28 Apr, 2013
3 commits
25 Apr, 2013
2 commits
-
Use https urls for gems fetched from git
10 Apr, 2013
2 commits
-
Add a line reminding Heroku deployers to edit seeds.rb before pushing and running rake db:seed
09 Apr, 2013
2 commits
-
Commenting on an error when there is only one user raises an exception
08 Apr, 2013
5 commits
-
Include Code Climate Score in README.md Badge
-
updates the image in the README.md to include the current code climate score.
-
Test coverage for comment observer [Fixes #452]
-
Ruby 2.0 gem update
26 Mar, 2013
2 commits
-
Update the Airbrake domain in the README
21 Mar, 2013
4 commits
-
Capistrano: deploy from fixed branch
-
Not the 'current' branch in the local clone fo the repo. I feel like deploying from the current branch in the local clone is a bit presumptous. I expect most people want to deploy from one branch while allowing for development of several feature branches. Right now you always have to remember to check out master before deploying or you may deploy your feature branch (which may not even be pushed).
-
Notice API: properly transform <var> tags
-
<var> tags with blank values were ending up completely wrong in the resulting hash. This: (a user with a blank name) <var key="user"> <var key="id">123</var> <var key="name"/> </var> Gets parsed by hoptoad_notifier into this: { 'key' => 'user', 'var' => [{ 'key' => 'id', '__content__' => '123' },{ 'key' => 'name' }] } Which when passed through the "rekey" method, ended up like this: { 'user' => { 'id' => '123', 'key' => 'name' # bad and wrong } } Now, after these changes, it correctly comes through as: { 'user' => { 'id' => '123', 'name' => nil # much better! } }
20 Mar, 2013
2 commits
19 Mar, 2013
1 commit
18 Mar, 2013
2 commits
-
Update to Rails 3.2.13
13 Mar, 2013
4 commits
-
Removed typo in comment deletion confirm dialog
-
There was a "you" missing in the sentence.
12 Mar, 2013
7 commits
10 Mar, 2013
4 commits
-
Fix padding between pagination links
-
Huge Performance Improvements
-
Enable the identity map for Mongoid. This allows subsequent finds for the same object to skip the DB and use an in-memory store of all retrieved records. In many cases this reduces the number of DB queries from _hundreds_ to a few tens. This is particularly noticable when showing lists of problems. Previously each `problem.app` reference caused a Apo.find() using the `app_id`, now it's an in-memory lookup.
