I have the following situation: I have an application checked out from GitHub, made some changes and checked it in a local repository (Bitbucket). During months I added new functionality into this local repo.
In the meantime, the GitHub project has been further developed.
Now I'm looking for a way to merge the GitHub master branch into my local master branch. It would be nice if I can do it in eclipse. How should I proceed and how can I avoid conflicts?
In the Git Repositories view:
Add an additional remote/upstream repository: right-click the node Remotes and choose Remote Remote...
Fetch from the GitHub repository: see EGit User Guide - Fetching from other Repositories
Merge changes from a remote branch of the GitHub repository into a local branch
Consider also to contribute back your changes to the GitHub project.
Related
I created an empty project in Git. I also created a Java EE projcet in Eclipse (note: Eclipse has the Git plug-in installed).
From within Eclipse, I accessed and duplicated the (empty) Git Repository project on my Local Machine (or rather laptop).
I then wrote some code in my Java EE project.
From within the project, I then used the Team -> Commit option on it to :
1) Add the project to the Local Repository
2) Commit and push the contents from the Local Repository to the Git Repository.
My problem is that all of this work was done using the Master branch.
The Git Repository is currently sitting on the Master branch.
The Local Repository is currently sitting on the Master branch.
Whenever I make changes to the project and do a Team-> Commit, all actions happen with the Master branches.
Master -> Master
What I would like to do is as follows:
1) create 3 more branches in Git Repository (called consolidation, testing, production) - each new branch would be a copy of the Master Branch
2) On the Eclipse side, I need to fix it so that when any changes are made and saved to the Local Repository, the data will be saved to the
Consolidation Branch of the Local Repository and ~NOT~ the Master Branch (which is what happens now).
3) After saving data in the Consolidation Branch, from then on out, when code is sent to the Git Repository, it would be done by using the Consolidation branch and NOT the Master branch.
How can I do this?
TIA
Update
#Howlger
Hi and thanks for the response! Your suggestion gave me a clue on how to solve the problem. First, I logged into Git and copied the Master branch to a "dev" branch.
Next, I switched to the Git perspective (locally) under Eclipse and created a "dev" branch for my project:
Branches -> Switch To -> New Branches
Then, I followed some of your advice except I created a path that would allow for
branch : dev (local) => branch : dev (remote)
It is under the "Advanced" button.
I did a small test and it worked. Thanks again!
Update
Hi again - just logging in to note that I changed the references so that
Source Ref = refs/heads/*
Target Ref = refs/heads/*
The mapping of a local branch to a branch of the remote/upstream repository is specified by the push ref mappings. It looks like your current push ref mapping is HEAD:refs/heads/master: your current local branch will be pushed to the remote/upstream repository master branch. To map local branches to the remote/upstream repository branches with the same name, just remove all push ref mappings:
In the Git Repositories view righ-click the Remotes > origin > push node and choose Configure Push...
Select the mapping and click Delete
Click Save
See EGit User Guide > Push Ref Specifications for more details.
I've been a great fan of Plastic SCM for a while now, using it locally for my own personal projects. Lately I've decided to contribute to an open source project on github. Since Plastic SCM can be used as a github client I decided to use it as such.
Now this is my first time using github, so I followed the classic github workflow as described in all the tutorials:
Fork the repo to get your own copy
Sync the Plastic SCM repo with the personal github repo
Commit changesets in Plastic SCM
Sync the Plastic SCM repo with github again
So far everything went fine. I contributed my code, uploaded it to github and saw all the changesets and branches appear in the github webui.
Next I put in a pull request to the original repo I forked from. This is where my problems start. The pull request was accepted, and a new changeset appeared in the original repo which contains my pull request. It seems all my intermediate branches did not transfer to the original repo either.
As shown in the network graph above, the original repo (purple) is now no longer in sync with mine. I started with further development in another branch (VEH003) but this one seems to be completely disconnected from the original repo (purple).
In my Plastic SCM client I can see exactly the same network graphs in the branch explorer, with the exception of the purple branch.
I read that to update your fork with the changesets from the original repo this has to be done explicitly, and in your local git repository (github help). Now I don't know how this is done using Plastic SCM.
To sync your local fork with the original repo, as described in the github help, you need to merge the upstream master branch with your local master branch.
The point here is that Plastic SCM does not track or sync the upstream references (the references of the original repo), just your github's repo references.
To be able to update your fork, and continue working with Plastic SCM you should use the command line git client to perform the update following the guide you linked, push the local git changes to your github repository and then sync Plastic SCM with github to get the changes. Step by step:
Use the git client to clone your github repository
git clone your-fork-url
Add the upstream repository.
git remote add upstream original-repo-url
Fetch the upstream references.
git fetch upstream
Merge upstream changes to your master branch.
git checkout master
git merge upstream/master
Push master branch to your github repository.
git push origin master
Sync Plastic SCM with your github repo.
I'm kind a newbie on Bitbucket so I have a lot of doubts.
I share an Eclipse project between two different computers. I have pushed the project from one computer to the Bitbucket repository using the Eclipse plug-in. Now when I go to the other computer I have fetched the repository that is saved on Bitbucket, and it has created a new remote tracking branch on my Eclipse local repository.
My problem comes when I want to merge that branch with the local branch, I have no idea how can I do that
Anyone could help me?
ps: sorry if the problem or the situation is not very clear
You should follow the section "Pulling New Changes from Upstream Branch ":
a git pull is a fetch + a merge.
Right-click on a project in the Package Explorer and select Team > Pull or right-click on a repository in the Git Repositories view and select Pull to pull new changes from the upstream branch your local branch is tracking.
The section "Fetch and Pull" of the Egit tutorial says as much:
When cloning remote repositories, Git creates copies of the branches as local branches and as remote branches.
A Fetch operation will update the remote branches only.
To update your local branches as well, you will have to perform a Merge operation after fetching.
I came to know about the git / egit version control system last week, seems too good to be true .So thought to shift from SVN to git..Since last week I am trying to understand the basics and concepts of git.
So I created a test environment for understanding the workflow of egit in eclipse as following.
I am following the strategy of remote tracking there are two repositories named local and remote used for understanding the workflow.
I created a repository named 'remote' with an emtpy index.php file and has one master branch and imported the project into eclipse.
I created another git repository named 'local' by cloning the above 'remote' git repo, this repository is now tracked by 'remote' repository has one master branch and origin/master remote tracking branch.
I imported the project from 'local' git repo. into my eclipse workbench and changed the index.php file using eclipse php editor -> committed changes to local's master branch and -> performed push from local master branch.
When checking the remote working directory .. there is no change updated which I did in local's index.php file , however master branch in the remote repository view shows the latest commit which I committed in the local's master branch, but unfortunately files are not updated, it just adds asterisk mark to all changed files in my remote project view.
So researching about the asterisk mark I found its in staged condition .So can anyone lead me to the right way explaining how to successfully perform push operation from local's master branch to remote's branch I will be grateful.
this is the picture of my egit test case set up in eclipse for understanding the workflow. you can see the asterisk mark in the remote project after performing push from local repository,you can see all three branches viz.local's master, origin/master and remote's master branches shows same latest commit.
1
Thank You.
I tried using bare repo. as a remote, when I push from changes from local to remote, I don't find any files in remote repo. Just branch gets updated with the latest commit.
That seems normal, since a bare repo has *no working tree, meaning no files.
You would need to clone that remote repo, and add this cloned repo (non-bare) as a project in your Eclipse, in order to:
push from your first Egit-managed repo to the bare remote repo
pull from that same bare repo to your second Egit-managed non-bare repo.
I'm new to git and am wrapping my head around how I'm supposed to be using git and egit. From the egit tutorial, I have setup a respository on GitHub, pushed my Eclipse projects to the remote GitHub repository from my local workspace, I can push changes to GitHub, switch branches, see the updates on GitHub, etc. This all makes sense.
Looking at the Git Repository explorer, I have a listing of "Local" branches and have no "Remote Tracking" branches and I have no "Remotes" listed. When I create a branch from a local branch, the egit dialog indicates "You are creating a branch based on a local branch" and suggests that I should be making a branch from a remote tracking branch.
So my question is, am I correctly using egit?
Should I just continue pushing changes to the remote GitHub repository? If so, what happens once I share the project and other developers clone the repository and start making changes to the remote repository?
Or should I now ditch the local repository and setup a new remote repository by cloning the existing GitHub repository that I initially created from my workspace?
Or do I create a new Push and Fetch "Remote" for my existing git repository?
Or something else?
Confused.
Since you created the repo on your local system and then pushed it to github without creating a remote you don't have a remote at hand. A remote is simply a short alias for the remote repository's URL. To fix this create a remote and a push and fetch configuration from the repositories view. In order to populate remote tracking branches in your local repo you need to run fetch once. As soon as this is done you can use "Push to upstream" instead of the more complex Team > Push... dialog which allows to define all parameters on the fly. When using native git command line you'll find the same concepts implemented there:
with
"$ git push [url] [refspec]" (e.g. "$ git push https://github/user/myrepo.git master:master")
you pass all parameters explicitly, this is similar to Team > Push... in EGit
with
"$ git push [remote]" (e.g. "$ git push origin")
you push to the repository defined by the configuration parameters of the given remote (check .git/config to see these parameters or open repository configuration from egit preference in Eclipse), this is similar to Team > Push to upstream in EGit. Usually the refspec used in this way is implicitly configured when creating a local branch based on a remote tracking branch. It's also possible to add this configuration later but since this is more
tedious manual configuration the other way is more handy.
If you clone a remote repository the repository you cloned from is stored as remote "origin" in your clone. This way you can skip configuring the remote manually. This is only needed if the repository is born when you create it from scratch.
The "Branching" section of the Egit User Guide can help:
There is no obligation to create a local branch which would be named like a remote tracking branch (see "Having a hard time understanding git-fetch" to have a good understanding of "remote tracing branches).
You can create as many local branches (i.e. branches that you won't necessary push anywhere) as you want/need.
But if you don"t see any remote branch, maybe you didn't fetch that GitHub repo in the first place: see Fetching.