cluesque

<%

require 'cgi'
require 'uri'

begin
  uri = URI.parse(ENV["DATABASE_URL"])
rescue URI::InvalidURIError
  raise "Invalid DATABASE_URL"
end

cluesque

require 'test_helper'

class ShouldaContextTest < ActiveSupport::TestCase
  def self.setup_setup_helper
    setup { puts "running the setup helper"; @wrapped = true }
    teardown{ puts "running the teardown helper"; @wrapped = false }
  end
  context "outer nesting" do
    context "without setup" do
      setup do

cluesque

# A migration that will continue to work after refactoring
class FutureProofMigration < ActiveRecord::Migration
  # Redefine models used in the migration as empty inner classes.  This removes all validations and magic.
  class Model < ActiveRecord::Base; end

  # Either don't rely on associations or declare them as referring to the new inner class
  class DependentModel < ActiveRecord::Base
    belongs_to :model, :class_name => "FutureProofMigration::Model"
  end

cluesque

# Include this in your development and test environments
# if you want to run all Resque processors inline 
# (avoiding the need for Redis and a separate worker process)

# In config/environments/{development,test}.rb:
# config.after_initialize do
#   require 'resque_inline'
# end

module Resque

cluesque

traceroute to 216.81.59.173 (216.81.59.173), 64 hops max, 52 byte packets
...
 8  100gigabitethernet11-1.core1.ash1.he.net (184.105.223.165)  14.978 ms
 9  10gigabitethernet1-2.core1.atl1.he.net (184.105.213.110)  33.392 ms
10  216.66.0.26 (216.66.0.26)  29.782 ms
11  *
12  episode.iv (206.214.251.1)  120.330 ms
13  a.new.hope (206.214.251.6)  73.192 ms
14  it.is.a.period.of.civil.war (206.214.251.9)  98.692 ms
15  rebel.spaceships (206.214.251.14)  68.083 ms

cluesque

d5toqvrtbm8bbu=> SELECT id, created_at FROM widgets WHERE id = 1155301;
   id    |         created_at         
---------+----------------------------
 1155301 | 2014-01-10 02:59:47.856214
(1 row)

d5toqvrtbm8bbu=> CREATE UNIQUE INDEX widgets_pkey ON widgets(id);
CREATE INDEX
d5toqvrtbm8bbu=> SELECT id, created_at FROM widgets WHERE id = 1155301;
 id | created_at 

cluesque

class Cryptor
  # Usage:
  # cipher = Cryptor.new(key, auth_data: '1234').encrypt(clear)
  # clear = Cryptor.new(key, auth_data: '12344').decrypt(cipher)
  attr_accessor :key, :auth_data

  # The auth_data should be something different for each text, perhaps a database row id?
  def initialize(key, auth_data: 'aad')
    self.mode = mode
    self.key = key

cluesque

04ac10d9df868268f354c546fb71262375a7fa654841cebc1ce5e3269d2f69de9c614b2235c2311d23e26f55a5c38601e240f0168c8ecc7255b2e940f50fa783b8

cluesque

Keybase proof

I hereby claim:

• I am cluesque on github.
• I am cluesque (https://keybase.io/cluesque) on keybase.
• I have a public key ASBvxQlEpGPq_kkRyixFNNvyPL7H1ag6fwoz4Dmb40Nsrgo

To claim this, I am signing this object:

cluesque

DHH's On Writing Software (well?) [episode 2](https://www.youtube.com/watch?v=m1jOWu7woKM) discussed callbacks, specifically the practice of hiding side effects of object creation using `after_save` and `after_commit` hooks.

It implicitly conceeded that `after_commit`, lacking information about what was just committed, was flawed, and uses the common workaround of using an `after_save` to decide to do something based on changes and set an @ivar, and then in the `after_commit` acting on the @ivar found.  Which is fun, but.

It also [at the end](https://youtu.be/m1jOWu7woKM?t=1151) discussed the possibility that in some use cases a caller might want to explicitly prevent side effects from happening.  Think about when audited temporarily suppresses auditing, or perhaps one might want to bulk update some data, without sending the email normally sent when said data is updated.

It teased the presence of a `Suppressible` mixin.  For fun, I wrote it.

Update: turns out the real one is public.  It uses `thread_mattr