Scott Tesoriere

Rails — A faster way for next and previous links on a post, article, or any model

2012-01-12T00:00:00+01:00

This is a simple, but common question I come across when dealing with people learning rails, as they inevitably do a bloggish type app first.

I assume you have a model like the one below

class CreateArticles < ActiveRecord::Migration
  def change
    create_table :articles do |t|
      t.string :title
      t.text :body

      t.timestamps
    end
  end
end

Adding the next/prev links

Personally, I prefer to add instance methods as well as a scope or a class method. Check it out below:

class Article < ActiveRecord::Base
    scope :next, lambda {|id| where("id > ?",id).order("id ASC") } # this is the default ordering for AR
    scope :previous, lambda {|id| where("id < ?",id).order("id DESC") }

    def next
      Article.next(self.id).first
    end

    def previous
      Article.previous(self.id).first
    end

    #### used below to benchmark
    # scope :next_by_date, lambda {|created_at| where("created_at > ?",created_at).order("created_at ASC") } # this is the default ordering for AR
    # scope :previous_by_date, lambda {|created_at| where("created_at < ?",created_at).order("created_at DESC") }

    # def next_by_date
    #  Article.next_by_date(self.created_at).first
    # end

    # def previous_by_date
    #  Article.previous_by_date(self.created_at).first
    # end

end

How it works

If you're confused, don't be! Let's assume we have an array of article ids: [1,2,4,6]. If we're looking for the next article for 2, it would obviously be 4. Programmatically, you can't just say Article.find(id+1), because it won't exist in some situations where articles are deleted.

We have to find the next id above 2, hence where("id > ?", id). However, if you order the ids descending you'll get 6, because 6 is the highest in the list while still being larger than 2. Knowing that, we then sort the list ascending, because 4 will be first on the list when ordering ascending.

This, of course, assumes you're using the instance methods as I did, if you used .last instead, the ordering for each would be reversed.

It's slightly faster than sorting by a date, or any other field not indexed

Since you're sorting by the primary key, it should be much faster than sorting by any other field, as the primary key is usually indexed in most databases. Plus, it's computationally easier to sort an integer, however slightly.

For example:

sql = "INSERT INTO articles (title, body, user_id, created_at) VALUES "
array = []
n = 100000
n.times do |i| 
  array << "('#{i}','#{i}', 1, '#{eval("#{i}.days.ago")}')"
end
sql += array.join(",")
ActiveRecord::Base.connection.execute(sql)
Benchmark.bm do |x|
  x.report { Article.first.next_by_date }
  x.report("by date: ") { Article.first.next_by_date }
  x.report { Article.first.next }
  x.report("by id: ") { Article.first.next }
end

I ran each twice just incase some caching occurred, and the results were:

      user     system      total        real
by date:   0.000000   0.000000   0.000000 (  0.017141)
by id:     0.000000   0.000000   0.000000 (  0.000682)

Changing the repetition will demonstrate a linear pattern, say O(n), for finding the next by, whereas by date will demonstrate a more exponential pattern, perhaps nearing O(n^2).

The awkward things Siri says

2011-10-16T00:00:00+02:00

A quick example of how easy nodejs easy to get up and running

I'll go into detail later, but essentially, this site uses nodejs, expressjs, railwayjs, knox for s3 uploads, a redis session store, and a mongodb backend via mongoose.

Site can be found here.

Questions?

You might have some questions, and I can answer them! I'll post some how-to's soon.

Node.js — Getting oAuth Up and Running Using Express.js and Mongoose

2011-10-10T00:00:00+02:00

Getting Started

If you're interested in how to incorporate this into railwayjs, see this other article instead. I normally use coffeescript, but for the purposes of this article, I converted it to javascript. For the impatient, this project can be found in my github repos.

Let's assume our app.js/server.js looks like this:

var express = require('express');
var app = express.createServer();

app.get('/', function(req, res){
  res.send('hello world');
});

app.configure(function() {
  app.set('views', __dirname + '/app/views');
  app.set('view engine', 'jade');
});

app.listen(3001);

Now, getting oAuth up and running can be easy, but you have to make sure you have mongodb up and running first, since we're using mongoose-auth. You could very well use everyauth, which mongoose-auth is based upon, though it would require a little more effort if you were to change to another database. This is pretty easy: in Ubuntu, you can just apt-get install mongodb, or in osx, brew install mongodb;. If you need to install homebrew for OSX, /usr/bin/ruby -e "$(curl -fsSL https://raw.github.com/gist/323731)".

Working with MongoDB

Let's connect to mongo:

var mongoose = require('mongoose')
  , Schema = mongoose.Schema
  , ObjectId = mongoose.SchemaTypes.ObjectId;
  
mongoose.connect('mongodb://localhost/example');

Incorporate mongoose-auth

To get this working, we have to add the following to our app.js:

var conf = require('./config/oauth_providers');
var UserSchema = new Schema({})
    , User;
var mongooseAuth = require('mongoose-auth');

UserSchema.plugin(mongooseAuth, {
  everymodule: {
    everyauth: {
      User: function() {
        return User;
      }
    }
  },
  facebook: {
    everyauth: {
      myHostname: 'http://local.host:3001',
      appId: conf.fb.appId,
      appSecret: conf.fb.appSecret,
      redirectPath: '/'
    }
  },
  twitter: {
    everyauth: {
      myHostname: 'http://local.host:3001',
      consumerKey: conf.twit.consumerKey,
      consumerSecret: conf.twit.consumerSecret,
      redirectPath: '/'
    }
  },
  github: {
    everyauth: {
      myHostname: 'http://local.host:3001',
      appId: conf.github.appId,
      appSecret: conf.github.appSecret,
      redirectPath: '/'
    }
  }
});

mongoose.model('User', UserSchema);

mongoose.connect('mongodb://localhost/example');

User = mongoose.model('User');

The configuration for this file resides in a different folder, I made a config folder and put the oauth_providers.js file in there, it looks like this:

module.exports = {
  fb: {
    appId: '111565172259433',
    appSecret: '85f7e0a0cc804886180b887c1f04a3c1'
  },
  twit: {
    consumerKey: 'JLCGyLzuOK1BjnKPKGyQ',
    consumerSecret: 'GNqKfPqtzOcsCtFbGTMqinoATHvBcy1nzCTimeA9M0'
  },
  github: {
    appId: '11932f2b6d05d2a5fa18',
    appSecret: '2603d1bc663b74d6732500c1e9ad05b0f4013593'
  },
  instagram: {
    clientId: 'be147b077ddf49368d6fb5cf3112b9e0',
    clientSecret: 'b65ad83daed242c0aa059ffae42feddd'
  },
};

NOTE: these keypairs really shouldn't be made public, but they exist in the mongoose-auth repo, so you can test using them. This also assumes you have something setup in /etc/hosts to map local.host to 127.0.0.1, example: 127.0.0.1 local.host

In addition, we have to configure express a little differently:

app.configure(function() {
  app.set('views', __dirname + '/app/views');
  app.set('view engine', 'jade');
  app.use(express.bodyParser());
  app.use(express.cookieParser());
  app.use(express.session({secret: 'secret'}));
  app.use(mongooseAuth.middleware());
});

mongooseAuth.helpExpress(app);

The above sets up cookies, adds a session key, hooks into mongooseAuth to the app, and exposes everyauth's routes to express' views.

Logging in

Now, just by visiting http://local.host:3000/auth/twitter, for example, and logging in, you'll be redirected back to the homepage. To check to make sure it's working, we have to add some views, make a folder called views and place the following in it named index.jade

- var items = ["facebook", "github", "twitter", "instagram"]
- if (!everyauth.loggedIn)
  h2 Not Authenticated
  each item in items
   a(href='/auth/' + item)
     span Connect with <span style="text-transform: capitalize">!{item}</span><br />

- else
  h2 Authenticated
  #user-id Logged in with `user.id` #{user.id} - aka `everyauth.user.id` #{everyauth.user.id}
  - if (everyauth.facebook)
    h3 Facebook User Data
    p= JSON.stringify(everyauth.facebook.user)
  - if (everyauth.twitter)
    h3 Twitter User Data
    p= JSON.stringify(everyauth.twitter.user)
  - if (everyauth.github)
    h3 GitHub User Data
    p= JSON.stringify(everyauth.github.user)
  - if (everyauth.instagram)
    h3 Instagram User Data
    p= JSON.stringify(everyauth.instagram.user)
  h3
    a(href='/logout') Logout

And don't forget to change the route, to

app.get('/', function(req, res){
    res.render('index');
});

And if you've been following, you'll see that all the twitter metadata is displayed. You're done! There's a lot of other things you can do, like being able to link each different account, whether or not to remember the user, and if anyone needs help with that, I can show you how to do it.

This project can be found in my github repo, expressjs & oauth.

Node.js — Getting oAuth Up and Running Using Express.js, Railway.js and Mongoose

2011-10-10T00:00:00+02:00

Let's get started

If you're interested in how to incorporate this without using railwayjs, see this other article instead.

This is assuming you already have a working railwayjs project working. If not:

sudo npm install railway -g
rw i blog && cd blog
npm install -l
rw g crud post title content
# make sure mongodb is running
rw s 3000
open http://localhost:3000/posts

If you need to install mongodb, it's simple. In Ubuntu, you can just apt-get install mongodb, or in osx, brew install mongodb;. If you need to install homebrew for OSX, /usr/bin/ruby -e "$(curl -fsSL https://raw.github.com/gist/323731)".

Setting up the environment

First things first, change your package.json to include mongoose-auth.

{ "name": "railwayjs project"
, "version": "0.0.1"
, "engines": ["node >= 0.4.0"]
, "main": "server.js"
, "dependencies":
  { "jade-ext":  ">= 0"
  , "jade":      ">= 0"
  , "express":          ">= 2.2.2"
  , "connect":          ">= 1.4.2"
  , "railway":          ">= 0.1.6"
  , "yaml":             ">= 0.1.2"
  , "coffee-script":    ">= 1.1.1"
  , "mime":             ">= 1.2.2"
  , "qs":               ">= 0.1.0"
  , "mongoose":         ">= 1.3.6"
  , "mongodb":          ">= 0.9.4-5"
  , "connect-mongodb":  ">= 0.3.0"
  , "mongoose-auth":    ">= 0.0.2"
  }
, "scripts":
  { "test": "nodeunit test/*/*"
  }
}

Next, we're going to change our config/environment.js to look like the following:

var express = require('express');
var mongooseAuth = require('mongoose-auth');

require('./mongoose_oauth');

app.configure(function() {
  var cwd;
  cwd = process.cwd();
  app.set('views', cwd + '/app/views');
  app.set('view engine', 'jade');
  app.use(express.static(cwd + '/public', {
    maxAge: 86400000
  }));
  app.use(express.bodyParser());
  app.use(express.cookieParser());
  app.use(express.session({
    secret: 'secret'
  }));
  // MAKE SURE THIS IS COMMENTED OUT, otherwise it will produce errors that are mostly nonsensical
  //app.use(app.router)
  app.use(express.methodOverride());
  app.use(mongooseAuth.middleware());
});
mongooseAuth.helpExpress(app);

Notice the use of the require statement, require('./mongoose_oauth');, this files contents are here and look like:

var mongoose = require('mongoose')
  , Schema = mongoose.Schema;

var conf = require('./oauth_providers');

var UserSchema = new Schema({})
  , User;

mongooseAuth = require('mongoose-auth');

UserSchema.plugin(mongooseAuth, {
  everymodule: {
    everyauth: {
      User: function() {
        return User;
      }
    }
  },
  facebook: {
    everyauth: {
      myHostname: 'http://local.host:3001',
      appId: conf.fb.appId,
      appSecret: conf.fb.appSecret,
      redirectPath: '/'
    }
  },
  twitter: {
    everyauth: {
      myHostname: 'http://local.host:3001',
      consumerKey: conf.twit.consumerKey,
      consumerSecret: conf.twit.consumerSecret,
      redirectPath: '/'
    }
  },
  github: {
    everyauth: {
      myHostname: 'http://local.host:3001',
      appId: conf.github.appId,
      appSecret: conf.github.appSecret,
      redirectPath: '/'
    }
  }
});
User = mongoose.model('User', UserSchema);
module.exports["User"] = mongoose.model("User");
module.exports["User"].modelName = "User";

Also notice, the configuration file, config/oauth_providers.js, this looks like:

module.exports = {
  fb: {
    appId: '111565172259433',
    appSecret: '85f7e0a0cc804886180b887c1f04a3c1'
  },
  twit: {
    consumerKey: 'JLCGyLzuOK1BjnKPKGyQ',
    consumerSecret: 'GNqKfPqtzOcsCtFbGTMqinoATHvBcy1nzCTimeA9M0'
  },
  github: {
    appId: '11932f2b6d05d2a5fa18',
    appSecret: '2603d1bc663b74d6732500c1e9ad05b0f4013593'
  },
  instagram: {
    clientId: 'be147b077ddf49368d6fb5cf3112b9e0',
    clientSecret: 'b65ad83daed242c0aa059ffae42feddd'
  },
};

The views

Now, if you were to run your app, rw s 3000, and go to http://local.host:3000/auth/twitter, you would find, after logging in to twitter, it redirects you back to the homepage, as per the mongoose_oauth.js config. But how do you know if you're actually logged in, and the user is being stored in the session?

Well, we have to update a view to contain the following information:

- var items = ["facebook", "github", "twitter", "instagram"]
- if (!everyauth.loggedIn)
  h2 Not Authenticated
  each item in items
   a(href='/auth/' + item)
     span Connect with <span style="text-transform: capitalize">!{item}</span><br />

- else
  h2 Authenticated
  #user-id Logged in with `user.id` #{user.id} - aka `everyauth.user.id` #{everyauth.user.id}
  - if (everyauth.facebook)
    h3 Facebook User Data
    p= JSON.stringify(everyauth.facebook.user)
  - if (everyauth.twitter)
    h3 Twitter User Data
    p= JSON.stringify(everyauth.twitter.user)
  - if (everyauth.github)
    h3 GitHub User Data
    p= JSON.stringify(everyauth.github.user)
  - if (everyauth.instagram)
    h3 Instagram User Data
    p= JSON.stringify(everyauth.instagram.user)
  h3
    a(href='/logout') Logout

And Done!

Now you should see various metadata depending on how you're logged in. All the files above can be found in this gist.

Migrating from Rails 3.1 RC4 to RC5 using Heroku's Cedar Stack (also compass, unicorn, and sendgrid)

2011-08-08T00:00:00+02:00

A mouthful, indeed

I'm currently developing a super duper secret project which uses Rails 3.1 and is hosted on Heroku. With the release of Rails 3.1 RC5, I've decided to migrate to Heroku's Cedar stack after reading Michael's article about how easily unicorn can be implemented on the new stack.

I've already migrated one app, Wasted on Steam, and the performance enhancements are amazing, as you might expect. Using siege, it was able deal with 100 concurrent requests repeated 10 times in 2.5 seconds per request. Previously, it had taken 10 seconds per request. This is amazing, especially considering Wasted on Steam uses a network blocking request to update data.

For this other app, this super duper secret one, I thought I would detail the process.

The initial setup

Currently, there's no way to migrate the stack automatically, so we have to clone and proceed:

git clone git@heroku.com:<your_app_name>.git cedar_testing

Next, create the app on the new stack:

heroku create --stack cedar
git config heroku.remote heroku

The code changes

Gemfile fixes

Rails 3.1.rc5 separated the asset-building gems into their own group in the Gemfile, called :assets, so we need to make some changes and clean it up so it looks like this:

source 'http://rubygems.org'

gem 'rails', '3.1.0.rc5'

group :assets do
    gem 'sass-rails', "~> 3.1.0.rc"
    gem 'coffee-rails', "~> 3.1.0.rc"
    gem 'uglifier'
    gem 'compass', '0.12.0.alpha.0.91a748a', :git => 'git://github.com/chriseppstein/compass.git', :branch => 'rails31'
end

gem 'pg'
gem "kaminari"
gem "redcarpet"
gem "devise"
gem 'sprockets'
gem "cancan"
gem "mime-types"
gem "treetop"
gem "ancestry", '>= 1.2.2'
gem "nokogiri", '1.4.4'
gem "oa-oauth", :require => "omniauth/oauth"
gem 'oa-openid', :require => 'omniauth/openid'
gem 'omniauth', '>= 0.2.6'
gem 'dalli'

group :production do
    gem 'unicorn'
end


group :development, :test do
    ...
end

Miscellaneous fixes

Please note, we no longer need to include the old version of rake, nor the therubyracer (the Cedar stack now includes node.js). It's very important that you update the path to use the native node.js by typing:

heroku config:add PATH=vendor/bundle/ruby/1.9.1/bin:/usr/local/bin:/usr/bin:/bin:bin

Supposedly, this is supposed to happen on new apps by default, but I had issues. Now, make sure you run bundle update now to update the Gemfile!

Next, we need to update our application.rb; it's changed in RC5. Around line, replace with the following:

Bundler.require *Rails.groups(:assets) if defined?(Bundler)

Now, if you're using Compass like I am, you'll need to create an initializer in config/initializers/sass.rb:

Rails.configuration.sass.tap do |config|
  config.load_paths << "#{Gem.loaded_specs['compass'].full_gem_path}/frameworks/compass/stylesheets"
end

Thanks to Ken Collins for that bit! I imagine, as some point when Rails 3.1 is released this will be done automatically.

Next, in your production.rb, change the config.action_dispatch.x_sendfile_header to:

config.action_dispatch.x_sendfile_header = nil

If you use sendgrid

The Cedar stack is supposed to be less magical and black-boxy then the others, so if you use sendgrid, we have to accommodate for it in our production.rb.

config.action_mailer.raise_delivery_errors = true
config.action_mailer.smtp_settings = {
  :address        => "smtp.sendgrid.net",
  :port           => "25",
  :authentication => :plain,
  :user_name      => ENV['SENDGRID_USERNAME'],
  :password       => ENV['SENDGRID_PASSWORD'],
  :domain         => ENV['SENDGRID_DOMAIN']
}

Using unicorn instead of thin

Our Procfile

This Procfile tells Heroku exactly which process groups need to exist for various reasons. Like a web group to run our unicorn. Our Procfile is quite simple as we only need the main web process, Unicorn.

web: bundle exec unicorn -p $PORT -c ./config/unicorn.rb

The unicorn config

Place this in your config directory and call it unicorn.rb

worker_processes 4 # amount of unicorn workers to spin up
timeout 30         # restarts workers that hang for 30 seconds

You'll have to play around with the configuration depending on how memory-heavy your app is. Wasted on Steam uses a ton of gems and it can easily run 4 worker processes.

Push the changes to heroku

git add Procfile config/unicorn.rb
git commit -a -m "Updated for Heroku and Unicorn"
git push heroku master

The data

If you have any configuration variables, you'll want to go to the old app and do heroku config -s and migrate them over using heroku config:add <VARIABLE>=<DATA> <HELLO>=<MOREDATA>.

Finally, we have to pull the database from our old app. This will only work if you have the gem taps installed, so gem install taps. Go to your old app folder and type the following:

heroku db:pull sqlite://backup.sql
mv backup.sql ../cedar_testing
cd ../cedar_testing
heroku db:push sqlite://backup.sql

If you have any issues, simply type git remote rm heroku; git remote add heroku git@heroku.com:<new_app_name>.git then proceed with the steps above. This usually fixes everything.

We're done!

Hopefully, everything should work now as planned. Make sure you go to http://<yourapp>.herokuapp.com and not http://<yourapp>.heroku.com, and you should see the site!

Comparing performance

This app is a simple app, but to give you an idea of how it performs, again with 100 concurrent requests repeating 10 times:

Old & Busted

Transactions:                 995 hits
Availability:              99.50 %
Elapsed time:              62.90 secs
Data transferred:           1.35 MB
Response time:              5.26 secs
Transaction rate:          15.82 trans/sec
Throughput:             0.02 MB/sec
Concurrency:               83.19
Successful transactions:         995
Failed transactions:               5
Longest transaction:           15.67
Shortest transaction:           0.12

New hotness

Transactions:                1000 hits
Availability:             100.00 %
Elapsed time:              19.99 secs
Data transferred:           1.36 MB
Response time:              0.61 secs
Transaction rate:          50.03 trans/sec
Throughput:             0.07 MB/sec
Concurrency:               30.33
Successful transactions:        1000
Failed transactions:               0
Longest transaction:           14.04
Shortest transaction:           0.11

Many articles helped get me here

Heroku's articles on migrating helped a lot. Michael's article on using unicorn on Cedar was the inspiration, and finally, Ken's article helped me get Compass integrated. Thanks!

Random Freeze Fix for GTX 460 in 10.6 (osx86)

2011-07-23T00:00:00+02:00

Update: completely rid yourself of freezes!

While I had no more random freezes in 10.6.*, upgrading to Lion allowed for full stability and full use of graphics acceleration via OpenCL/CUDA/OpenGL. So please, upgrade to Lion, go purchase it on the App Store. http://itunes.apple.com/us/app/os-x-lion/id444303913?mt=12. There are many guides available detailing installation.

So far, anyway!

Newly released drivers for the Quadro 4000 for Mac can be used for the GTX 460, and this combination with 10.6.7 and Chameleon rc750 have completely removed any random freezes thus far. The system has been running for 24 hours now.

But it still locks up :(

Ultimately, if you still have freezes, consider going here and buying one of these. If that's not the way you want to spend your money, then go here and looking for one of these, which work without issue, without injection in OSX 10.6.7.

http://us.download.nvidia.com/Mac/Quadro_Certified/256.01.00f03/Retail_256.01.00f03v7.dmg

Wasted on Steam - an analytic tool for the Steam platform

2011-07-09T00:00:00+02:00

I was always curious...

Steam has existed for about 8 years now, and I was always curious how much time and money I've ~~wasted~~ spent with the service. I created a webapp to calculate how many hours have been spent as well as an approximate cost of an account's worth. There are gaps in steam's data, so it's not entirely accurate, but it's a good estimate. Go on, check it out! http://wastedonsteam.com

What's it made out of?

It's built using Rails 3.1, which was the other main motivator for creating this - testing out the new assets pipeline in Rails 3.1. It's hosted on heroku, uses gruff to make simple price graphs, nokogiri to update the data, the money class to store the prices, and delayed_job to push all the updating into the background.

The gemfile is below:

gem 'rails', '3.1.0.rc4'
gem 'sass'
gem 'coffee-script'
gem 'uglifier'
gem 'attempt'
gem 'gruff', :require => false
gem 'rmagick', :require => false
gem 'nokogiri', '1.4.4'
gem 'sprockets', '> 2.0.0.beta.2'
gem 'compass', :git => "https://github.com/chriseppstein/compass.git"
gem 'jquery-rails'
gem 'money'
gem 'steam-prices', :git => "https://github.com/scottkf/steam-prices.git"
gem 'linguistics'
gem 'redcarpet'
gem 'kaminari'
gem 'devise'
gem 'dalli'
gem 'delayed_job'

group :production do
    gem 'hirefire'
end

Rails 3.1 — SQL logging to STDOUT during testing (with rspec, test::unit, or cucumber)

2011-05-28T00:00:00+02:00

Seemingly random SQL logging?

A simple fix, you probably have config.logger = Logger.new(STDOUT) in your test.rb or application.rb!

Rails 3.1 — Using ERB/HAML/etc within a Coffeescript JS file

2011-05-28T00:00:00+02:00

Need to use some information from your application in your new Coffeescript file? All you have to do is rename it with .erb.coffee or ..coffee instead of .coffee. This lets you do things like

$ ->
    $('#currency_select').change ->
        window.location = window.location.pathname.replace(/^\/(<%=Country.currencies.keys.join '|'%>)|(?!(<%=Country.currencies.keys.join '|'%>))\//i, '/'+$(this + "option:selected").text())

The best part about it is that Rails still knows when to generate the file, it doesn't do it each time! The only thing you can't directly access are the routes (e.g. root_path(), etc), but a gem called js:routes can take care of that for you.

Rails 3.1 — 'load_missing_contant': Expected ... to define ... (LoadError)

2011-05-28T00:00:00+02:00

Seeing this error?

This is a simple one to fix. I had two Classes of the same name but with different scopes, one was a model and the other a class in a separate module under lib. You have three options in this scenario:

Choose more appropriate names
Move the module out of lib into a gem
Rename the filename of the Class under the lib/ folder, while retaining the same class name.

Rails 3.1 — Testing with remote XML/HTML using Nokogiri, stubs, and rspec

2011-05-19T00:00:00+02:00

So you're pulling remote data, woo!

I suppose this works in Rails 3.x, but, in order to not be a burden on the server you're grabbing the data from, it would be nice to not have to grab the data each and EVERY time you have to run the unit tests, so you have to mock Nokogiri, but how. It's quite simple, assuming you're using Nokogiri like this: doc = Nokogiri::XML(open("http://tesoriere.com?xml=1")). In your unit test, all you have to do is mock the parse function and you're done!

doc = Nokogiri::XML(open(Rails.root + 'spec/support/tesoriere.xml'))
Nokogiri::XML::Document.stub!(:parse)
Nokogiri::XML::Document.should_receive(:parse).and_return(doc)

Rails 3.1 — use the jQuery Google Library instead of the default

2011-05-19T00:00:00+02:00

Rails 3.1 supports all sorts of fun new technologies out of the box, Coffeescript, Sprockets and more. However, the default jQuery 1.6.1 library that is included by default isn't minified, as well as the jQuery-rails library. This results in a whopping 250kb javascript overhead.

To reduce the overhead, and use the Google jQuery library instead, you have to make a couple changes. First, in app/assets/javascripts/application.js:

//= require jquery
//= require jquery_ujs
//= require_tree .

Change this to:

//= require jquery_ujs
//= require_tree .

And in your app/views/layouts/application.html.* change the header by adding an additional line:

<head>
    <title>Steamhours</title>
    <%= stylesheet_link_tag    "application" %>
    <%= javascript_include_tag "https://ajax.googleapis.com/ajax/libs/jquery/1.6.1/jquery.min.js" %>
    <%= javascript_include_tag "application" %>
    <%= csrf_meta_tags %>
</head>

Note, you can't just add //= require https://../jquery.min.js because sprockets builds the resulting js file each time Rails receives a request (assuming it has changed). It really wouldn't make much sense to download the api and redistribute it, it would defeat the point of even using it in the first place.

All the additions in Rails 3.1 are wonderful, go try them out!

Rails 3.1 — Fixing the 'ajax:loading' event

2011-05-19T00:00:00+02:00

Having difficulty getting the ajax:loading event to trigger? All you have to do is use ajax:beforeSend instead! An example:

Our example coffeescript:

$ ->
    toggleLoading = -> $("#loading").toggle()
    $("#loading").toggle()
    $("form[data-remote]")
        .bind('ajax:beforeSend', toggleLoading)
        .bind('ajax:complete', toggleLoading)
        .bind('ajax:success', (event, data, status, xhr) ->
            $("#list").html($.parseJSON(data))
        )
        .bind('ajax:error', (xhr, status, error) ->
        )

Our form:

<%= form_for @profile, :id => "hours_form", :url => {:action => :show }, :remote => true do |f| %>
    
    <%= field_set_tag do %>
        <%= f.label :profile_name %>
        <%= f.text_field :profile_name %>
    <% end %>
    
    <%= submit_tag 'Submit', :id => "submit" %>
    
<%end%>
<div id="loading">
Loading...
</div>

And finally, the controller:

class ProfilesController < ApplicationController

  def index
    @profile = Profile.new
  end
  
  def show
    @profile = Profile.find(p.id, :include => :games)
    
    respond_to do |format|
      format.html
      format.js { render :json => @profile }
    end
  end
end

Integrating and maintaining a calendar of events within the Symphony CMS

2011-05-03T00:00:00+02:00

Symphony doesn't easily offer a way to integrate a calendar of events into an existing setup, so I've come up with an XSLT Utility to make it a little easier.

After having to revisit a calendar of events and kind of hacking my way through it originally, I've come up with another way to incorporate an event calendar more easily into an existing site. A very simple example is hosted here. The source is available at github. If you have any issues, please discuss them at github or at the Symphony forums: http://symphony-cms.com/discuss/thread/66262/.

Requirements:

Nils' Date and Time Field Extension - Because it uses a nice interface and allows for easier management of events; make sure to take the experimental branch.
This event I created, found on github, placed in your workspace/events folder.
A section to handle all the events with a single Date & Time field, it must contain at least one other field, like name, description, etc. But you can have as many as you like. They will all be displayed, and can be disabled by using CSS display:none; on the <li>. For example, if we had a field called description it would be produced as <li class="description">, so .description { display: none; } would get rid of it.
Attach the event to whatever page you want to display the calendar on.
Attach a data-source pulling whatever events you wish, you can limit it by date to reduce the amount of entries, but do not change the grouping.
Finally, in the page you wish to drop the calendar in, the php xmlns must be added to the <stylesheet> tag, and the function to generate the calendar must be used. An example page would look like is below. The second parameter next to the function is the name of the data source. The third is an optional path to link to the event by id so you provide the root path (if none is provided it will just be listed as text). The fourth is an optional class to provide the table with, the default is calendar.

      <?xml version="1.0" encoding="UTF-8"?>
        <xsl:stylesheet version="1.0"
                xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
                xmlns:php="http://php.net/xsl">

        <xsl:import href="../utilities/master.xsl"/>

        <xsl:template match="/data">

        <h1><xsl:value-of select="$page-title"/></h1>
        <xsl:value-of select="php:function('generate_calendar', 
          calendar-main-events, concat($root,'/calendar/event'), 
          'calendar')" disable-output-escaping="yes"/><br />
        </xsl:template>
        </xsl:stylesheet>

Bugs/features:

It's not possible to have a date field in the same section you plan to use for the events, it conflicts with a date time field.
If you have two calendars on the same page, when you switch months/years, it will switch them on both calendars.

You can download it at my Github repo.

How to get CM7 on your HTC Incredible (Android 2.3/Gingerbread)

2011-02-09T00:00:00+01:00

Why this is important and why I am not responsible for your phone exploding :(

I originally switched from an iPhone 3GS to the HTC Incredible because ATT was molesting my wallet. Only, I discovered the default OS wasn't too responsive, and there was a ton of crap on it. Aside from that, I didn't really have any complaints, and so enter custom roms and Cyanogen Mod 7 (CM7) and the point of this article.

CM7 is a customizable, nearly universal Gingerbread/Android 2.3 ROM. There's a port for almost each Android device that can be rooted. A phone being rooted is something akin to jailbreaking an iPhone. It just let's you do more stuff. I'm not talking about you pirates :(

BEFORE YOU BEGIN!

Please note than this will start your phone fresh, so you might considering backing up all of your data if you don't already have a backup. To make a backup of your apps, please consider using AppBrain.

How do I do it?

To install the rom is fairly easy. You have to (1) root the phone, and (2) install the rom. I won't explain it step by step, because there are plenty of resources. To begin, please read about more of the terminology and disclaimers here.

Rooting the phone. The software to do it is called Unrevoked Forever and can be downloaded here. After download the software, the process is described in detail at their website, unrevoked.com. It's just involves running the programs, and clicking through the prompts.
Downloading CM7. You can either download Rom Manager after the phone is rooted and download CM7 through there, which is located under Cyanogen nightly, or download it here, copy it to the phone and use Rom Manager to install it. Supposedly #29 is stable, I'm on an earlier version and everything works fine.

If you get past step one, which you should, you can backup your current phone (which is one nice thing about rooting) as a nandroid backup, which is akin to an iTunes iPhone backup, and always revert back to it if nothing works well.

GTX 460 in Mac OSX 10.6.5 (or osx86)

2010-11-16T00:00:00+01:00

Completely rid yourself of freezes!

But... But... That's illegal!

Actually, no, it's not. I'm not selling, nor condoning the use or creation of a hackintosh. If you however, are a hobbyist and are curious, the GTX 460 family of nVidia graphic cards now work in Snow Leopard (osx86).

All one has to do is go here and download the latest drivers for Snow Leopard. This, however, requires you to update to 10.6.5, and use an appropriate boot loader with the following in com.apple.boot.plist:

<key>GraphicsEnabler</key>
    <string>Yes</string>

More information can be found on insanelymac.

Cucumber, capybara, lightboxes, and you

2010-10-21T00:00:00+02:00

Selenium wasn't playing very nice with lightboxes during one of my tests, it just said it wasn't visible! Thanks! That helps all you have to do is wait silly:

When I wait until I can see "new_category"

And the step definition:

When /^I wait until I can see "([^"]*)"$/ do |selector|
  page.has_css?("#{selector}", :visible => true)
end

Cucumber, capybara, and checking for the existence of selected options within a select box

2010-10-21T00:00:00+02:00

Checking for a selected option within Cucumber/Capaybara is quite simple:

Then "hello" should be selected for "article_category_id"

And the step definition:

Then /^"([^"]*)" should be selected for "([^"]*)"$/ do |value, field|
  assert page.has_xpath?("//option[@selected = 'selected' and contains(string(), '#{value}')]") 
end

You could easily extrapolate this for use within an unselected value, just by removing the selected bit:

Then /^"([^"]*)" should be seen within "([^"]*)"$/ do |value, field|
  assert page.has_xpath?("//option[contains(string(), '#{value}')]") 
end

When should I use Ruby on Rails, or something else? (Comparing Symphony and Rails)

2010-10-18T00:00:00+02:00

When beginning a new project, the first thing I come across is what software/language/methodology do I use to complete the project? The answer is quite simple, it depends.

Oh, you clever chap, what a cheeky answer

The truth of that matter is, it does depend solely on your application requirements and your knowledge of what's available. Nearly 90% of the time in my situation, it either comes down to either of two things:

Ruby on Rails or
Symphony CMS

The other 10% of the time I use my own solution, as in SAMOS, but that's for another day.

Both Rails and Symphony subscribe to a Model-View-Controller (MVC) architectural software design methodology. Rails treats the methodology quite literally, while Symphony does so a little more loosely. The nomenclature in each are as follows:

	Rails	Symphony
Model	Model	Section
View	View	Pages
Controller	Controller	Data-Sources/Pages/Components

I won't really describe rails much, because there's plenty of information about it. A great place to start is here: http://edgeguides.rubyonrails.org/getting_started.html.

Then I'll go somewhere else!

However, Symphony could use some explanation. Symphony's models are called sections (which contain many fields, just like fields in a database), and they are just like rails models. It's views are controlled by pages, and it's controllers are controlled (hah!) by data-sources, pages, and components. While that may be a little confusing, it works quite wonderfully.

Pages control what gets output as HTML, but you can also include logic in them, just like Rails does with ERB/HAML. However, a best practice is to stub out a lot of the logic into components (templates) to make it more DRY.

Data-sources are the link between the pages/components and the sections/fields. In rails, this would be the ActiveRecord interface. They grab data in a certain way based on lots of different things in lots of different ways. I couldn't be more vague, I know.

Components are similar to helpers/partials in rails, where you can lump similar things together so you can reuse them.

Go on then, answer the original question

How I decide which to use basically involves me looking at this table and deciding how I feel like torturing myself for that certain project. Most of the time the biggest issues are whether or not it contains a lot of logic that needs to be tested, and which template language I want to use.

	Rails	Symphony
Testable?	Yes, I use BDD	Not so much
Template Language	ERB/HAML	XSLT (this is quite amazing)
GUI	Nope	Yes, you can see an example here
Output	HTML/XML/JSON/etc	HTML (as xml), JSON
Database	Everything you can throw at it	MySQL
Hosted at Godaddy?	FML	DIAF

There are a couple other things to compare, but these are often the most important.

Great! Losing patience, how do you decide?

If it's more of a web application and not as much as website, I often use rails, because they usually involve logic that should be tested.

The template language shouldn't be, but is a huge issue to a shallow individual such as myself. You can do exactly the same stuff in both ERB/HAML and XSLT. They both have their advantages and disadvantages. ERB/HAML's main pro going for is that it's easy to use and quite succinct. XSLT's plus is that it's XSLT and lets you traverse data (XML) in such an amazing way, but it comes at price, it's quite verbose and ancient (it's from 1999). For me? It depends, again, and how bored I am with one over the other.

Whether or not I need Symphony's GUI is also another important issue. If I don't feel like building an interface and using the default Symphony UI for tech-savvy clients I will often choose this (though you can also make your own interface as well).

Once again, if it's more of web app, I'll often use rails because I might need an API (where I might need to output XML/JSON) right away, or at some point in the future. In Symphony, you can output XML right out of the box, and can create a REST API, but the CUD part of CRUD is a little harder to manage in symphony this way.

The database usually isn't as huge issue, and mainly depends on where the app/site is being hosted. Symphony currently only supports MySQL, but nearly every host supports MySQL, so whatever!

And there you have it, it simply depends, on... lots of stuff. </shill>

WPInks, our new partner in elegant photography

2010-09-28T00:00:00+02:00

Well I certainly didn't take that picture.

Someone else must have! In an effort to maintain my focus on mostly development and design, I've partnered with a local business, WP Inks (West Palm). WP Inks is a design firm specializing in Photography, Calligraphy and Interior Design. Examples can be found at their website. Their main focus is family photography, still photography, engagement photos and the like.

I've known the President of the company for a long time now, and I'd stake my reputation on the quality and commitment of their company.

Rails 3.x — Self-referential has_many :through Parent<->Child relationship modelling

2010-09-01T00:00:00+02:00

Seriously?

Srsly, that's more of an SEO title than anything, because damnit, I know how the hell it works at an SQL level and it was frustrating to get working in Rails because of syntax, namely :class_name => :object, when it should be :class_name => "object". I hope someone finds this useful, because when it comes to rails 3 there's not too much stuff that's recent.

Get to the point, show me the goods

rails new confusingTitleProject
cd confusingTitleProject
rails g model Person name:string
rails g model PersonRelationship parent_id:integer child_id:integer

#App/Models/PersonRelationship.rb
class PersonRelationship < ActiveRecord::Base
  belongs_to :parent, :class_name => "Person"
  belongs_to :child, :class_name => "Person"
end

#App/Models/Person.rb
class Person < ActiveRecord::Base
  
  validates_presence_of :first_name, :last_name

  has_many     :parent_child_relationships,
               :class_name            => "PersonRelationship",
               :foreign_key           => :child_id,
               :dependent             => :destroy
  has_many     :parents,
               :through               => :parent_child_relationships,
               :source                => :parent

  has_many     :child_parent_relationships,
               :class_name            => "PersonRelationship",
               :foreign_key           => :parent_id,
               :dependent             => :destroy
  has_many     :children,
               :through               => :child_parent_relationships,
               :source                => :child
end

Then, as an example, you can do the following to add a parent:

a = Person.new; b = Person.new
b.parents = [a]

Migrating from mephisto to jekyll

2010-08-25T00:00:00+02:00

My life story:

Pfft, I'll save my virtual breath.

I've finally got around to moving my website (this one you're looking at!) to Jekyll. Why? Well, lots of reasons! But, they have already been mentioned numerous times.

However, if you don't know what Jekyll is, I implore you to visit the wiki and learn about it, it's great fun if you're familiar with the terminal in osx, unix, *nix, or just about anything without a super fancy user interface. It lets you:

write your posts in Markdown or Textile (or anything you want to extend it to)
use plain old css and html
host your site anywhere with space, (even mobileme)
use git or svn as a naturally way of supporting article revisions, which mephisto did beautifully (I prefer git)

So long demon, it's been... hell.

I came from mephisto, which is dead now, but luckily there is a way to migrate everything (except the tags :() to Jekyll. Comments also! Huzzah! The process is listed there, so I won't rehash what was said.

Boy that sure is a lot of purple.

The design hasn't changed much, only tweaked a little. Let's face it, I'm no amazing designer, I usually hire other people for that. However, I've never found a great way to represent myself, so this will do for now. The important technical bit is that is uses HTML5 and CSS3 (via compass, which is amazing).

In order to maintain compatibility with older browsers, there's something called modernizr, which translates HTML5 elements such as <section> and <article> into something older browsers can deal with. It really is as simple as adding this into your header:

<script src="/javascripts/modernizr-1.5.min.js"></script>

and this into your body tag

<body class="no-js">

You can view my default layout for this site on github, right here. I make liberal use of the newer HTML5 elements (and <video> should I ever need it)!

Articles with name-tags are fun!

I had to re-tag all the articles after migrating, but I also needed a way to support /tag/something here urls. I described this process in a separate post, called Automatically generated tags and tag clouds with Jekyll (multiple word tags allowed). I should probably fire my ghost-writer.

Musings from the peanut gallery.

Comments are a tricky thing since everything is statically generated, hence that whole "you can do this on MobileMe" thing. But as everyone knows by now, disqus is a great way to implement them.

The whole process to implement them is actually a relatively easy one, you just have to use some universal code after making an account from the disqus website to add it in.

Sending tesoriere.com to it's home.

Deployment is relatively easy with a rake task using rsync to send over all that data to wherever. It's really as simple as:

sh 'rsync -rtzh --delete _site/ #{DEPLOY_USER}@#{DEPLOY_HOST}:~/#{domain}/'

Please don't steal me :(

You can view the full source of this entire site on github! It really only takes a couple days to do, and it's worth it if you want to try something new, simple, and I'm sure as hell my hosting company appreciates less load.

Highlighting liquid code in a liquid template with Jekyll (Escape a liquid templating tag)

2010-08-25T00:00:00+02:00

This is a somewhat confusing thing to achieve, highlighting liquid templating code in a code block, and made even more confusing by the title. Take this example from this article about I wrote about mephisto:

#previous link
{% if article.previous? %}
  {{article | previous_article | link_to_article }}
{% else %}
  <a href="/">home</a>
{% endif %}

To actually display the liquid template code you have to get crazy with the curly brackets; the code in the markdown (in my case) file actually looks like:

#previous link
{{ “{% if article.previous? “}}%}
    {{ “{{article | previous_article | link_to_article  “}}}}
{{ “{% else “}}%}
  <a href="/">home</a>
{{ “{% endif “}}%}

Typing that above leads to a sweet infinity effect, but do yourself a favor and don't copy/paste that, just take it from this post on github, because to actually type that I had to use “. Of course a proper way to do would be to extend markdown/textile in pygments to support the liquid templating language, but that's for another day.

Automatically generated tags and tag clouds with Jekyll (multiple word tags allowed)

2010-08-25T00:00:00+02:00

Jekyll is a great thing, and supports tags right out of the box, but in order to get urls like in mephisto, /tags/something here, you have to either add an extension (which do exist, though I haven't tried), or generate them with the Rakefile, or however really, but I chose the Rakefile. Luckily raimonds simanovskis had already done so, but his solution didn't support tags with spaces, which I had used in mephisto, and to not break compatibility when switching, I had to edit the relevant part of his Rakefile to support tags with spaces.

So an an example post looks like this:

--- 
layout: post
title: Cloning an ISO to a USB flash drive (or anywhere) ...
syntax-highlighting: yes
tags: 
- dd
- snow leopard
---
Oh, and Leopard too...

The original method (can be viewed in it's entire, original, unedited form here). To allow support for tags with space (which some people don't like but I do!), you have to change it so it looks like:

desc 'Generate tags pages'
task :tags  => :tag_cloud do
  puts "Generating tags..."
  require 'rubygems'
  require 'jekyll'
  include Jekyll::Filters

  options = Jekyll.configuration({})
  site = Jekyll::Site.new(options)
  site.read_posts('')

  # Remove tags directory before regenerating
  FileUtils.rm_rf("tags")

  site.tags.sort.each do |tag, posts|
    html = <<-HTML
---
layout: default
title: "tagged: #{tag}"
syntax-highlighting: yes
---
  <h1 class="title">#{tag}</h1>
   
   {% for post in site.posts %}
       {% for tag in post.tags %}
           {% if tag == '#[tag]'%}
               {% include post.html %}
           {% endif %}
       {% endfor %}
   {% endfor %}
HTML

    FileUtils.mkdir_p("tags/#{tag}")
    File.open("tags/#{tag}/index.html", 'w+') do |file|
      file.puts html
    end
  end
  puts 'Done.'
end

Note the line if tag == "#[tag]" should be if tag == "#{tag}" but liquid 2.0 doesn't support escaping, and my ghetto method doesn't work. So just pull the entire function from github.

View my entire Rakefile @github.

GTX470/480 (Fermi) in OSX

2010-06-21T00:00:00+02:00

This may or not be relevant to you, but it's interesting. How to use a GTX470/480/465/480M in OSX.

http://www.insanelymac.com/forum/index.php?showtopic=214219

Premier Medical Care

2009-10-13T00:00:00+02:00

Premier Medical Care's new website, he also happens to be my physician. http://premiermedicalcarepb.com.

Cloning an ISO to a USB flash drive (or anywhere) using DD in Snow Leopard 10.6 with a progress-ish meter

2009-08-31T00:00:00+02:00

Oh, and Leopard too.

What a title. I recently needed to use G-Parted, but didn't have any blank disc's laying around, and the method they describe on the gparted wiki to do this with a USB disk doesn't work on OSX. So, I used DD.

Open up terminal (from Applications/Utilities or use Spotlight) and type what follows substituting as necessary. My USB device was located at /dev/disk3. You can find the device name in Disk Utility by right-clicking on the device, and choosing 'information'.

THIS WILL OVERWRITE WHATEVER'S ON THE DISK.

sudo dd if=/path/to/iso.iso.dmg.img of=/dev/disk3 bs=1024

And voila, you have a read-only replica of the iso on your usb device. Please note the emphasis on read-only in the case of an ISO.

To see a progress bar (of sorts), open up a new terminal window and type the following

ps auxww | grep " dd " | grep -v grep | awk '{print $2}' | while read pid; do kill -SIGINFO $pid; done

I originally got the progress bit from here, but changed it to work with OSX.

Using a 3.0 iPhone on T-Mobile? Get MMS Working!

2009-06-16T00:00:00+02:00

I am, and have been. T-Mobile's customer service is beyond amazing.

Go to settings -> general -> network -> cellular data

APN: wap.voicestream.com
MMSC: 216.155.174.84/servlets/mms
MMS Proxy: 216.155.165.50:8080
Username: leave empty
Password: leave empty

Click the picture --^ to see how it looks on the phone.

Keep Flippin' website v3

2009-06-15T00:00:00+02:00

New Hotness (updated)

I've recently changed the Schedule page, and a hidden, non-publicly page, to use a Rails Application as a 'data-source' to provide the Schedule of classes, as well as other stuff on that non-public page. This was done in an effort to help ease the transition between the software that Keep Flippin' currently uses to manage itself, and the new software I will be writing. More will follow in the coming months!

Old and Busted (original)

Below this is all tech-talk. To summarize, I used new stuff to make the site better, make it quicker, and more responsive. This is what I made.

The new keep flippin website is built using compass for the CSS (this is amazing), symphony, some extensions I created, SASS, XSLT and GitHub to keep track of everything.

I know I've been living under a rock

2009-06-14T00:00:00+02:00

But have you seen jekyll?. It makes blogging seem like it might be fun. I bet the people at my hosting company would appreciate it too.

Dealing with database concurrency (locking) in the Symphony CMS

2009-06-13T00:00:00+02:00

Symphony doesn't natively supported pessimistic database locking, so I created an extension in symphony to handle it, both in the backend and the frontend. It's somewhat complicated to implement, but so far it works pretty great. I created it because one of the next big projects I'm working on will be used by many people at once, and I was considering using symphony to create it, as opposed to Rails.

You can get it here.

Report any issues at GitHub.

Uploading multiple files with Symphony CMS

2009-06-01T00:00:00+02:00

Many CMS' don't offer a way to manage picture galleries easily, and neither does symphony out of the box. But, after playing with symphony some, I decided that I liked it and the next project I would do would be with symphony.

So I created a plugin (in symphony: extension) that allows you to upload multiple files, and create entries (in this example, gallery items, or rows in the database, etc) to allow you to accomplish this easily.

You can download it at my Github repo.

Symphony CMS

2009-05-28T00:00:00+02:00

I've been working on moving this to this. It's got a bit of a learning curve, but I would say it's like Rails for designers with a GUI. That's not a bad thing. The bad thing is that it uses XSLT 1.0. Which is the least powerful language ever written (spec was written 8 years ago). It's not a procedural language, and if you don't treat it as one, you'll be ok.

Some posts about tutorials will follow.

mephisto next article link, and previous article link

2009-04-17T00:00:00+02:00

8.2 either removed the two functions I used to manage article navigation at the bottom (move your eyes dowwwwn just above the comments), or I added them and thought they came for free. After looking at the date in the master repo the latter appears to be true!

#article_drop.rb

## this goes in app/drops/article_drop.rb
def next?
  !@source.next.blank?
end

def previous?
  !@source.previous.blank?
end

Then to use it, go to your liquid template and add the following somewhere important:

#previous link
{% if article.previous? %}
  {{article | previous_article | link_to_article }}
{% else %}
  <a href="/">home</a>
{% endif %}

#next link
{% if article.next? %}
{{article | next_article | link_to_article }}
{% else %}
  <a href="/">home</a>
{% endif %}

If you want the links to have different text, say next/previous, you need to change the previous like so

#previous link
{% if article.previous? %}
  {{article | previous_article | link_to_article: 'PREVIOUS' }}
{% else %}
  <a href="/">home</a>
{% endif %}

#next link
{% if article.next? %}
{{article | next_article | link_to_article: 'NEXT' }}
{% else %}
  <a href="/">home</a>
{% endif %}

It would appear that the prev/next_article filters check to see if one exists, but I prefer to link back to home for usability purposes, and this lets me do that.

embed a mobileme gallery in a design, on any website

2009-04-11T00:00:00+02:00

On the right there's a new link for 'pictures', I'm embedding a mobileme gallery within the frame of the extremely wide tesoriere.com layout. It's not thaaaat terrible, but honestly, don't most people use facebook now?

The relevant code is simple, and can be found through iWeb, but to save you the trouble:

<iframe src="http://gallery.me.com/<username>" frameborder="0" scrolling="no">
</iframe>

You have to give it a big enough frame [w:855px,h:630px], or allow it to scroll.

PS. It's also hosted on MobileMe with a CNAME pointing to web.me.com from me.tesoriere.com.

up next

2009-03-07T00:00:00+01:00

Last year, I completed a simple commerce website where I used my own solution written in rails. It worked, but it was a very simple application, not much flare, not much ajax (which is omg exciting). So for this, it would've wound up being a little too simplistic, and to avoid having to reinvent the square wheel, I tried using substruct. Substruct is great for what it is, but it hasn't been updated in some time now, so I tried redoing it with spree, and that's where I am now. The transition has been pretty smooth since I had finished most of the views already. As far as the functionality, it seems like a winner! Huzzah.

mephisto - section with multiple pages on homepage (static intro page)

2008-06-21T00:00:00+02:00

The way mephisto works, you can't have a static intro page (one article per page, a 'paged' section), and a blog in a subdirectory (ie. /journal). These options are found under the 'sections' tab.

The reason it breaks is because of the way mephisto uses permanent links. It won't prefix permanent links with the URL associated with their section.

There are a number of ways to fix this:

sneaky template manuvering (the solution I took described below)
changing the way permanent links are constructed in the mephisto code-base (seemingly the proper way) and submitting a patch to include it in future revisions
just moving the whole mephisto installation to a subdirectory. This has it's problems because you wouldn't have access to the articles at all to slap on your introduction page. I use them to represent the 'portfolio' on this website.

The first method is the quickest; using the templates to achieve this is fairly simple and straightforward. In your 'layout.liquid', which is the default template, you can add the following:

Include file ''journal'' contains invalid characters or sequences

There also exists a 'home.liquid' and 'journal.liquid' template with however you want them to be designed. And that's it!

datamapper - many-to-many parent child relationship

2008-06-17T00:00:00+02:00

I started playing around with datamapper to see if I could model an many to many parent/child relationship (self-referential habtm?) for an upcoming project. The plan was to use sproutcore for the gui (once I figured it out!), and merb as a webservice that it interacts with.

Justin gave me a kickstart with datamapper, and I eventually came up with:

#Person.rb
class Person
  include DataMapper::Resource


  property :id, Integer, :serial => true

  has n, :relationships
  has n, :parents, :through => :relationships, :class_name => 'Person'
  has n, :children, :through => :relationships, :class_name => 'Person'


  # parent = Person.create()
  # child = Person.create()
  # relationship = Relationship.new(:parent => parent)
  # child.relationships << relationship
  # child.save

end

#Relationship.rb
class Relationship
  include DataMapper::Resource
  
  property :id, Integer, :serial => true
  belongs_to :person
  belongs_to :parent, :class_name => 'Person'
  belongs_to :child, :class_name => 'Person'
end

I'll eventually write some spec's for it.

new design

2008-06-04T00:00:00+02:00

There's a new design up today, and it came out pretty well. I still believe it needs more cowbell for it to truly shine. It runs on mephisto and uses slimbox for the lightbox popups.

dr lorne s. stitsky - revised

2008-05-29T00:00:00+02:00

I just finished the 2nd version of Dr. Stitsky's website to match his current marketing materials.

gardenhouse

2008-05-28T00:00:00+02:00

I just finished developing a website for Cheryl Maeder, for her business, Gardenhouse. They designed the majority of it, and I used textpattern to control the backend, using Slimbox to display pictures, and Slideshow 2.0 for the 'art' page. Everything worked out pretty well.

dr lorne s. stitsky

2008-05-28T00:00:00+02:00

I recently finished a web site for Dr. Lorne L. Stitsky, for his personal practice. His slogan is, "putting the care back into healthcare... one house call at a time". It's not just a marketing ploy either, he actually makes housecalls. I designed and developed the site, it uses Mephisto as a backend and Lightbox for the gallery.

Update: the site is pending a redesign to match Dr Stitsky's new theme. The website was created before he had a theme set in mind.