I have also recently switched from xapian to elasticsearch as my search engine of choice.

BTW, I am on Ubuntu 12.04, installed elasticsearch via a deb file downloaded from their site and installed sentry via pip.

Once I was all setup I wanted to enabling logging of elasticsearch to my centralised sentry instance. This was a real pain and would have taken me a lot longer to complete (if at all) if it wasn’t for ColinHebert – details. Thanks again!

Here is what I had to do to get it working:
Elasticsearch uses log4j for logging but has a yaml file to configure it instead of a properties file. I had to append this to my /etc/elasticsearch/logging.yml file:
sentry:
type: net.kencochrane.raven.log4j.SentryAppender
dsn: "http://XXXX:YYY@myserver.com/N"

I also change the first line to read:
rootLogger: DEBUG, console, file, sentry

All I did was append sentry to the list of rootLoggers.

Once the config was sorted I had to download the dependencies:
raven-3.0-20130516.191912-9.jar
raven-log4j-3.0-20130516.191940-9.jar
commons-codec-1.8.jar
jackson-core-2.2.2-20130512.040915-3.jar
slf4j-api-1.7.5.jar

and add them to /usr/share/elasticsearch/lib on the machine hosting elasticsearch.

Once that was all done restarting elasticsearch was all that remained and I was logging to sentry!

This post is about how to compile your very own copy of bootstrap! Why would you want to do that when the nice guys maintaining the package allow you to choose your own colour scheme? I am very grateful for the customise options but I wanted to be able to able to build the package myself for three reasons: 1) I am a geek and it is a geeky thing to do, 2) I wanted to be able to customise my theme and revisit it improving it as a I continue on my project (the customise page does not have a save/resume feature) and 3) I wanted to create a colour scheme that did not match the default – I didn’t just want to choose the colours and watch them cascade through the package. I wanted to change how the colours cascade through the package.

Without further delay this is what I did:
brew install node.js
You must have the excellent HomeBrew installed to run that command.

Once I had node.js installed it was a simple case of
cd twitter-bootstrap-src
npm install
make clean
make bootstrap

This uses the excellent npm package manager to install the dependencies for bootstrap and then uses make to clean and build bootstrap. Once it completed you will have a working bootstrap installation in twitter-bootstrap-src/bootsrap.

I hope this helps and if you have any questions post a comment and I will try to help!

iOS is the software that runs on the iPhone, iPod touch, iPad and AppleTV devices.  It enables apps to be installed, removed, run etc.  It also has some utility features such as battery monitoring and display etc.

In order for a developer to have an app they have written run on iOS it must be signed using a developer certificate issued by Apple.  This signing means two things – all apps are developed by people have approved and every time you build an app is built the author of the app is verified using the certificate – so nobody can claim to be author of your app or author an app in your name.

The developer certificate is not the only thing used in the signing process.  There is a provisioning profile file used.  During development and testing apps are signed using a development provisioning profile.  This profile lists the devices the app can be installed on.  To add items to the list you must use Apples provisioning center.  When you are ready to upload your app into the App Store you have to sign your app using your distribution profile.

Once you have a signed app you can install the app on your device using iTunes.

I was responsible for all of the data modelling and the migration of two sites with over 5k users from an off the shelf (undocumented) Ning platform onto a bespoke Django build.

I established standards for documentation (using UML) and introduced a language the business owners used to describe their product to others – this was the business owners first development project. I introduced the concept of a backlog and held daily standups to ensure the business owners could track progress when desired.

I introduced game mechanics and social engagement circles to help increase user engagement and provided a solid platform for the business owners to continue on with. I created the ‘cluster’ feature they are using to earn revenue and I have provided suggestions of other features (at scrum story level) with revenue making suggestions.

I provisioned the whole stack – nginx, gunicorn, celery, mysql, redis, varnish-cache. I defined and implemented the caching policy – all not logged in users see the same rendered page so there are no db or django calls after the first page load until the page is edited. I wrote every line of python code the site used to go live using libraries like django-allauth, django-model-utils, django-dajax, easy-thumbnails, pybbm and django-celery.

I then went on to build the front end game and back end leader board.

CDN usage

I moved all my static assets and media to dedicated subdomains so that the session and csrf cookies would not be transmitted whilst the assets were being requested. This was an easy win thanks to Djangos STATIC and MEDIA setup.

GZIP compression

I turned on gzip compression within nginx:

http {
...
gzip on;
gzip_http_version 1.1;
gzip_vary on;
gzip_comp_level 6;
gzip_proxied any;
gzip_types text/plain text/html text/css application/json application/x-javascript text/xml application/xml application/xml+rss text/javascript application/javascript text/x-js;
gzip_buffers 16 8k;
gzip_disable "MSIE [1-6]\.(?!.*SV1)";
...
}


Expires and cookie turn off

I set far future expiration dates on my static content and disabled cookies for the urls:

server {
listen 80;
...
location ~* \.(?:ico|gif|css|js|jpe?g|png)$ {
access_log off;
expires max;
add_header Pragma public;
add_header Cache-Control "public"; break;
}
...
}


JS and CSS minify

I used jezdez‘s great django_compressor library to minify my js and css. This basically read my JS and CSS declarations and imported them into a single file for each. I had to tweak this due to possible name conflicts but it worked well in the end.

Next steps

The next thing for me to do is to use more sprite sheets instead of seperate images but there are not too many images anyway so I am going to stop for now.

Next post

In order to improve performance more I introduced varnish-cache to my stack. I had to do a lot of reading to come up with a working setup and will post that in another blog post coming soon – I need to finish my testing first!

It is written in Python and requires you to write Python code to use it.  Python differs from PHP.  PHP comes with most of the libraries you need out of the box and then has PEAR and PHAR to supply the rest.  Python is a little more fragmented.  With Python, you have to install the library and then install the python bindings, which is the python code that allows you to communicate with the libraries.  For example, if you want to use mysql you must install mysql and then the mysql python bindings.  This can be confusing and you can end up in a world of dependency nightmare!

With libraries like pip, vagrant and virtualenv and PaaS like dotCloud and Heroku I do not think it is such a pain anymore.  These problems can now be solved once and shared across projects.  For my own development (and sanity) I have put together a bootstrap vagrant file which allows me to build a virtual server.  It using vagrant, which in turn uses VirtualBox, to build a virtual server and then uses Chef to install the recipes (which are individual dependencies).  These recipeies include nginx, postgres etc.  At the end of a couple of commands you end up with a working virtual server because somebody else has saved a working set for you.  Great eh!?!

Here is my git repo.  To use this all you need to do it install git and VirtualBox and then run the following commands in a terminal window:

[code lang="bash" ]

git clone https://github.com/ineedarobot/geodjango-env

cd geodjango-env

vagrant up

[/code]

Once it is up you can login using vagrant ssh and your server is ready! Any questions please post a comment and I will try to help.  If you spot any issues with my setup feel free to post an issue on the github repo.

BTW, this is actually a working setup for geodjango, meaning it installs Postgres, postgis and the gis libraries needed.

I was getting the following error:

File “/Library/Python/2.6/site-packages/Django-1.3-py2.6.egg/django/db/backends/mysql/base.py”, line 14, in
raise ImproperlyConfigured(“Error loading MySQLdb module: %s” % e)
django.core.exceptions.ImproperlyConfigured: Error loading MySQLdb module: dlopen(/Users/eamonnfaherty/.python-eggs/MySQL_python-1.2.3-py2.6-macosx-10.6-universal.egg-tmp/_mysql.so, 2): no suitable image found. Did find:
/Users/eamonnfaherty/.python-eggs/MySQL_python-1.2.3-py2.6-macosx-10.6-universal.egg-tmp/_mysql.so: mach-o, but wrong architecture

This was a complete mix up between 32bit and 64bit Python libraries working together with MySQL.

To fix it I put the following in my .profile file

export PATH=$PATH:/usr/local/mysql/bin

export CC=”gcc-4.0″
export CXX=”g++-4.0″

export PYTHONPATH=”/opt/local/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/”
export DYLD_LIBRARY_PATH=/usr/local/mysql/lib/
export ARCHFLAGS=’-arch i386′
export VERSIONER_PYTHON_PREFER_32_BIT=yes

After adding this I ran a clean, build and install on the mysql-python library and everything worked! Phew!

I really like the stricter typing of signals compared to events.  For me, strict typing means that my code is checked by the compiler (instantly, at compile time) as apposed to being checked by the AVM (eventually, at runtime).  I have had countless times where I have seen code fail due to the event types declared in the handler method being wrong, even when the string values are stored as variables instead of using literals. The checking that event handlers are added correctly is not checkable by the compiler.  This means we need to either blindly hope it all works out or rely on testing.

Signals (out of the box) fixes this issue but there is still another issue for me, the contract between an event (also a signal) and its handler.  With events there are two contracts, the event type (the Event.type String) and the event data.  Signals (out of the box) handles the event type issue by moving the name of the type into an interface as a signal, which is a great idea!  The data problem is harder to solve.

There are two checks that should be performed on the data.

The first is validation. At the moment this is limited to type checking.  When you set up a signal, you specify that the signal should be an int, String etc.  I think this is a great start but I think there should be more (optional) validation checks available; range checks for ints and regex checks for Strings.  When these validations fail then the signal should have a built in mechanism to throw an InvalidSignalError.  This would tighten up the robustness of projects using signals and would definitely reduce the amount of code I write.

The second check is verification. This is a lot harder to check but is a much more valuable check to perform. I think this is up to developers to check when developing. I am currently using signals to send messages between objects. I like to keep the implementation of my objects hidden from others so when writing a class I am always mindful of what values are available to other objects. So, when writing a signal I often find myself doing the following:

[code lang="as3" ]
public class ShowScreenSignal extends Signal
{
private static const MENU:int = 0;
private static const PLAY:int = 1;
...

public function ShowScreenSignal()
{
super(int);
}

public function dispatchMenu():void
{
dispatch(MENU);
}

public function dispatchPlay():void
{
dispatch(PLAY);
}

...

public function isMenu(message:int):Boolean
{
return message = MENU;
}

public function isPlay(message:int):Boolean
{
return message = PLAY;
}

...
}
[/code]

This means in my controller/mediator I call signal.dispatchPlay() and in my listener/command I check signal.isPlay(message) to check the message. This means nothing knows about the message values other than the signal. I can change the message to be a String without changing any of my controllers/mediators/command. This is better encapsulation. This works really well when using RobotLegs + Signals as you can subtype your command classes and have a condition in your execute method if (message.isPlay(message)).

This is solving one problem. I want to have one signal with many handlers based on the message. eg:

[code lang="as3" ]
public class ShowPlayCommand extends SignalCommand
{
[Inject]
public var message:int;

[Inject]
public var signal:ShowScreenSignal;

override public function execute():void {

if (!signal.isPlay(message)) {
return;
}

//do something
}
}
[/code]

Here is the PHP I used to get the feed of mypage.

[code lang="php" ]
<?php

//config
$appId = "999";
$appSecret = "I_LOVE_ROBOTS";
$feedURL = "https://graph.facebook.com/mypage/feed";
$limit = 10;

//script
$accessTokenUrl = "https://graph.facebook.com/oauth/access_token";
$parameters = "client_id=$appId&client_secret=$appSecret&grant_type=client_credentials";
$accessToken = file_get_contents($accessTokenUrl .  "?" . $parameters);

$parameters = "$accessToken&limit=$limit";
$result = file_get_contents($feedURL . "?" . $parameters);
?>
[/code]