Search Results

Search found 4953 results on 199 pages for 'git commit'.

Page 25/199 | < Previous Page | 21 22 23 24 25 26 27 28 29 30 31 32  | Next Page >

  • How to do simultaneous builds in two Git branches?

    - by james creasy
    I've looked at git-new-workdir, but I don't want the history to be shared because the branches have a release-main relationship. That is, changes in the release branch I want to propagate to the main line, but changes in the main line I don't want in the release line. A common pattern for me is to fix a bug in the release line, integrate it to the main line, then start builds in both branches at the same time. Is there a way to do this with git-new-workdir, do I need to clone, or is there a better solution? Thanks

    Read the article

  • Can't get gitosis and ssh to play nice on cygwin

    - by Noel Kennedy
    I have followed this guide to setting up gitosis on a windows 2003 server via cygwin. I have now got to a point where it largely works. I can clone, pull and push. The problem I am having is that I think I have not got the ssh bit right at all. When I connect via msysgit from machines and accounts where I have not created or uploaded ssh keys it works. Every time I clone, pull or push I get a password challenge for the 'git' user running on the server but basically I can execute git commands. When I connect with users with an ssh key in the ~/.ssh folder, I don't get the password challange and instead I get a permissions failure: DEBUG:gitosis.serve.main:Got command "git-upload-pack '/cris.git'" DEBUG:gitosis.access.haveAccess:Access check for 'teamcity@hhit24808' as 'writable' on 'cris.git'... DEBUG:gitosis.access.haveAccess:Stripping .git suffix from 'cris.git', new value 'cris' DEBUG:gitosis.access.haveAccess:Access check for 'teamcity@hhit24808' as 'writeable' on 'cris.git'... DEBUG:gitosis.access.haveAccess:Stripping .git suffix from 'cris.git', new value 'cris' DEBUG:gitosis.access.haveAccess:Access check for 'teamcity@hhit24808' as 'readonly' on 'cris.git'... DEBUG:gitosis.access.haveAccess:Stripping .git suffix from 'cris.git', new value 'cris' ERROR:gitosis.serve.main:Repository read access denied fatal: The remote end hung up unexpectedly I have uploaded the public rsa key into the key_dir folder. Here is my conf file: [gitosis] loglevel = DEBUG [group gitosis-admin] writable = gitosis-admin members = myemail@mydomain [group cris-developers] members = myemail@mydomain TeamCity@HHIT24808 writable = cris If it matters, I have generated a key without a passphrase as I believe this is necessary to enable ssh for automated scripts. When I use keys with a passphrase, I get challanged for the phrase but then get the same permissions problem. I have tried 'writable' and 'writeable' for permissions. Help!! Update 1: When I try to clone a non-existant repo, I get the same error message, co-incidence? Update 2: Wierd, I've got one machine and one login working. It seems to be something to do with the syntax for addressing git over ssh. This now works on one machine for one login: git clone git@servername:cris.git The same command fails for a user on another machine without an uploaded ssh key. But this command works (after being challanged for git@servername's password) git clone git@servername:/home/git/repositories/cris.git neither command works on a 2nd login whose ssh key has been uploaded

    Read the article

  • NFS share access - Permission denied

    - by rgngl
    I'm trying to share a directory on my NAS device(WD Mybook WE) with NFS to another machine on my local network. The directory on the NAS device looks like this: drwxr-x--- 15 git git 4096 Nov 17 01:05 git/ And id's of the user git on the NAS device is like this: [root@myhost DataVolume]# id git uid=505(git) gid=505(git) I played with many different parameters in the /etc/exports file and this is what I got there currently: /DataVolume/git 192.168.0.20(async,rw,no_root_squash,no_subtree_check) On the client side I have the user git and group git with the same id's to match the ones on the server. user@myclient:~$ id git uid=505(git) gid=505(git) groups=505(git) I mount the directory with: sudo mount myhost:/DataVolume/git -t nfs git/ and the mounted directory looks like: drwxr-x--- 15 git git 4096 Nov 17 01:05 git After these steps I can't seem to cd to that directory with any user, including git and root. I am getting a Permission denied error. Thanks in advance for any help.

    Read the article

  • Pushing to bare Git repository (remote) causes it to stop being bare

    - by NSD
    I have a local repository called TestRepo. I clone it with the --bare option, zip this clone up, and throw it on my server. Unzip it, and it's still bare. I then clone the bare remote repository locally over ssh with something like git clone ssh://[email protected]/~/TestRepo.git TestRepoCloned The local TestRepoCloned is not bare and has a remote called "origin." It appears to be tracking correctly from the looks of its config file [core] repositoryformatversion = 0 filemode = true bare = false logallrefupdates = true ignorecase = true [remote "origin"] fetch = +refs/heads/*:refs/remotes/origin/* url = ssh://[email protected]/~/TestRepo.git [branch "master"] remote = origin merge = refs/heads/master I edit an existing file. I commit the change to the current branch (master) via git commit -a -m "Edited a file." The commit succeeds and all is well. I decide to push this change to the remote repository via SSH with a git push The remote repository is now no longer bare, but has a complete working directory, and I get continuous error messages on all further attempts to push to it. Everything I've read seems to suggest that what I'm doing is correct, but it simply is not working. How am I supposed to push changes to a bare remote repo and actually keep it bare?

    Read the article

  • Can't get automated release working with Hudson + Git + Maven Release Plugin

    - by Christopher Maier
    As the title says, I'm trying to get an automated release job working on Hudson. It's a Maven project, and all the code is in Git. Manually, I do the release on my personal machine like so: git checkout master mvn -B release:prepare release:perform This works perfectly. The Maven release plugin properly pushes the release tag to the origin repository as well as the next commit that bumps the version to the next SNAPSHOT. However, when I run this same Maven job through Hudson (either by creating my own "release" job or by using the M2 Release Plugin) it doesn't work so well. The release tag gets pushed out to the origin repository, and the release gets pushed out to our Nexus repository, but the subsequent commit that bumps the version to the next SNAPSHOT doesn't go out. Furthermore, the "master" branch in the origin repository doesn't get changed at all. I've looked in Hudson's workspace for the job, however, and the version has been updated. After looking at the output from the Hudson job, it appears that the Git plugin does not actually checkout "master", but rather it's SHA1 id. That is, if the "master" branch label points to commit "f6af76f541f1a1719e9835cdb46a183095af6861", Hudson does git checkout -f f6af76f541f1a1719e9835cdb46a183095af6861 instead of git checkout -f master As a result, the changes that the Maven release plugin is making are not actually on any branch (certainly not on "master") and these changes don't make it to the origin repository. It runs on the right code, but bookkeeping-wise, the changes seem to get lost because no branch label points to them. Has anybody gotten the Hudson + Git + Maven Release Plugin combo to work properly? Is there some additional configuration somewhere I can set to make this happen? Or is this a bug in the Hudson Git plugin? Thanks in advance.

    Read the article

  • Git on Windows 7 expecting Linux? /dev/null not found error

    - by Klikini
    I have installed git (not GitHub) on Windows 7 x64 Home Premium, and I cannot get it to work. Opening Git Bash outputs the following: Welcome to Git (version 1.9.4-preview20140815) Run 'git help git' to display the help index. Run 'get help <command>' to display help for specific commands. sh.exe": /dev/null: No such file or directory sh.exe": /dev/null: No such file or directory sh.exe": /dev/null: No such file or directory sh.exe": /dev/null: No such file or directory sh.exe": /dev/null: No such file or directory sh.exe": /dev/null: No such file or directory Andy@ANDY-DELL ~ $ If I open the Git GUI, I get a this box: Title: git-gui: fatal error Content: fatal: open /dev/null or dup failed: No such file or directory Git Gui requires Git 1.5.0 or later. I also tried GitHub for Windows, but I got an internet connection error when attempting to clone a repo, even though my connection is fine. Is this possibly related? I have learned so far that /dev/null is the Linux version of the Windows NUL, but why is it trying to do this on Windows? Thanks in advance.

    Read the article

  • Add comment to subversion commit automatically

    - by Paul Alexander
    I've already got my subversion repository set up to require comments of a minimum length to accept a commit. However, I'd like to start tagging those comments with information from our bug tracking system when committed. I've already got the scripts set up to pull data from the bug tracker and just need a way to get that info into the subversion commit comments. How can append to the existing comment in subversion automatically? For reference, the subversion repository is hosted on a linux server with Ubuntu 9 something installed and I have complete root access to the machine.

    Read the article

  • Multiple git repos in plesk virtual hosts

    - by icc97
    Each plesk vhost only has one user access that httpdocs directory. I want to manage a whole bunch of separate virtual hosts using Git. Does this mean installing a separate Git repository with a separate user / ssh public key combination for each virtual host or is there a way of centralising it at all? Gitosis sounds like it might help - but I'm not sure if it gets round pushing the files to each virtual host.

    Read the article

  • AnkhSVN client side pre-commit hook

    - by santa
    Basically I want to do the same thing as the fella over there. It seems that everybody was thinking about server-side hooks (with all their evil potential). I want a client side script be run before commit so astyle can format the code the way my boss likes to see it. Since my IDE (VS2010Pro) automatically checks when a file changed on the disk an opts me in for reloading it, there is no real evil with all that. Is there any (clean) way to accomplish that with AnkhSVN? Maybe there's also a way to extend VisualStudio to call my pre-commit-script...

    Read the article

  • Git sh.exe process forking issue on windows XP, slow?

    - by AndyL
    Git is essential to my workflow. I run MSYS Git on Windows XP on my quad core machine with 3GB of RAM, and normally it is responsive and zippy. Suddenly an issue has cropped up whereby it takes 30 seconds to run any command from the Git Bash command prompt, including ls or cd. Interestingly, from the bash prompt it looks likes ls runs fairly quickly, I can then see the output from ls, but it then takes ~30 seconds for the prompt to return. If I switch to the windows command prompt (by running cmd from the start menu) git related commands also take forever, even just to run. For example git status can take close to a minute before anything happens. Sometimes the processes simply don't finish. Note that I have "MSYS Git" installed as well as regular "MSYS" for things like MinGW and make. I believe the problem is related to sh.exe located in C:\Program Files\Git\bin. When I run ls from the bash prompt, or when I invoke git from the windows prompt, task manager shows up to four instances of sh.exe processes that come and go. Here I am waiting for ls to return and you can see the task manager has git.exe running and four instances of sh.exe: If I ctrl-c in the middle of an ls I sometimes get errors that include: sh.exe": fork: Resource temporarily unavailable 0 [main] sh.exe" 1624 proc_subproc: Couldn't duplicate my handle<0x6FC> fo r pid 6052, Win32 error 5 sh.exe": fork: Resource temporarily unavailable Or for git status: $ git status sh.exe": fork: Resource temporarily unavailable sh.exe": fork: Resource temporarily unavailable sh.exe": fork: Resource temporarily unavailable sh.exe": fork: Resource temporarily unavailable Can I fix this so that git runs quickly again, and if so how? Things I have tried: Reboot Upgrade MSYS Git to most recent version & Reboot Upgrade MSYS to most recent version & Reboot Uninstall MSYS & uninstall and reinstall MSYS Git alone & Reboot I'd very much like to not wipe my box and reinstall Windows, but I will if I can't get this fixed. I can no longer code if it takes me 30 s to run git status or cd.

    Read the article

  • Add a git repository to FishEye

    - by lostInTransit
    Hi I am trying to find some help on the FishEye documentation to help me add a git repository to it. This is all I can get and I have no idea what to put in the repository location (git://, ssh://, https:// which URL do I put there?) Can someone please help me out! Thanks.

    Read the article

  • Multiple CheckOut or Multiple Commit .

    - by marco
    Hello sorry for my English ... I'm using Visual studio 2008 with C# and SharpSVN dll (for Subversion 1.6) I wish to commit to multiple different folders in my repository, but without having to checkout the entire repository. I tried using the method: SVNClient.Commit (ICollection coll, CommitArgs ca); Where eg coll [0] = "c:\svnCheckoutDir\dirLic001\sect\file_ext.ini" and coll [1] = "c:\svnCheckoutDir\dirLic121\sect\file_ext.ini" but I get the error message: "c:\svnCheckoutDir" is not a working copy. when I download the entire repository, I have no problems. but when I do a checkout of the folders you just want to modify the process does not work. Help me please . Marco Untiveros Novatronic Peru SAC.

    Read the article

  • subversion problem - commit access

    - by Calvin
    Hi everyone, I'm new to setting up subversion but originally when I made a repository, all my team members could update and commit without problem. There was a problem with it so we decided to recreate it, but now only I can commit changes to it. And my username/password doesn't work on their computers, so I'm sure it's something obvious and silly, but I just don't know enough to know what's causing it. The passwd and svnserve.conf files are the same as the original repository that worked for everyone. Any ideas? Thanks in advance.

    Read the article

  • Subversion post-commit hook to sync rep with FTP server ( for a website )

    - by Brett
    I've installed a repository on my computer locally. What I'm trying to do is be able to work on a website locally on my computer and see changes using something like MAMP. When I commit a change though I'd like it to sync my repo with the live website source files on a remote FTP server. I've done a bit of digging and I know that people keep saying to use a post-commit hook but I'm not sure how to configure it or even how to install it locally. Also i'm not sure if it's possible to do from my computer to an FTP. Could someone be a huge help and walk me through how to do this I've been trying for hours to figure out how to do it. thanks so much.

    Read the article

  • Git: push via ssh to a root owned repository with ssh root logins disabled

    - by anthonysomerset
    is that even possible? Summary, i'm running puppet master on a server and ideally we want root logins via ssh disabled, we want to force all access via sudo if root access required however we have puppet installed using a git repo to manage the manifests, this repo is currently owned by root and currently i only know of 2 solutions (less ideal) allow root access via key auth only - if so, what can i lock it down to to only allow the git push commands? own the repo in /etc/puppet as a different owner - will puppet work reliably with this?

    Read the article

  • push commits to git (gitolite) repository messes up file permissions (no more trac access)

    - by klemens
    already posted here so feel free to answer there. everytime i commit/push something to the git server the file permissions change (all added/edited files in the repository have no read and execute access for the group). thus trac can't access the repository. do I need to change permissions of the folder differently? chmod u=rwx,g=rx,o= -R /home/git/repositories or do i need to setup gitolite somehow to write files with different permissions??? regards, klemens

    Read the article

  • How do I edit git's history to correct an incorrect email address/name

    - by Chas. Owens
    When I started using git I just did a git init and started calling add and commit. Now I am starting to pay attention and I can see that my commits are showing up as cowens@localmachine, rather than the address I want. It appears as if setting GIT_AUTHOR_EMAIL and GIT_COMMITTER_EMAIL will do what I want, but I still have those old commits with the wrong email address/name. How can I correct the old commits?

    Read the article

  • Git push from post-receive

    - by meka
    I have two servers, let's call them first and second. First one is where the real development is done, and second one should be the replica. What I would like to do is put "git push" in post-receive, but there is one problem. Post-receive is executed as the user doing git push to first server, so I can't chmod 600 ssh key with no pass. What is the best practice for this? Thanx!

    Read the article

  • Rescuing files and commits from "no branch" in git

    - by Xeoncross
    I started working on some files I had in a git submodule under another project. However, since it was a git submodule it never checked out "master" and instead just checked out the head and placed all the files in the folder in "no branch". Now that I've made some changes by accident to these files I just realized that I was working in a "no branch", submodule of my project. How do I get those files into a branch (like master) so I can rescue them?

    Read the article

  • How do I correctly install dulwich to get hg-git working on Cygwin?

    - by Erik Vold
    I have a similar issue as in this issue, but in my case I am trying to use cygwin. First I followed the instructions here, and I ran: $ easy_install hg-git The I created ~/.hgrc, with: [extensions] hgext.bookmarks = hggit = Then when I typed 'hg' at a command prompt, I'd see: "* failed to import extension hggit: No module named hggit" So I did a search for "hggit" and found /cygdrive/c/Python26/Lib/site-packages/hg_git-0.2.1-py2.6.egg/hggit, so I updated .hgrc: [extensions] hgext.bookmarks = hggit = /cygdrive/c/Python26/Lib/site-packages/hg_git-0.2.1-py2.6.egg/hggit Then when I type 'hg' I get "No module named dulwich.errors" If you read this question, it's the same problem. In python shell I cannot import dulwich: >>> import dulwich Traceback (most recent call last): File "<stdin>", line 1, in <module> ImportError: No module named dulwich I checked out my easy-install.pth and it does contain the dulwich egg: import sys; sys.__plen = len(sys.path) ./hg_git-0.2.1-py2.6.egg ./dulwich-0.5.0-py2.6-win32.egg import sys; new=sys.path[sys.__plen:]; del sys.path[sys.__plen:]; p=getattr(sys,'__egginsert',0); sys.path[p:p]=new; sys.__egginsert = p+len(new) So how can I fix this so that import dulwich works, which should fix my problem using hg-git I assume..

    Read the article

  • Git already up to date unless I reset

    - by Chris
    I have a cloned repo I use for the live site and I have it pull from a bare repo. For some reason every time I execute "git pull" I get the already up to date message. But it's not up to date and not updated. If I do a git reset --hard HEAD^ and then git pull again, the changes come in. What's the problem here and how do I fix it so that git pull will pull and merge changes in without needing a hard reset?

    Read the article

  • Textbased issue-tracker/todo list for Git?

    - by anon
    I've been managing all of my todo-lists as ~/git-repo/todo which is kept under git. THen I add/delete files from the todo list, and have git autocommit all changes. However, I feel there should be more powerful tools. Besides "cil" and git-issues [neither of which I've tried], what tools are available? PS I want something that's entirely text/command line based. Thanks!

    Read the article

  • git merge should ignore one directory

    - by dorelal
    I have tons of data in directory called reports. While doing git merge with another branch I am getting lots of conflicts for files under reports directory. I would like git merge to ignore files under reports. In another words I would like all the data from reports from master and not from lab branch. Is that possible? This is what I am doing right now. git checkout master git merge lab

    Read the article

  • Simplify a batch of git commands

    - by Bogdan Gusiev
    When I want to merge one branch to another I use to do the following(in this example master to custom): git checkout master && git pull && git checkout custom && git merge master Can somebody suggest how to simplify this? Thanks, Bogdan.

    Read the article

< Previous Page | 21 22 23 24 25 26 27 28 29 30 31 32  | Next Page >