OmbuLabs Blog

The Lean Software Boutique

Articles by Sirko Sittig

Follow @schmierkov on Twitter

Use session variables to optimize your user flow

Sessions provide you a nice little data storage feature where the application does not need to get the information directly from the database. So you do not have to persist data in your database and can easily store info about the user on the fly. This is a nice way to enhance the user experience on your page.

Let's say that you want to show some users a new fancy sign up form and the rest the old form. If you store the version of the sign up form in a session variable, you don't need to persist this info in your database.

Read more »

Protect your sensitive data in Git

If you are working with open source or if you are going to open source a repository, you should ensure that none of your sensitive data (API Keys, Credentials, Passwords) can be accessed by anyone.

One thing that a lot of people forget, is that this information stay forever in your repository history, if you do not rewrite the history of your repository.

For instance, what usually happens is that you commit a file with sensitive information. In this Example I added accidentally my ssh-key to the repo:

$ git commit -am 'init git repo'
[master (root-commit) 917a1e1] init git repo
 2 files changed, 52 insertions(+)
 create mode 100644 id_rsa
 create mode 100644 id_rsa.pub

After doing a couple of additions, working and editing, I realise that I should never have commited the ssh-key. *facepalm*

Alright, then I just do a simple git rm --cached id_rsa and everything is back to normal. I also add this file to a .gitignore, so that this cannot happen in the future anymore.

(master) $ git rm --cached id_rsa
rm 'id_rsa'
(master) $ git status
A  .gitignore
D  id_rsa
(master) $ git commit -am 'remove id_rsa'
[master c69deb9] remove id_rsa
 2 files changed, 1 insertion(+), 51 deletions(-)
 create mode 100644 .gitignore
 delete mode 100644 id_rsa

So if we now have a look in our commit list, we can still see our first commit where I added my ssh-key. If I checkout this commit, I still have the contents of my ssh-key available.

(master) $ git log
917a1e1 - init git repo (24 minutes ago) <Sirko Sittig>
(master) $ git checkout 917a1e1
((detached from 917a1e1)) $ cat id_rsa
-----BEGIN RSA PRIVATE KEY-----
MIIJKQIBAAKCAgEAoequrqsM42na3OpvBFYOpqvzJumr3/kxJTuluXbPyJzVjMXf
d/uhFUJgSqq4AJGOFLLPpQ+9jwfA+WraIxZ9R7p8LgpNdUwKsmGnUvofeD/9Rs1y
YZO8EAjl1URLJ379nN+L5KKPS/48Q4iGp57iwuGzrXLHccLyW5+Z0iMuHlKBQzPx
...

To ensure that ALL of this data gets properly removed, I need to remove this file from all the commits in the repository with git filter-branch. The command git rm --cached git rm docs is not sufficient in this case.

(master) $ git filter-branch --tree-filter 'rm -f id_rsa' HEAD
Rewrite c69deb9779a30e6335ab1a8ac1a0825cfc9302e4 (6/6)
Ref 'refs/heads/master' was rewritten

So far so good, but what about my other branches that have been created? bash (master) $ git checkout new-feature (new-feature) $ ls drwxrwxr-x 3 sirko sirko 4096 Dec 31 13:20 ./ drwx------ 56 sirko sirko 12288 Dec 31 12:37 ../ drwxrwxr-x 8 sirko sirko 4096 Dec 31 13:20 .git/ -rw-rw-r-- 1 sirko sirko 3243 Dec 31 13:20 id_rsa -rw-r--r-- 1 sirko sirko 748 Dec 31 12:41 id_rsa.pub -rw-rw-r-- 1 sirko sirko 64 Dec 31 13:20 my_document.txt

Apparently git filter-branch is applying this changes only to the current branch, which is actually not what I want. To make this work, it seems that I have to run git filter-branch in every existing branch, which makes it pretty annoying. After reading more in the (git docs)[https://git-scm.com/docs/git-filter-branch], I found that I need to apply the --all option.

(master) $ git filter-branch --tree-filter 'rm -f id_rsa' HEAD --all
Rewrite c69deb9779a30e6335ab1a8ac1a0825cfc9302e4 (7/7)
Ref 'refs/heads/master' was rewritten
WARNING: Ref 'refs/heads/master' is unchanged
Ref 'refs/remotes/origin/master' was rewritten
WARNING: Ref 'refs/remotes/origin/master' is unchanged
Ref 'refs/remotes/origin/new-feature' was rewritten
(master) $ git checkout new-feature
(new-feature) $ ll
total 44
drwxrwxr-x  3 sirko sirko  4096 Dec 31 13:41 ./
drwx------ 57 sirko sirko 12288 Dec 31 13:41 ../
drwxrwxr-x  8 sirko sirko  4096 Dec 31 13:41 .git/
-rw-rw-r--  1 sirko sirko   748 Dec 31 13:41 id_rsa.pub
-rw-rw-r--  1 sirko sirko    64 Dec 31 13:41 my_document.txt

That seems to be exactly what I want and in the end I just need to git push --all --force my changes. After doing this, all collaborators should dump their local versions and clone a fresh version from the origin.

Another alternative to working with git filter-branch is BFG which has some more nifty features.

This tool provides some commands to completely remove big files as well as passwords from your Git history. Sadly I could not get it properly working, big files are still persistent as a git object and passwords can not be deleted because they are protected by 'HEAD'. I could not really find a solution for these problems. Maybe you are more lucky!

The easiest and much simpler solution is to initialize a new git repository, after making sure to have all sensitive information removed. The downside is obviously the loss of the project's Git history.

Read more »

Adding Csrf-Protection to your Rails-Backbone App

When integrating Backbone.js in your Rails App, you might face the problem of the inability to verify the CSRF-Token.

The CSRF Protection secures your app with a token. Rails makes sure that the person who is interacting with your app is someone who started a session in your site, not some random attacker from another site. So you should not turn it off, unless you know what you are doing.

For more information on this Topic, check out the Rails Security Guide.

This problem occurs as soon as you are trying to send form data, without the CSRF-Token provided by Rails.

Started POST "/products" for 127.0.0.1 at 2015-12-16 10:06:05 -0300
Processing by ProductsController#create as JSON
  Parameters: {"product"=>{"name"=>"foo"}}
WARNING: Can't verify CSRF token authenticity
...
Completed 302 Found in 7.6ms (ActiveRecord: 0.8ms)

After this request, Rails will terminate your session and you will have to login again.

This problem is caused by your Backbone.js application, which is sending the data directly to the backend without providing the CSRF-Token.

To solve this problem you need to add the token to your Backbone request. One of the simplest solutions I came across is the following by Anton Shuvalov.

// https://github.com/shuvalov-anton/backbone-rails-sync
Backbone._sync = Backbone.sync;
Backbone.sync = function(method, model, options) {
  if (!options.noCSRF) {
    var beforeSend = options.beforeSend;

    // Set X-CSRF-Token HTTP header
    options.beforeSend = function(xhr) {
      var token = $('meta[name="csrf-token"]').attr('content');
      if (token) { xhr.setRequestHeader('X-CSRF-Token', token); }
      if (beforeSend) { return beforeSend.apply(this, arguments); }
    };
  }
  return Backbone._sync(method, model, options);
};

It grabs the CSRF-Token provided in the meta tags of your Rails application and sets it for the request header field X-CSRF-Token.

After adding this to the Backbone code, it works as expected.

Started POST "/products" for 127.0.0.1 at 2015-12-16 10:08:29 -0300
Processing by ProductsController#create as JSON
  Parameters: {"product"=>{"name"=>"foo"}}
...
Completed 200 OK in 40.6ms (Views: 0.6ms | ActiveRecord: 10.3ms)
Read more »