3462fe1fdfd510d5b 
>
> It seems to me that the first style is much more useful, and indeed it 
> should be augmented so that it can show the complete traceback back to 
> a "root" of the object structure (typically a "refs" value).  Is there 
> any way to activate the first style? 
>
>
Just for the record, I scanned through the git mailing list looking for 
traces of these two modes, but couldn't find anything. You could try 
reading the fsck source...

There are some perl scripts around on StackOverflow that can help you find 
the same information.

My guess is that this is some undocumented feature which is only enabled in 
some particular cases.. 

Here's what happens when I try removing some random object in my Git repo 
(using Git 2.0.0):

git init; git add .;git commit -m "inital"
rm .git/objects/00/79c7f07eb4bfbc08cc21f3324e7997e0b05415
git fsck

Checking object directories: 100% (256/256), done.
broken link fromtree c862bb1ab77b0d118009961c1f06d8954f98db8f
  totree 0079c7f07eb4bfbc08cc21f3324e7997e0b05415
missing tree 0079c7f07eb4bfbc08cc21f3324e7997e0b05415 

Here I tried removing a blob instead:

rm .git/objects/ac/9ea1b673c0bb4501415550d492474700d6bd2f
git fsck   
  
Checking object directories: 100% (256/256), done.
missing blob ac9ea1b673c0bb4501415550d492474700d6bd2f

So, recreated the issue, in any case. The only difference was that I found 
the blob-to-delete by looking with ls-tree, while the first one was totally 
random.

I'd keep trying around a bit, see if you can find a pattern, and then ask 
on the git-dev list.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: newbie: when, exactly, do I "git init" ???

[Thomas Ferris Nicolaisen]

On Monday, December 1, 2014 9:30:53 PM UTC+1, derell...@gmail.com wrote:
>
> Okay, I've got a basic *working* directory that I want to use for all of 
> my git projects: c:\SourceCode\git_work
>
> Under c:\SourceCode\git_work, I have the following directories:
>
> + library_code  # general-reference library code that is used by 
> other projects
> + winagrams# project which will access library_code via 
> subtree, when I learn how
> + wfontlist# project which will access library_code via 
> subtree, when I learn how
> + ndir  # project which will *not* access library_code
> + files  # other single-file functions which will 
> *not* access library_code
>
> What is the correct way to build all of this structure??
> I initially ran "git init" only in the c:\SourceCode\git_work, then used 
> "git add ." and "git commit" in each of the subdirectories, to create the 
> repository.
>

Git init will initialize one Git repository in the current directory. So 
what you did there was to make git_work into a Git repository, followed by 
5 commits that added stuff to the repository.
 

> However, many of the online "newbie" references that I have looked at, 
> appear to run "git init" in each subdirectory...  That seems a little odd 
> to me; using the terminology that I've seen, it seems like each project 
> would be a separate "repository", but I actually want one repository that 
> contains *all* of my projects... am I mis-understanding exactly what 
> "repository" means??
>

There are cases where it's not clear what is the right structure of your 
repository or repositories. I suggest the following rules of thumb:

* Things that belong together go in the same repo (tightly coupled modules 
for example)
* Things that have nothing to do with each other go in separate repos (two 
unrelated projects)
* Things that follow the same release cycle and versioning go in the same 
repository (two related projects that are shipped together)
* Things that are built separately, or in isolation, go in separate repos

Now, again, there might be cases where the above guidelines are in 
conflict, and to that I'll say generally: When in doubt, leave stuff in the 
same repo. If need be, you can always split them again in the future (but 
try keeping them in distinct subdirectories to make this work easier).
 

> I've been using Subverson for years, but I'm new to git... that is 
> probably causing some confusion for me.
> I would be grateful for *any* clarification that I could get here!!!
>

I'd add to the above: Any "project" in SVN that has the trunk/branches/tags 
structure typically becomes one Git repo.

By the way, if you want to start over with Git-ifying your directories, all 
you have to do is to delete the .git folders created by the git init 
command. Warning: This will remove all the history/commits you made, but 
will leave all other files in place so you can start over cleanly.

And to be clear: I cannot say whether your git_work should be one repo or 
not. You'll have to say more about their relationships and purposes.

Finally, I suggest to put your subtree plans on hold until you've fully 
grokked what they are and when to use them.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [git-users] any suggestions for pruning all upstream branches after a github fork?

[Thomas Ferris Nicolaisen]

On Tuesday, September 30, 2014 8:49:11 PM UTC+2, Sam Roberts wrote:
>
>
> So what I want to do in effect is: 
>
> for each $branch in remote "origin", 
>   if head(origin/$branch) is in ancestry of upstream/$branch 
> git push --delete origin $branch 
>   end 
> end  

 
>

How about this approach:

(The thing is to avoid GitHub's Fork button, instead create it manually 
with only the branches you want)

* git clone github:joyent/node
* cd node
* git branch #one branch: master
* git remote add sam github:sam/node #where node is a freshly initialized 
GH repo
* git push sam master # pushes only one branch to your GitHub repo

Now you can clone the one-branch repo:
* git clone github:sam/node
* cd node
* git remote add joyent github:joyent/node
* git fetch joyent

Now you have your fork with just one branch in it. The branches from joyent 
are available in their remote, and I believe they will be updated with 
every `fetch joyent` that you make. Note that once you start creating 
tracking branches, you need update them manually or use a script (like 
git-up ). 

If you want to see all remote branches in your github repo, but not see the 
50 remote joyent branches, you can use

git branch --list -a origin\* 

(might want to make an alias for it).

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: any suggestions for pruning all upstream branches after a github fork?

[Thomas Ferris Nicolaisen]

On Tuesday, September 30, 2014 7:35:05 AM UTC+2, Sam Roberts wrote:
>
> Basically, I'd like to delete every branch in my fork of an upstream 
> repo that is the same as an upstream branch. 
>

Not sure if I understood correctly here. You want to delete all local 
branches which are already remote branches?

I don't understand why you want this though. Local branches that track 
remote branches are sort of just aliases for the remote branches. You don't 
save any space or anything by removing them all. Usually, you only create 
local branches manually when you want to work with a remote branch, or 
create one, and once the work is done, you manually delete it once it has 
been merged. If you have a lot of these merged branches lying around and 
you want to clean them up, there are scripts like (*use with caution, 
scripts may permanently delete important stuff*):

* https://github.com/arc90/git-sweep 
* http://mloughran.github.io/git-cleanup/  
* https://gist.github.com/Espenhh/3804251
* (google for more actively maintained alternatives)

In the case of a GitHub fork, I'm not sure what it does there on the 
serverside. Maybe it does a mirroring clone. Read more about this in the 
elaborate SO answer 
here: 
http://stackoverflow.com/questions/3959924/whats-the-difference-between-git-clone-mirror-and-git-clone-bare
 

> With sufficient scripting, I can write something to enumerate and 
> compare branch heads, but I'm hoping someone knows of a tool that does 
> this, or can point me to some useful git plumbing to help. 
>
>
Have a look at the scripts above and see what they use. I'm guessing you'll 
find git fetch --prune helpful, as well as git branch --merged.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git on workstations? is that good?

[Thomas Ferris Nicolaisen]

On Sunday, August 24, 2014 5:03:09 PM UTC+2, guru prasad wrote:
>
> Version control novice here. Read through git docs, watched videos on git. 
> One of the advantages being talked about is the ability to have entire 
> project repository on user workstation. In my company there is major effort 
> going on to make user workstation as thin clients where it has much less 
> software than it used to. So if we are already accessing stuff via 
> network(vpn), what advantage is having distributed version control. Also 
> our thin client images sometimes gets lost and they restore from last back 
> up if possible. I loose file on c drive all the time. Would that mean I 
> have to put git repository on network drive? So am nervous advocating use 
> of distributed VC where repositories end up on unstable workstations. 
> Please help me understand so I can turn my team git as well. Thanks
>

Duplicate 
thread: https://groups.google.com/d/topic/git-users/ivI9ZEiqeks/discussion 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Measures to take regarding large binary files in a Git repository

[Thomas Ferris Nicolaisen]

On Monday, August 18, 2014 9:31:23 AM UTC+2, fabian@gmail.com wrote:
>
> 1. Have I got everything right in my analysis above? Am I missing anything 
> important, any problems I should expect?
> 2. Would you recommend setting core.bigFileThreshold, pack.packSizeLimit 
> or other options to non-default values proactively on all clients, or 
> should I rather postpone this until (if ever) we're experiencing problems? 
> If I don't set these values proactively, is there a chance that the Git 
> repository could be ruined?
> -- What is a good value for core.bigFileThreshold, given my concrete 
> binary files of 10 to 400 MB, some of which have up to 17 revisions?
> -- What is a good value for pack.packSizeLimit? Git for Windows defaults 
> it to 2g, is there any reason not to leave it at that?
> 3. Since pack.packSizeLimit does not affect the packs created for pulling 
> and pushing - what problems can I expect there? How could I tackle them?
> 4. "git repack -afd" and "git gc" currently fail with an out of memory 
> error on the migrated repository [1][2]. Should I worry about this?
> -- I can make "git repack -afd" work by passing "--window-memory 750m" to 
> the command. After that, git gc works fine again) Again, is setting 
> pack.windowMemory to 750m something I should do proactively?
>


I think you should bring this to the Git development mailing list, as the 
technical depth is a bit beyond what we usually discuss on this list.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Problem with commit

[Thomas Ferris Nicolaisen]

On Saturday, August 9, 2014 4:24:05 PM UTC+2, DavidFreed wrote:
>
> Dear Friends,
> I have this problem with Git , please advice me.
>
> git commit -am "commit all"
>
>
>
As the output tells you, there's nothing to commit here. You can add new 
files to Git by doing git add , but all those untracked files 
there look like build-output you don't to keep as part of your source code. 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: repo init not working in GIT BASH

[Thomas Ferris Nicolaisen]

On Thursday, August 7, 2014 2:54:49 PM UTC+2, scmmanage wrote:
>
> I have installed GIT GUI and then in GIT BASH tried to install repo. When 
> I try to do repo init I am getting errors.   Please let me know if anyone 
> used repo commands in git bash if so let me know how to install.
> I installed repo and found that python was not there in GIT BASH. Then 
> downloaded Python for windows and installed. Set the env path to bashrc.  
>
> $ repo init
> sh.exe": /c/Users/vikram.sadasivam.LGE/bin/repo: /usr/bin/env: bad 
> interpreter:
> Permission denied
>
> Please do ask me if anything I left out.
>
>
repo has their own mailing list 
here: http://groups.google.com/group/repo-discuss
 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Unstaged Changes with Brand New Repository

[Thomas Ferris Nicolaisen]

On Thursday, August 7, 2014 7:31:51 PM UTC+2, Ben Ruppel wrote:
>
> Hi,
> I'm very new to git.  I just created a new repository in a directory with 
> existing code with Git Gui 1.9.4 in windows.  All of the files that were in 
> the directory appear to be listed in "Unstaged Changes" in the upper left.  
> I'm not sure how to proceed.  Should I commit them?  I did set git up to 
> use Windows end of line for checkouts and then to check back in with Unix 
> style end of line, which is desired as this code is eventually heading out 
> to a linux system.  
>
> Can anyone explain what is going on with the files in the "unstaged 
> changes" box?
>

This is normal. Git is just telling you that there are a bunch of files in 
the current directory that haven't been added to the Git repository yet. 
You would see the same unstaged changes if you run `git status` on the 
command line.

After initializing a repository, you have to explicitly tell Git which 
files are to be version controlled. Doing so is a two-phase operation. You 
first stage, or add, which files you want to track, and then you commit the 
first revision, typically with a commit message like "Initial commit".

Example:

git add src (adds everything in the directory src/)
git add README (adds the single file)
git add :/ (adds absolutely everything)

Then:

git commit -m "Initial commit"

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Adding plugins from specific revision Github project to the Cordova project

[Thomas Ferris Nicolaisen]

On Wednesday, August 6, 2014 2:17:54 PM UTC+2, Surinaidu Majji wrote:
>
>
>
> On Wednesday, August 6, 2014 5:36:17 PM UTC+5:30, Thomas Ferris Nicolaisen 
> wrote:
>>
>> On Wednesday, August 6, 2014 8:01:36 AM UTC+2, Surinaidu Majji wrote:
>>>
>>> I have setup the android project using cordova command-line-interface 
>>> with the help of node.js. I have added the required plugin from Github to 
>>> our project by the following command:
>>>
>>> C:Users/Admin/>cordova plugin add 
>>> https://github.com/phonegap-build/PushPlugin.git
>>>
>>> For now it is working fine, but what if they suddenly changes the 
>>> plugin.So I want to stick with the specific revision of plugin. So how can 
>>> i get the plugin which is specific to particular revision. Shall i add the 
>>> revision number or commit number to the above url?
>>>
>>
>> I think you'd better ask in some Cordova forum. I have no idea what the 
>> above command does.
>>
>
>Yes you are right, but how do we  get the code which is specific to 
> particular revision, from git repository. 
>

If you just want to check out the code from a certain revision for quick 
inspection, use checkout:

git checkout 

In the long run, you'd probably mark it with a tag:

git tag  
git checkout 

 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Adding plugins from specific revision Github project to the Cordova project

[Thomas Ferris Nicolaisen]

On Wednesday, August 6, 2014 8:01:36 AM UTC+2, Surinaidu Majji wrote:
>
> I have setup the android project using cordova command-line-interface with 
> the help of node.js. I have added the required plugin from Github to our 
> project by the following command:
>
> C:Users/Admin/>cordova plugin add 
> https://github.com/phonegap-build/PushPlugin.git
>
> For now it is working fine, but what if they suddenly changes the 
> plugin.So I want to stick with the specific revision of plugin. So how can 
> i get the plugin which is specific to particular revision. Shall i add the 
> revision number or commit number to the above url?
>

I think you'd better ask in some Cordova forum. I have no idea what the 
above command does.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git over bbcp

[Thomas Ferris Nicolaisen]

On Thursday, July 31, 2014 9:08:58 PM UTC+2, Joanna Anthony wrote:
>
> Hello,
>
> Is anyone familiar with bbcp file transfer protocol? 
>
> I have some large git repos I'm trying to clone over a high latency vpn 
> link. I get really awful transfer rates using any of the standard 
> protocols. For regular file transfers I've been using bbcp for a while, and 
> get rates that can actually saturate my link.
>
> Now I've been trying to get git to work over bbcp, but without much luck. 
> Would anyone have any hints how I might get this to work?Unfortunately I'm 
> not too familiar with the internals of git but it seems like this should be 
> doable. 
>
>
As far as I know, Git only supports SSH, HTTP and the "native" Git 
protocol. There was some ftp support, but I'm not sure if this is supported 
anymore. See GIT URLS section 
on https://www.kernel.org/pub/software/scm/git/docs/git-clone.html for more 
info.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: how to create git patch based on different directories

[Thomas Ferris Nicolaisen]

>
> is there any similar option as "--relative" option in "git format-patch"?
> I try to use git format-patch based on different directory level, but I 
> cannot find the similar option as "--relative".
>
> appreciate your help in advance,
>
>
format-patch accepts common diff options, including --relative. 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Issues on the Git reference site need fixing

[Thomas Ferris Nicolaisen]

On Tuesday, July 22, 2014 5:51:13 PM UTC+2, Mickey Kay wrote:
>
> Hi there,
>
> Two issues that need addressing on http://git-scm.com/ (not sure if this 
> is the right place to post, but the support email isn't working for the 
> site).
>
> *1. Code formatting*
>
> Hi there, not sure if this is intentional or not, but all of the  
> elements on the site are rendering as block level, which makes it kind of 
> hard to read a lot of them that are meant to be inline:
>
> [image: Inline image 1]
>
> It'd be great if you implmented a method for specifying block vs inline 
> code to enhance readability, like so:
> [image: Inline image 4]
>
> *2. Support email doesn't work*
> I tried to email this issue to the listed support email, 
> g...@vger.kernel.org , however I got a permanent failure 
> auto-response. Any idea who needs to know about this one?
>
> That's all - thanks!
>

The git-scm.com website isn't actively nursed by the git-core developers 
. You're better off 
registering 
an issue on their GitHub page .

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Need help for updating git...

[Thomas Ferris Nicolaisen]

On Sunday, July 27, 2014 8:05:34 PM UTC+2, Bill CX wrote:
>
> Hi there,
> I have an issue to update the git on Mac, hope someone can help me.
>
> I am trying to use this command line to update my git to 2.0.3. git clone 
> https://github.com/git/git. 
>
> I currently have git 1.8.4, and I have no idea to remove the older version 
> either. Please help me. 
>
 
>

I've personally come to believe that any other way than homebrew [1] is the 
wrong way to install Git on a Mac these days.

brew install git

Then make sure that homebrew's bin folder (usually /usr/local/bin) comes 
first in your $PATH, so `which git` points to /usr/local/bin/git.

[1] http://brew.sh/

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: What version of SSL is supported by GIT 1.8.3

[Thomas Ferris Nicolaisen]

On Friday, August 1, 2014 9:05:22 PM UTC+2, RK wrote:
>
> I would like to know the SSL version supported by GIT 1.8.3 or 1.9.0
>

It appears you (?) have got some good answers on the Git development list 
already: http://git.661346.n2.nabble.com/Git-Help-needed-td7616237.html 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git svn preserving empty directories

[Thomas Ferris Nicolaisen]

On Thursday, July 31, 2014 10:19:13 AM UTC+2, Thomas Ferris Nicolaisen 
wrote:
>
> On Wednesday, July 30, 2014 12:04:01 PM UTC+2, Sebastian Kleine wrote:
>>
>> Im' trying to clone a svn repository into git with "git svn clone" 
>> command. It works except that empty directories are not kept in the git 
>> repository (they don't exists after git svn clone). I know this is normal 
>> git behaviour and that you can put a .gitignore file in an empty directory 
>> to keep the directory in git. I've read about the --preserve-empty-dirs 
>> parameter to git svn and thought this would handle the problem of missing 
>> empty directories after with git svn clone. But it doesn't. The directories 
>> are still missing. Why? How can I keep them?
>>
>> I'm using git svn version 1.7.10.4.
>>
>
> You have to specify the  --preserve-empty-dirs when doing the initial 
> git-svn clone. Doing it during an update (git-svn rebase) won't suffice.
>
> By default, this should create a .gitignore file in every empty directory 
> coming from SVN.
>

Oh, in an existing clone, you could try doing git svn mkdirs 
<https://www.kernel.org/pub/software/scm/git/docs/git-svn.html>: 

Attempts to recreate empty directories that core git cannot track based on 
information in $GIT_DIR/svn//unhandled.log files. Empty 
directories are automatically recreated when using "git svn clone" and "git 
svn rebase", so "mkdirs" is intended for use after commands like "git 
checkout" or "git reset". (See the svn-remote..automkdirs config file 
option for more information.) 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: git svn preserving empty directories

[Thomas Ferris Nicolaisen]

On Wednesday, July 30, 2014 12:04:01 PM UTC+2, Sebastian Kleine wrote:
>
> Im' trying to clone a svn repository into git with "git svn clone" 
> command. It works except that empty directories are not kept in the git 
> repository (they don't exists after git svn clone). I know this is normal 
> git behaviour and that you can put a .gitignore file in an empty directory 
> to keep the directory in git. I've read about the --preserve-empty-dirs 
> parameter to git svn and thought this would handle the problem of missing 
> empty directories after with git svn clone. But it doesn't. The directories 
> are still missing. Why? How can I keep them?
>
> I'm using git svn version 1.7.10.4.
>

You have to specify the  --preserve-empty-dirs when doing the initial 
git-svn clone. Doing it during an update (git-svn rebase) won't suffice.

By default, this should create a .gitignore file in every empty directory 
coming from SVN.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Remove files and folders permanently from all commits, history and pack files

[Thomas Ferris Nicolaisen]

On Thursday, July 17, 2014 8:04:23 AM UTC+2, Thomas Ferris Nicolaisen wrote:
>
> 
>

Oh, that's what I get for reading this list oldest -> newest :)

Duplicate post got some answers already here: 
https://groups.google.com/d/topic/git-users/HSVsp6Xv7OE/discussion 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Edit git patches with --minimal diff option

[Thomas Ferris Nicolaisen]

On Wednesday, July 16, 2014 8:44:59 AM UTC+2, Taowei Luo wrote:
>
> When running git diff or git format-patch, we know we may get a more 
> simply patch by --minimal option.
>
> I also want to split the changes in a file to several patches. So I use 
> git add -e or git add -p or git add -i.
>
> My problem is I can't get these two things together. Say, using git add -e 
> to edit patches on the minimal changes of a file. (It seems the --minimal 
> is not a default option and not even an option for git add)
>
> Anyone has ideas about this?
>

I'm not sure what the question is here. Do you just want to have 
patches/diffs of the things you've staged? git diff --staged > patch oughta 
do that..  

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Remove files and folders permanently from all commits, history and pack files

[Thomas Ferris Nicolaisen]

On Tuesday, July 15, 2014 4:41:20 PM UTC+2, cemico wrote:
>
> I have a git project in which all files are of about 700 MB size.
> In the .git folder lies a "pack" file which is 1.13 GB (Gigabyte)
>
> I accidently commited files and folders, which don't have to be under 
> version control.
>
> How do i remove let's say a "concept_art" folder from all commits (i have 
> about 50 commits by now) and from the pack file?
>
> I tried many things now. e.g:
> - 
> http://dound.com/2009/04/git-forever-remove-files-or-folders-from-history/
> - http://rtyley.github.io/bfg-repo-cleaner/
> - filter-branch with "rm"
> - git gc with prune=now and aggresive
>
> and the thing i thought which must really work:
>
> http://git-scm.com/book/en/Git-Tools-Rewriting-History#The-Nuclear-Option:-filter-branch
>
> None of the above work as expected or better: None work like i need it to 
> work.
> Some delete all commits (the most do that), some don't delete the files 
> from all commits, all(!) don't make the "pack" file smaller after deleting 
> all files and folders i don't want.
>

You're not giving us a lot to work with here. What were the commands you 
made? What was the output? What was the expected output?
 

> What i need is:
> - Exclude files and folders from my repo
> - Delete all references to theses files and folders from all commits
> - But don't delete any old commit
> - Rewrite the pack file, so that not any of the excluded files and folders 
> are in it
>
> What is the way to achieve that?
>

Filter-branch is the fallback tool when you know exactly want, and you 
don't mind the work of finding the right command to write, plus waiting for 
it to run (can take some time in large repositories).

Be aware that even after running filter-branch your repo will still be 
pretty big until you rinse out all the old stuff.

Here's an example I ran on some repos a while back:

# remove folders 'sandbox' and 'servers', and prune all empty commits:
git filter-branch -f --index-filter "git rm -r -f --cached --ignore-unmatch 
sandbox servers" --prune-empty

# clean away original refs left behind by filter-branch:
git for-each-ref --format="%(refname)" refs/original/ | xargs -n 1 git 
update-ref -d
# empty reflog:
git reflog expire --expire=now --all
# remove remotes:
git remote rm origin
# remove work tree changes:
git reset --hard
# clean up:
git gc --aggressive --prune=now

Remember to always make backups before you start filter-branching!

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: how to create git patch based on different directories

[Thomas Ferris Nicolaisen]

On Wednesday, July 9, 2014 9:49:18 AM UTC+2, loody mil wrote:
>
> hi all:
> below is my directories level
> level 0 --> level 1 --> level 2 --> level 3
>
> Below is my environment:
> the git repository is located at level 0.
> my modifications are located in level2
> 1. I use git diff create patch
> 2. receiver's root directory is at level2
> 3. the receiver need to use patch -p3 < xxx.patch.
>
> Below are my questions:
> how could I create patch that the receiver will only need to use -p1 when 
> his root directory is located in level2
> I try option "--no-prefix" but only eliminate one level.
>


See the git diff --relative 
option: https://www.kernel.org/pub/software/scm/git/docs/git-diff.html
 

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: How did I make this mess?

[Thomas Ferris Nicolaisen]

On Thursday, July 10, 2014 6:28:01 PM UTC+2, Joe Strout wrote:
>
> I joined a project yesterday that's using git for their version control.
>
> I pulled down the repo, changed one file, and added several new ones.  
>

How did you pull down the repo exactly? In doing this pull, you created 
your second change: the merge commit that obviously merged in a lot of old 
crap.
 

> Everything was in good shape and ready to share with the team.  So, on the 
> advice of someone who has used git a lot more than me, I did a Commit, then 
> a Pull (which apparently was required because a few other changes had been 
> made while I was working), and then a Push up to the master.
>

It's usually safer to do a git fetch. Then have a good look in SourceTree 
what you want to do - pick a target branch where you want your changes to 
end up (probably origin/master).

Now either merge or rebase your local master branch onto origin/master, and 
*then 
you check that everything looks OK*, both files and history. 

If everything looks all right, you can push. It's best if you have someone 
experienced guide you through this the first time though.
 

> This morning I heard from the team leader (who was remarkably calm) that 
> everything was broken.  Looking in the log, there are two entries from me: 
> the first one (9:10 PM) contains just the files I thought I was committing, 
> but the second one (9:11 PM) contains hundreds of files that I never 
> touched.  And these, apparently, leave the master in a nonworking state.
>

You created the second one when you made the pull command. Tell us what you 
did (parameters, etc), and show us your branch configuration (git remote 
-vv) if you want advise on what went wrong.
 

>
> I've attached a screen shot of SourceTree, which hopefully makes sense to 
> somebody here.  It baffles me in a variety of ways.  For example, between 
> my two commits 60 seconds apart, there are half a dozen other commits, 
> including July 2nd, July 5th, etc.  How in the world did those get in 
> between my two actions?
>

These are changes on other branches. Changes on other branches can happen 
in parallel. The displayed order in your GUI is based on some other logic 
than when something was committed.
 

> Our hypothesis is that me pushing my changes to the master somehow caused 
> a bunch of other changes, which previously were in other branches, to get 
> pushed to the master as well.  Can anyone explain this?
>

Yes. You merged in some branch when you pulled. You didn't notice that this 
was a bad merged, and you pushed it.
 

> Finally, what did I do wrong, and what should I have done instead?  My 
> goal is simple:
>
> 1. Get the latest version of the files
> 2. Make some minor edits
> 3. Share these edits with the group.
>
> This would be a simple svn update, edit, svn commit in my comfort zone, 
> but apparently in git it's not so easy.  Can anyone explain, as simply as 
> possible, how to accomplish this with git and SourceTree?  (Yes, I've read 
> git documentation and tutorials, but clearly I still don't get it.)
>
>
Yes. Git is not easy, and Git is not SVN. Git is a power-tool, and you need 
to get on top of branches, remotes, merging and rebasing. If you have no 
one to help you out, read chapter 2 and 3 here for starters: 
http://git-scm.com/book 

And try out these things one some dummy repo before you push to a 
repository other people are using.

And get some help from your colleagues.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Git doc on custom commands

[Thomas Ferris Nicolaisen]

On Monday, July 7, 2014 1:43:42 PM UTC+2, Giuseppe Manzo wrote:
>
> I red a few articles around the web on how to implement custom git 
> commands, but I cannot find anything about in the official site or in the 
> documentation.
> Does anyone knows how to find some official documentation about? I'd like 
> to knew more than the simple 'put any executable file in your path' I found 
> on the web.
> Regards,
> Beppe
>

I think git help leaves the job to the system's man-pages, so if you've 
managed to get a man page for your command in on your system, git help will 
delegate to it, I guess.

Have a look at 
https://www.kernel.org/pub/software/scm/git/docs/git-help.html for some 
hints.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Issue with exorbitant number of non-sensical conflicts following a second git rebase on branch

[Thomas Ferris Nicolaisen]

On Saturday, July 5, 2014 7:33:12 AM UTC+2, Matthew Johnson wrote:
>
> I usually only 'lurk' on this forum, since I am pretty new to Git myself, 
> too. But there are two things that stood out here I feel I need to speak up 
> about: 1) be very, very careful about following the 'guidance' from a Stack 
> Overflow post where neither the question nor any of the answers got a 
> rating higher than 3 2) rebase is NOT part of the normal Git workflow. Its 
> use should be rather rare. I cannot imagine why you would want to do it 
> after every checkout.
>

I agree, I would personally only bother doing it every once in a while, due 
to the overhead work. However, the more often you do it, the smaller amount 
of conflicts you'll have to deal with at a time. In a busy repository, it 
could be the right thing to do.
 

>
> In particular, the man page for git-rebase (recall that to look up git 
> commands in man, you need to add the '-') has two warnings that are likely 
> quite pertinent.
>
> 1) Rebasing (or any other form of rewriting) a branch that others have 
> based work on is a bad idea
>

As long as thread-starter is the only one using this feature branch in the 
shared repository, it should be all right.
 

> Taking a slightly different approach, the online Git book covering rebase (
> http://git-scm.com/book/en/Git-Branching-Rebasing) seems to imply that if 
> what you want to do can be easily done with merge, then you should do so, 
> resorting to rebase only when you want to do odd things like apply not the 
> whole current branch, but only a patch in it to another branch.
>

I think that's a bit old-fashioned view on things. I think we're better off 
embracing rebase as a tool earlier than later, teaching new users how and 
when to use it appropriately. There are several use-cases for rebase that I 
would classify as normal procedure:

* Doing minor commits on master for which you don't want a branch: git pull 
--rebase
* Starting off on branches that you end up wanting to just rebase on master 
instead (and fast-forward onto master)
* Rebase interactive to clean up commits

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Issue with exorbitant number of non-sensical conflicts following a second git rebase on branch

[Thomas Ferris Nicolaisen]

On Friday, July 4, 2014 10:11:57 AM UTC+2, ayrshi...@gmail.com wrote:
>
> However, I have just tried to do the same thing again and I have been hit 
> with a vast number of conflicts. Now, conflicts are fine but some files 
> seem to almost be in a permanent state of conflict. For example, commit 2 
> will show as a conflict of commit 1 so rather than git realising that 
> commit 2 comes after commit 1 - it shows as a conflict.
>
> Is this a symptom of rebasing on top of a branch that I have previously 
> rebased on?
>

You will keep running into the same conflicts when rebasing like this, I 
think. You can make Git remember the conflict resolutions from previous 
rebases by enabling "rerere": http://git-scm.com/blog/2010/03/08/rerere.html

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: two tree alias ?

[Thomas Ferris Nicolaisen]

On Tuesday, July 1, 2014 9:12:58 AM UTC+2, lei yang wrote:
>
> Hi expert,
>
> I have two tree 
>
> git://xx-git.ws.com/git/layers/A (nothing contains,just init)
> git://xx-git.ws.com/git/layers/B (lots of content)
>
>
> for now I want to realize this, when I clone A,it actually clone B? 
> can we create a alias?
>

If you have access to the filesystem where the repositories are stored on 
disk, maybe you can replace A with a symlink to B?

cd /var/git/layers # or wherever they are stored
rm -rf A
ln -s B A

>

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[git-users] Re: Unable to get colors with git 2.0 on os x 10.9

[Thomas Ferris Nicolaisen]

On Saturday, June 28, 2014 9:33:16 AM UTC+2, Martin H. Andersen wrote:
>
> I have installed git 2.0 with Homebrew an changed my $PATH in the 
> .bash_profile to: export PATH=/usr/local/bin:$PATH so that I don't work 
> with the build in apple git version.
>
> This is the git command that should work: *git config --global color.ui 
> true *But nothing happens
>
> I have tried all the color settings in the terminal without any effect.
> Is this a git 2.0 problem or just "something"?
>

Colors are fine in my terminal on OS X with Git 2.0.  

Please post a screenshot of some output where you were expecting colors, 
and the output of `git config -l | grep color'. Run git --version if you 
are in doubt whether the right Git installation is being used.

-- 
You received this message because you are subscribed to the Google Groups "Git 
for human beings" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to git-users+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.