Source control is all over the place these days, with git becoming a primary tool, not just for managing development, but for moving code from place to place.

For a simple example of pushing code from place to place, what better place to start than with this Jekyll site, which I’m using to write about the process in the first place?

First, we have to get started with the installs to use Jekyll on the remote machine. I’m borrowing from other sources here, but note that the -V at the end helps to let you know that the install is doing work and not just frozen. For me, this is going on a Debian distribution.

sudo apt-get install ruby ruby-dev make gcc nodejs
sudo gem install jekyll --no-rdoc --no-ri -V

Now github offers a fairly easy way to bootstrap our repo. The .git for this very Jekyll repo is just at the github repo. To do all this, I am trying to loosely follow the instructions from git-scm.com, but with the exception that all the basic certs have already been set up. Given all this, I clone the git repo on the server, and then add the remote on my own machine. You’ll have to mentally substitute in the specific usernames and addresses.

git remote add my_server username@server_ip_address:/opt/git/AlanCoding.github.io

So here I go, I’m going to make one commit, push it directly to this server, then after that I’ll make another commit and push this to github.

…

Well, that didn’t quite go off without any problems. It turned out there was a reason that tutorial started off with a bare repository. I opted to just set the config variable:

git config receive.denyCurrentBranch ignore

After that, and a “push my_server master”, it accepts the push. Then ssh in, and “git log”, and we have the last commit I just made.