Skip to main content

The Foolproof Git Setup

A confused Git user.
There was a recent complaint on Git User mailing list. Quote:
Why do I have to be a source control engineer just to be a software developer?
Now, the main issue of this poor person is that his organization has failed to on-board him into his job. They told him to use Git, but didn't give him the means to do so productively (as far as we know, anyway).

What I started wondering was: Could his employer somehow have set up Git for him so that he could work with it just as easily as with Subversion?

Given a prepared .gitconfig with all the aliases and configurations that you wanted, could you really dumb down Git so much that it would be analogous to the simpleness of SVN?

Well, here's my take on it: To a certain degree: yes. If you could narrow down Git to do the most typical two SVN commands:
  • update (pull without merging), and
  • commit (with push)
.. that would help a lot.

Dumbing down pull

The first step would be to make git pull as stupid as an svn update - always rebase local commits on top of the ones being fetched:

git config --global pull.rebase true

The above setup has a weakness though. If pull-rebase runs into a conflict, even if there is only one local commit to rebase -  you get stuck in rebase mode and you have to start figuring out what Git wants you to do. This is what it looks like (and I can see why newcomers still find Git scary when I look at this):

  ~/temp/bar/[master]>git pull
remote: Counting objects: 5, done.
remote: Compressing objects: 100% (3/3), done.
remote: Total 3 (delta 2), reused 0 (delta 0)
Unpacking objects: 100% (3/3), done.
From /Users/tfnico/temp/foo
   24b3b7f..4705398  master     -> origin/master
First, rewinding head to replay your work on top of it...
Applying: A random commit message: makeshiftness
Using index info to reconstruct a base tree...
M readme.txt
Falling back to patching base and 3-way merge...
Auto-merging readme.txt
CONFLICT (content): Merge conflict in readme.txt
Failed to merge in the changes.
Patch failed at 0001 A random commit message: makeshiftness
The copy of the patch that failed is found in:
   /Users/tfnico/temp/bar/.git/rebase-apply/patch

When you have resolved this problem, run "git rebase --continue".
If you prefer to skip this patch, run "git rebase --skip" instead.
To check out the original branch and stop rebasing, run "git rebase --abort".

That's a lot of text for one little conflict.

While SVN will present all conflicts at once, Git will insist you handle them one commit at a time, which is nicer, but still enters you into rebase-mode where you have to go git rebase --continue when conflicts have been resolved.

So, until I can figure out a way around that, you'll have to teach your newcomer at least enough to be able to resolve conflicts, do git add <resolved file> and git rebase --continue.

OR, we go back to the drawing board.

There must be a better way

I realized this after trying to publish this blog-post, and by the time I realized it hadn't gone live, I had a better idea.

What if we just squash all local commits into being changes in the work tree?

We could make an alias: git update - which does the following

  1. git reset --soft origin/master
  2. git stash
  3. git pull
  4. git stash pop
This would essentially guarantee that any conflicts occur after we pop the stash. This is essentially the same behavior as SVN update. Here's the alias:

git config alias.update "!git reset --soft origin/master && git stash && git pull && git stash pop"

Of course this would mean that we lose the commit messages of any local commits, but that wouldn't be a problem if we also implemented the next step:

Some very pushy commits


Next up, I would imitate svn commit by automatically pushing after each commit by configuring up this post-commit hook (in .git/hooks/post-commit):

#!/bin/sh
git push

Voila. Every commit is immediately pushed away, and if there are remote changes, you have to update first. Just like in SVN.

(I know SVN will allow you to commit anyway if there are no conflicts, but I don't think that's a good idea. We could actually script our way into this too, by detecting the rejection, doing an update, and then commit again with the same commit message unless there are any conflicts.).

That's it

These two adaptions would actually go a long way, until the user wants to get into stuff like branching, and by that time he should be ready to learn to use Git properly anyway.

We could maybe alias git add to git resolve. We could also maybe wrap Git in a script and somehow capture the output messages to make them more readable and SVN-like. In the end, I guess there's also a limit to how much energy we should put into making training wheels for people..

Oh, on a final note, there's an old related project called EasyGit. They wrap Git itself and provide a smoother experience for people coming from SVN. They're not as hardline as my two adaptions above though.

Popular posts from this blog

Encrypting and Decrypting with Spring

I was recently working with protecting some sensitive data in a typical Java application with a database underneath. We convert the data on its way out of the application using Spring Security Crypto Utilities. It "was decided" that we'd be doing AES with a key-length of 256, and this just happens to be the kind of encryption Spring crypto does out of the box. Sweet!

The big aber is that whatever JRE is running the application has to be patched with Oracle's JCE in order to do 256 bits. It's a fascinating story, the short version being that U.S. companies are restricted from exporting various encryption algorithms to certain countries, and some countries are restricted from importing them.

Once I had patched my JRE with the JCE, I found it fascinating how straight forward it was to encrypt and decrypt using the Spring Encryptors. So just for fun at the weekend, I threw together a little desktop app that will encrypt and decrypt stuff for the given password and sa…

Managing dot-files with vcsh and myrepos

Say I want to get my dot-files out on a new computer. Here's what I do:

# install vcsh & myrepos via apt/brew/etc
vcsh clone https://github.com/tfnico/config-mr.git mr
mr update

Done! All dot-files are ready to use and in place. No deploy command, no linking up symlinks to the files. No checking/out in my entire home directory as a Git repository. Yet, all my dot-files are neatly kept in fine-grained repositories, and any changes I make are immediately ready to be committed:

config-atom.git
    -> ~/.atom/*

config-mr.git
    -> ~/.mrconfig
    -> ~/.config/mr/*

config-tmuxinator.git  
    -> ~/.tmuxinator/*

config-vim.git
    -> ~/.vimrc
    -> ~/.vim/*

config-bin.git   
    -> ~/bin/*

config-git.git          
    -> ~/.gitconfig

config-tmux.git  
    -> ~/.tmux.conf    

config-zsh.git
    -> ~/.zshrc

How can this be? The key here is to use vcsh to keep track of your dot-files, and its partner myrepos/mr for operating on many repositories at the same time.

I discovere…

Always use git-svn with --prefix

TLDR: I've recently been forced back into using git-svn, and while I was at it, I noticed that git-svn generally behaves a lot better when it is initialized using the --prefix option.

Frankly, I can't see any reason why you would ever want to use git-svn without --prefix. It even added some major simplifications to my old git-svn mirror setup.

Update: Some of the advantages of this solution will disappear in newer versions of Git.

For example, make a standard-layout svn clone:

$ git svn clone -s https://svn.company.com/repos/project-foo/

You'll get this .git/config:

[svn-remote "svn"]
        url = https://svn.company.com/repos/
        fetch = project-foo/trunk:refs/remotes/trunk
        branches = project-foo/branches/*:refs/remotes/*
        tags = project-foo/tags/*:refs/remotes/tags/*

And the remote branches looks like this (git branch -a):
    remotes/trunk
    remotes/feat-bar

(Compared to regular remote branches, they look very odd because there is no remote name i…

The Best Log Viewer Ever

This is what it looks like when I want to have a look through the logfile, to see what a user did on one of our machines one day:


Read the whole story about how it works on the Viaboxx Systems blog (and upvote on DZone!).

Microsoft ups their Git efforts another notch

This week Microsoft announced first class Git support embedded in the coming version of Visual Studio.

Now, it's not completely shocking. We could have seen it coming since Microsoft started offering Git repos on CodePlex, and more recently offering a Git client for TFS. In any case, these are some big news. Scott Hanselman weighs on some features and some more background here.

For those who are a bit unaware of what the Git situation on Windows looks like these days, I've dotted down these notes:
Some explanation on these:

msysGit has long been The Way to use Git on Windows. It's basically a port of Git itself, so it's a command-line tool.GitExtensions (includes Visual Studio integration), TortoiseGit, Git Shell, posh-git and most other tools are powered by msysGit.libgit2 is a native library for doing Git stuff. It is developed completely separate from Git itself. The above tools could (and should) probably use libgit2 instead of hooking onto and around msysGit.Github…