pawlik

/home/greg/.rvm/gems/ruby-1.8.7-p358/gems/sequel-3.20.0/lib/sequel/adapters/postgres.rb:211: [BUG] unknown type 0x22 (0xc given)
ruby 1.9.3p194 (2012-04-20 revision 35410) [x86_64-linux]

-- Control frame information -----------------------------------------------
c:0034 p:---- s:0153 b:0153 l:000152 d:000152 CFUNC  :initialize
c:0033 p:---- s:0151 b:0151 l:000150 d:000150 CFUNC  :new
c:0032 p:0089 s:0141 b:0141 l:000140 d:000140 METHOD /home/greg/.rvm/gems/ruby-1.8.7-p358/gems/sequel-3.20.0/lib/sequel/adapters/postgres.rb:211
c:0031 p:0012 s:0134 b:0134 l:000768 d:000133 BLOCK  /home/greg/.rvm/gems/ruby-1.8.7-p358/gems/sequel-3.20.0/lib/sequel/database/misc.rb:45
c:0030 p:---- s:0131 b:0131 l:000130 d:000130 FINISH
c:0029 p:---- s:0129 b:0129 l:000128 d:000128 CFUNC  :call

pawlik

# file extract_addresses_from_companies

load 'spec/spec_helper.rb'
load 'db/migrate/20121109212148_extract_addresses_from_companies.rb'

describe ExtractAddressesFromCompanies do
  before do
    @previous_migration_version = '20121108165838'
    @my_migration_version = '20121109212148'
  end

pawlik

"Write safe migrations" approach.

This is what I've mentioned on our call. Let me prove this could work and please prove me wrong or point weak spots.

Assumptions

• We can agree on how many steps we want to be able to rollback to (capistrano supports it, now on webstore is 2. I think increasing this too much makes no sense. Personally I think 2 is perfect - can't imagine that we release 1.19, after two weekes we release 1.20, after another 2 weeks 1.21 and at this point PO decides we want to rollback to version from a month ago).
• don't think that database migrations is about just database migrations - it's also about writing code in a way it can handle this migrations.
• future actions can be easily managed with something similar to roadsigns. For migrations, in case you need column to be removed in some unspecified future you can leave file CHECK_THIS_BEFORE_MAKING_NEW_MIGRATION, with entry 2015-01-12 prepared for column foo removal. Please remove after at least 5 releases (see first assumpti

pawlik

cache_dir="/tmp/git-cache/"
if [ ! -d "$cache_dir" ]
then
  mkdir -p $cache_dir
fi
cache_path=$cache_dir$1
if [ ! -d "$cache_path" ]
then
  git clone --mirror $1 $cache_path
else

pawlik

Recent code review dragged my attention to repeatedly delegated methods. (For nexway teammates https://github.com/NexwayGroup/N3/pull/2352/files#diff-1333ef8b1eee5de57912fa4ed14919dbR34)

example:

public function getBillingAddress()
{
  return $this->getOrder()->getBillingAddress();
}

pawlik

👍 👎 ✈️ 🎨 🐻 🍺 🚲 💣 📖

pawlik

package main

import (
	"fmt"
	io "io/ioutil"
	"os"
	"flag"
)

func usage() {

pawlik

<?php
class Tappable {
  private $whatever = null;

  public function __construct($whatever) {
    $this->whatever = $whatever;
  }


  public function tap(callable $callback) {

pawlik

goal: give helper A possibility to inject 'tax helper', if none injected - Mage::helper('tax') should be used by default.


both ways work:

1)
```php
    /**
     * @var Mage_Tax_Helper_Data
     */

pawlik

http://johnsythinks.com/hex/

Thanks to http://www.javascripter.net/faq/hextorgb.htm

function hexToR(h) {return parseInt((cutHex(h)).substring(0,2),16)}
function hexToG(h) {return parseInt((cutHex(h)).substring(2,4),16)}
function hexToB(h) {return parseInt((cutHex(h)).substring(4,6),16)}
function cutHex(h) {return (h.charAt(0)=="#") ? h.substring(1,7):h}

this is my quick resolve function: