WinstonYW

Responsible Engineering

Sat, 17 Oct 2015 00:00:00 +0800

I wrote a post originally to talk about "Responsible Engineering" and showed it to folks at Jolly Good Code for feedback. Then Choon Keat showed me this particular set of slides..

After going through the slides, I decided that my post wasn't worth reading and you should read that set of slides instead.

Dan McKinley covered everything I wanted to say about being responsible and more in his talk titled "Choose Boring Technology". The deck is golden and every engineer should read and memorize it.

Also, there are some interesting conversations on Twitter too.

~~The only thing I would add is that IF anyone still decides to try out that new shiny thing on production, please, please insist for it to be tested. Any kind of tests!~~

Scrap that.

Even writing tests is not justification enough to introduce a new shiny thing to the mix. Does the app really need that to work, or was it added to satisfy your own personal desire? Think of the big picture and the impact that the new shiny thing would bring to the team. Would that be a positive or negative impact? Brings to mind this little story.

Maybe a responsible engineer is more likely to stick with "boring" technology?

Half A Decade Of RedDotRubyConf

Thu, 30 Jul 2015 00:00:00 +0800

This year, 2015, is the fifth run of RDRC and my third time organising it. Even though it's my third time doing this, it's not in anyway easier because expectations have definitely gone up over the years, and I have to do something about the feedback collected last year. :)

The preparation started 6 months before the conference, and it always begin with the venue and dates. I picked Biopolis again because it's really affordable, and that means ticket prices can stay similar.

Then I started to invite several key speakers, one of whom is Matz. I really wanted to have Matz at RedDotRubyConf 2015, as he was here in 2011 and so I thought it would be special to have him back again 4 years later. But alas! Matz wasn't able make it for the dates (May 28, 29) which I planned for.. What? OMG!

Like I said, I really, really wanted to have Matz at RedDotRubyConf 2015! So I went back to the venue and asked for a few other open dates and luckily one of it suited Matz's schedule. Phew! And the rest was history as Aaron, Sam, Linda, Laurent and Jesse all agreed to come too!

From January onwards, the months went by quickly as we took care of the logistical challenges of the conference. Too much to say here, but did I mention I am trying to write an open source Conference Planning Guide?

Finally June 4, 5 came and went by quickly and I really couldn't have done it without the help of Guo Xiang, Jingwen, Elisha, Anna, Laurence and Cheeaun. They covered the grounds that I wasn't able to attend to and were instrumental in making the conference a success. Thank you!

Thank You Sponsors!

Our sponsors were awesome and I can't thank them enough! They just wanted to do their bit to help build the Ruby community and I hope RDRC helped them snag a few more deals or new hires!

Why are sponsorships important?

Firstly, sponsorships are boosts of confidence for what we were doing, such that organisations find it advantageous for them to be a sponsor.

Secondly, because ticket sales alone are never enough to cover the full cost of the conference, and so sponsorships are really important to make sure we don't go into red.

Thank you:

Thank You Speakers

Speakers - They are the soul of the conference. Thank you all so much!

25 Speakers - 6 Key Notes, 9 CFPs and 9 Lightning Talks.

We had topics about Ruby (and mRuby), Rails, RubyMotion, Robotics, Security, Refactoring, Optimization, Open Source, etc, and this year, we also did a Panel discussion on Ruby which was very well-received!

I learned a lot from all the talks and I hope you do too. You can enjoy the talks again as they are now available on Confreaks.

Thank you Matz, Aaron, Sam, Linda, Laurent, Jesse, Aman, André, Sau Sheong, Shipeng, Prathamesh, Laura, Juanito, Hiroaki, Christopher, Joy, Vaidehi, Paolo, Hiroshi, Tania, Guo Xiang, Radamanthus, Elisha, Grzegorz, Tomoya and Yuki!

Thank YOU!!!

As I have always said, RDRC is for you, and you, and you!

Almost 350 Rubyists from all over the World gathered together for 2-days to celebrate Ruby on the tiny island called Singapore.

Thank you for making this possible. Thank you all for your support!

Let's continue to make the Ruby community even better!

Infographics

Piktochart made generating this infographic enjoyable and truly A BREEZE!!!

Resources

Just refer to Cheeaun's awesome curated list!

Closing

Honestly, imo, I felt that this was the best RDRC amongst that last 3 that I organised. First year was about making sure I don't mess up and that the brand continued to grow. Second year was about repeating the same feat (while juggling with starting my own consultancy). This year, it was about taking care of the finer details, so that everyone felt welcomed.

Things that I felt we did well:

We have private buses that brought speakers (who were staying at the conference hotel) to Biopolis
We have better breakfasts and tea breaks
We have barista-style coffee/tea because developers care for their daily fix
We have vegetarian lunch boxes for attendees who indicated on registration that they were vegetarian
We have cleaners to make sure the venue is kept spick and span
We did a RubyMotion workshop before the conference, and RailsGirls and Git Workshop after the conference

If I were to give RDRC 2015 a score, I would say we scored 90 marks. Haha.

Organising a conference is challenging and tiring, but at the same time, fun and rewarding.

That's all folks! :)

Drop me an email - suggestions, compliments or criticisms. Thank you!

Setting Up Ruby on Rails On Digital Ocean

Fri, 24 Oct 2014 00:00:00 +0800

Edit This has been updated for Ubuntu 14.10, Ruby 2.2.2 and Postgres 9.4.

Documentation of how I setup SGBusFeedback.com, a basic Ruby on Rails application backed by a PostgreSQL database on Digital Ocean.

Noting it down for my future self.

Create Droplet on Digital Ocean

Login to Digital Ocean and start Create-ing a Droplet
Size : $5/mth
Region : Singapore
Image : Linux Distribtions > Ubuntu 14.04 x64
Add SSH key for easy access to the Droplet
Create Droplet!

Set Up Droplet

Assuming that your SSH key has been set up properly, SSH into your Droplet.

machine$> ssh root@<ipaddress>

Add a Deploy User

Create a deploy user on the Droplet.

droplet$> sudo adduser deploy
droplet$> sudo adduser deploy sudo #add to sudo group

Create a SSH key on your machine, and upload the key.

machine$> ssh-keygen
Generating public/private rsa key pair.
Enter file in which to save the key (/Users/winston/.ssh/id_rsa): id_deploy_digital_ocean
...

Copy the SSH key to the Droplet.

On Mac? brew install ssh-copy-id. On Linux? You already have ssh-copy-id.

machine$> ssh-copy-id -i ~/.ssh/id_deploy_digital_ocean deploy@<ipaddress>

Now you can login to your Droplet with:

machine$> ssh-add ~/.ssh/id_deploy_digital_ocean
machine$> ssh deploy@<ipaddress>

Harden Droplet

Modify SSH settings /etc/ssh/sshd_config in the Droplet.

Add PermitRootLogin without-password (What?)
Add UseDNS no (What?)
Change Port to something unique and take note. E.g. 8888

Reload SSH with reload ssh.

Verify your SSH settings with your new user. In a new Terminal:

machine$> ssh -p 8888 deploy@<ipaddress>

Bonus. Add this to ~/.ssh/config on your machine so that login will be a breeze.

# in ~/.ssh/config
Host    <appname>
    HostName        <ipaddress>
    Port            8888
    User            deploy
    IdentityFile    ~/.ssh/id_deploy_digital_ocean

machine$> ssh <appname>

Update and Upgrade Ubuntu

A Ruby on Rails server requires quite a few things. Installed these as the deploy user.

droplet$> sudo apt-get update
droplet$> sudo apt-get upgrade
droplet$> sudo apt-get install git-core
droplet$> sudo apt-get install nodejs # for assets compilation.

RVM

Install RVM.

droplet$> curl -L https://get.rvm.io | bash -s stable
# If an error about gpg occurs, follow the instructions..
droplet$> gpg --keyserver hkp://keys.gnupg.net --recv-keys D39DC0E3
# https://github.com/wayneeseguin/rvm/issues/3110

Relogin to shell, install RVM requirements, and Ruby.

droplet$> source /etc/profile.d/rvm.sh
droplet$> echo "source /etc/profile.d/rvm.sh" >> ~/.bashrc
droplet$> rvm requirements
droplet$> rvm install 2.2.2
droplet$> rvm use 2.2.2 --default
droplet$> ruby -v

RubyGems Setup

Prevent RubyGems from installing docs.

droplet$> echo "gem: --no-ri --no-rdoc" > ~/.gemrc

Install Nginx and Passenger

Follow the instructions on the Official Documentation.

Condensed Version:

Install Phusion's PGP key and add HTTPS support for ATP.

droplet$> sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys 561F9B9CAC40B2F7
droplet$> sudo apt-get install apt-transport-https ca-certificates

Add Passenger's repository. Just one.

droplet$> sudo sh -c 'echo "deb https://oss-binaries.phusionpassenger.com/apt/passenger trusty main" >> /etc/apt/sources.list.d/passenger.list'
droplet$> sudo chown root: /etc/apt/sources.list.d/passenger.list
droplet$> sudo chmod 600 /etc/apt/sources.list.d/passenger.list
droplet$> sudo apt-get update

Install Nginx and Passenger packages.

# The Official Documentation uses `nginx-extras` but `nginx-full` is good enough.
droplet$> sudo apt-get install nginx-full passenger

Edit /etc/nginx/nginx.conf and uncomment passenger_root and passenger_ruby.

Also find out where your ruby is, and use that for passenger_ruby.

droplet$> passenger-config --ruby-command
  Command: /usr/local/rvm/gems/ruby-2.2.2/wrappers/ruby
  Version: ruby 2.2.2p95 (2015-04-13 revision 50295) [x86_64-linux]
  To use in Apache: PassengerRuby /usr/local/rvm/gems/ruby-2.2.2/wrappers/ruby
  To use in Nginx : passenger_ruby /usr/local/rvm/gems/ruby-2.2.2/wrappers/ruby
  To use with Standalone: /usr/local/rvm/gems/ruby-2.2.2/wrappers/ruby /usr/bin/passenger start

droplet$> sudo vi /etc/nginx/nginx.conf
... Edits ..
droplet$> sudo service nginx restart

Install PostgreSQL

Install it.

droplet$> sudo apt-get install postgresql-9.4 libpq-dev

Set up Postgres user.

The password that you used here will be used in your app's database.yml.

droplet$> sudo su - postgres
droplet$> psql -c"alter user postgres with password 'strong123password';"

Create the Production database.

droplet$> psql
psql> create database <production_dbname>;
psql> exit

Connect GitHub Repo to Droplet

Create a SSH key on the Droplet, and then add it to the GitHub repo.

Follow the instructions here: https://help.github.com/articles/generating-ssh-keys/

Make sure you run this droplet$> ssh -T git@github.com.

This will ensure that Mina can deploy with GitHub.

Setup Mina

Follow the instructions on Mina's README.

Condensed Version:

Init Mina on your application.

machine$> mina init
Created config/deploy.rb.

Use rvm or rbenv.

require 'mina/rvm'
# require 'mina/rbenv'

Update config/deploy.rb with Droplet's Details.

set :domain,       <ipaddress>
set :deploy_to,   '/home/deploy/<app_domain'
set :repository,  'git@github.com:<user>/<repo>.git'
set :branch,      'master'
set :user,        'deploy'
set :port,        '8888'
....
task :environment do
  queue 'source ~/.bash_profile'

  # For those using RVM, use this to load an RVM version@gemset.
  invoke :'rvm:use[ruby-2.1.4]'
end

SetUp Mina on Droplet.

machine$> mina setup
-----> Using RVM environment 'ruby-2.1.4'
       Using /home/demo/.rvm/gems/ruby-2.1.4
-----> Setting up /home/deploy/<app_domain>

       total 16
       drwxrwxr-x 4 deploy deploy 4096 Nov  5 02:00 .
       drwxr-xr-x 8 deploy deploy 4096 Nov  5 02:00 ..
       drwxrwxr-x 2 deploy deploy 4096 Nov  5 02:00 releases
       drwxrwxr-x 4 deploy deploy 4096 Nov  5 02:00 shared

-----> Done.
-----> Be sure to edit 'shared/config/database.yml'.
-----> Be sure to edit 'shared/config/secrets.yml'.
       Elapsed time: 1.00 seconds

Important! Update shared/config/database.yml and shared/config/secrets.yml on the Droplet.

#database.yml
production:
    adapter: postgresql
    encoding: unicode
    database: <production_dbname>
    username: postgres
    password: strong123password
    host: localhost

#secrets.yml
production:
    secret_key_base: <Generated from `rake secret`>

You might also need these settings in .bash_profile.

# vi .bash_profile
export LANGUAGE=en_US.UTF-8
export LANG=en_US.UTF-8
export LC_ALL=en_US.UTF-8

Now, you can deploy your application to the Droplet.

machine$> mina deploy

Update Nginx Configuration

Update the file /etc/nginx/sites-enabled/default with these settings:

server {
    listen 80 default_server;
    listen [::]:80 default_server ipv6only=on;

    server_name         <app_domain>;
    passenger_enabled   on;
    rails_env           production;
    root                /home/deploy/<app_domain>/current/public;
}

Restart the Nginx server, you should be able to see your app at the ipaddress.

droplet$> sudo service nginx restart

Next Steps

At this point, you already have your app up and running at ipaddress.

You will probably need to do the following to make it work nicely:

Use figaro or add ENV variables to .bash_profile.
Point DNS to the ipaddress.
Explore other configurations for your Nginx (static 40 or 50)
Put in place backup strategies for your PostgreSQL (maybe use RDS instead?)

That's it!

Troubleshooting

Sometimes, the $5/mth Droplet might complain of inadequate swap space. Read this.

Also, Nginx error logs are pretty handy. Available at /var/log/nginx/error.log.

Resources

Reading these really helped!

RedDotRubyConf 2014 In Review

Wed, 02 Jul 2014 00:00:00 +0800

RedDotRubyConf is 4 years old now, and this is the 2nd year that I am running it.

Although it's not my first, it's actually even tougher than the previous year, because I really wanted to make sure that it would be bigger and better than before.

Judging from the feedback so far, I think I did ok for the second time. :)

At this moment, I am full of gratitude to everyone who helped made RDRC a success, especially Jaryl, Yink, Kevin, Alarice - my Ruby Heroes, and many others!

Thank You, Sponsors!

We have a lot more sponsors this year, and all of them are awesome!

They didn't ask for much but just wanted to do their bit and help to build the community. Hope the conference helped to boost their brand and even hire a few good folks too.

Thank You, Speakers!

Speakers, we cannot do without any of you. All of you were awesome!

28 speakers in all - 6 Key Notes, 14 CFPs and 6 Lightning Talks.

I am sure all of us learned a ton from all the speakers in those 2 days.

As I was running around most of the time, I didn't manage to listen to all the great talks. Luckily, we have Confreaks covering the show this year, so I'll catch up on the talks soon.

Thank YOU!

What is a conference without an audience? RedDotRubyConf is for all of you!

Almost 350 Rubyists from all over the World gathered together for 2-days to celebrate Ruby on the tiny island called Singapore.

This year, we also have the most number of female attendees and students, and it's my hope to continue to grow the numbers for these two segments in the next year.

Thank you all for your support!

Infographics

Let's look at some infographics now, powered by Piktochart, one of our sponsors.

Piktochart is awesome! It's my first time using the service and I am blown away. Try it!

Improvements

Of course, nothing is perfect, and the day after it ended, I have already been thinking of improvements that I could do for 2015.

Ticketing and Registration

I really wanted to try out Peatix's Colorsync, but alas, we have some restrictions that's better handled by the traditional spreadsheet method. Bummer..

Anyway, next year, I hope to make the entire ticketing and registration experience better, and also to find ways to make discovery and interactions amongst attendees easier.

Swag

I got some pretty bad lanyards, I know. Sorry! I'll fix it next year.

Also, someone suggested actually giving out a book. Plausible. I'll consider that.

Breakfast and Water

Yes! There were breakfasts this year! I hope it helped to tide you till lunch time.

But, I'll definitely up the quantity and variety next year.

And it also seems like water is actually more popular than the sodas. Someone mentioned that maybe we are all more health conscious nowadays?

Ok. Let's do more water and less sodas.

Cleanup

I received good feedback about the cocktail tables and how it encouraged conversations during lunch.

However, at the same time, I am disappointed that a lot of the plates and soda cans were left on the cocktail tables after lunch. :(

Also, with 350+ people, A LOT of trash is generated!

Next year, I'll need to think about getting cleaning help.

Local Speakers

Same problem as last year. We need more local speakers! Where are all of you?

How can I make more local speakers submit CFPs? Hmm..

Social Media

This year, I got the team contribute to @reddotrubyconf during the conference.

I think we can do more next year, with a "social media plan" to keep it entertaining for all.

Resources

I'll be updating these, but otherwise, Cheeaun's list is awesome too!

Slides

My Slides, Day 1 and Day 2
Speakers' slides are linked on www.reddotrubyconf.com

Photos

Videos

Videos available on Confreaks.

Jobs Board

The full glory of our Jobs Board.

That's it. See you all in 2015!

Drop me an email - suggestions, compliments or criticisms. Thank you!

Time.now vs Time.zone.now

Mon, 03 Mar 2014 00:00:00 +0800

Ruby Time can get confusing when used in your Rails app. What Time is it now?

Suppose I have a server in UTC time and all of my users are from Singapore (GMT +08:00), then your Rails app's time_zone should be configured to SGT.

# config/application.rb

# Set Time.zone default to the specified zone and make Active Record auto-convert to this zone.
# Run "rake -D time" for a list of tasks for finding time zone names. Default is UTC.
config.time_zone = 'Singapore'

Besides that, instead of using Time.now, use Time.zone.now especially when you are printing time attributes as display to your users.

Why? Let's explore the difference:

irb> Time.zone = "Singapore"
=> "Singapore"
irb> Time.now
=> 2014-03-03 11:07:05 +0000
irb> Time.zone.now
=> Mon, 03 Mar 2014 19:07:12 SGT +08:00

Time.now uses the system time because it's is part of the Ruby standard library.

However, Time.zone.now returns the time, corrected to our Singapore time zone because it's been extended by ActiveSupport to have the zone method.

Therefore, in order for your users to see the time in SGT, use Time.zone.now in your code. In fact, use Time.current. Read the code to find out what's the difference.

What about DateTime?

irb> Time.zone = "Singapore"
=> "Singapore"
irb> DateTime.now
=> Mon, 03 Mar 2014 11:15:08 +0000
irb> DateTime.now.in_time_zone
=> Mon, 03 Mar 2014 19:15:14 SGT +08:00

DateTime returns the system time too because it's also part of the Ruby standard library.

The method to convert DateTime to our time zone is the .in_time_zone method which also accepts a time zone as a param.

Now then.. What do you do when you query the database with ActiveRecord?

You can use either Time.now or Time.current. Both will work because ActiveRecord will convert the times to UTC before sending it to the database, as all time columns in the database are also stored in UTC. Let's verify..

irb> Feedback.all
  Feedback Load (0.4ms)  SELECT "feedbacks".* FROM "feedbacks"
=> [#<Feedback id: 1, created_at: "2014-03-03 11:29:26", updated_at: "2014-03-03 11:29:26">]

Notice that created_at and updated_at columns are stored in UTC.

irb> Feedback.first.created_at
  Feedback Load (0.4ms)  SELECT "feedbacks".* FROM "feedbacks" ORDER BY "feedbacks"."id" ASC LIMIT 1
=> Mon, 03 Mar 2014 19:29:26 SGT +08:00

When you retrieve an object and print created_at, it will be converted to your time zone.

irb> Time.now
=> 2014-03-03 11:41:31 +0000
irb> Feedback.where("created_at < ?", Time.now)
  Feedback Load (0.5ms)  SELECT "feedbacks".* FROM "feedbacks" WHERE (created_at < '2014-03-03 11:41:32.687457')
=> [#<Feedback id: 1, created_at: "2014-03-03 11:29:26", updated_at: "2014-03-03 11:29:26">]

When you use Time.now, it generates a SQL query with time in UTC.

irb> Time.current
=> Mon, 03 Mar 2014 19:42:22 SGT +08:00
irb> Feedback.where("created_at < ?", Time.current)
  Feedback Load (0.4ms)  SELECT "feedbacks".* FROM "feedbacks" WHERE (created_at < '2014-03-03 11:42:23.549020')
=> [#<Feedback id: 1, created_at: "2014-03-03 11:29:26", updated_at: "2014-03-03 11:29:26">]

When you use Time.current, it also generates a SQL query with time in UTC.

Hope this helps to explain Time in Rails!

Spring And Failing Specs

Mon, 03 Mar 2014 00:00:00 +0800

Spring will be included in Rails 4.1 as the defacto Application Preloader. It's similar to Zeus which keeps your application running in the background during development, so that any rake or spec command will run speedily without having to reload the app.

Anyway, I was using spring, spring-commands-rspec and enumerize gems in my app and was running specs with spring rspec when this happened:

$> spring rspec

........F.............................

Failures:

  1) Feedback validations
     Failure/Error: it { should enumerize(:emotion).in(:happy, :sad) }
     NoMethodError:
       undefined method `enumerize' for #<RSpec::Core::ExampleGroup::Nested_3::Nested_2:0x000001029bcd20>
     # ./spec/models/feedback_spec.rb:22:in `block (3 levels) in <top (required)>'
     # -e:1:in `<main>'

Finished in 0.27399 seconds
38 examples, 1 failure

However, if I just do rspec, the specs passed!

$> rspec

......................................

Finished in 0.27061 seconds
38 examples, 0 failures

Randomized with seed 2852

Dug into the problem and found that this was a loading issue with enumerize.

This was how the enumerize matchers were loaded:

if defined?(::RSpec)
  require 'enumerize/integrations/rspec'
end

And it works fine for plain rspec command because rspec loads before rails. However, when spring starts the application, rspec is not yet loaded and so the matchers will not be required as well.

The fix is to load rspec like this:

begin
  require 'rspec/matchers'
rescue LoadError
else
  require 'enumerize/integrations/rspec'
end

This makes sure that the enumerize matchers will only required when rspec/matchers can be required without errors, which means it will only load in dev or test environment because you shouldn't have rspec bundled in your production environment.

Pull request over here.

Keep this in mind when you encounter weird failing specs when using spring.

Passion Drives You To Do Crazy Things

Sun, 22 Dec 2013 00:00:00 +0800

Many of you would know, I left Pivotal Labs/Neo Innovations in October to join a startup, and now, I'll be leaving the startup too.

Since the day I learned about Ruby, I know it's going to change my life.

Since the day I learned about Agile practices, I know it's going to change many people's lives.

I might have exaggerated a little, but right now these are the two things I am passionate about.

I believe that the World (or Singapore for a start) can do with more Ruby engineers, and companies can benefit from Agile practices such as Pair Programming, Test Driven Development, Continuous Integration, Retrospectives etc.

Having worked with Friendster.com, Viki.com, many other startups and even Government agencies while I was with Pivotal Labs, I am confident of the value that Ruby and/or Agile can bring to a company.

With that in mind, I have decided to step out to pursue my ambition of helping companies in Singapore grow through the sharing of my knowledge in these areas.

Being employed with a startup unfortunately doesn't afford me the convenience to help others or startups and although I have been doing this in my spare time (e.g. Rails InstallFest), I realized that I really want to do this on a fulltime basis.

It's crazy, but the calling is strong and this could be my version of self-actualization.

In the last year, I have met lots of people in the startup industry when I was running monthy Ruby meetups and organizing RedDotRubyConf 2013.

I realised that despite the availability of funding for these startups in Singapore, a lot of them do not have access to a ready pool of engineering talents or knowledge of how to manage an engineering team effectively.

This means that the startups have to first solve the problems of engineering shortage and engineering management before they can tackle the problem that the startup was originally formed to solve.

Hence I see an opportunity for me to try and help solve these problems with my experience in Ruby and in Agile practices.

As an experiment, I worked with a friend, who runs a company doing Joomla, for two weeks and shared with him Ruby on Rails and Agile practices.

We did pair programming and he's now a convert to Ruby on Rails, and exploring the possibility of using Ruby more for their projects. The Agile practices have also helped improved the processes in his company, and in his words "made my engineers very happy".

To hear that really touched me and further reaffirmed my decision to step out.

I have the ability to help make engineers happy and I should continue to do it!

So what am I going to do?

I have several ideas right now on what I can do, but nothing concrete yet. I'll be working for the startup until January to help ease the transition, though my notice period is just one week. After that, I'll take some time off for a few weeks to distill my thoughts and to firm up my action plans.

I could be doing Ruby on Rails courses, Agile consultation, Pair Programming etc. But whichever it is, my guiding principle is clear - I want to share my knowledge and do more for the local startup community.

Meanwhile, if you would just like to chat or you have advice for me, feel free to reach out!

Anyway, fail also ok. At least I tried! It's going to be an exciting 2014 ahead.

Singapore Rails InstallFest

Mon, 07 Oct 2013 00:00:00 +0800

Background

For the last year, I have been toying with the idea of starting a basic Ruby on Rails class to help people get started with Ruby on Rails.

I wanted to do it to grow the Ruby community in Singapore, and because teaching is FUN! However, I didn't have time to work on the materials, and so I never got anything started.

One day, I came across this thread on "Ruby or Rails Oceania" about Rails InstallFest, and I was immediately excited as it seems to be aligned with what I had in mind.

And so, I emailed Mikel of reInteractive to enquire about the possibility and he referred me to Chloe, reInteractive's Community Manager who was really enthusiastic about helping us get started with the Rails InstallFest.

The Event

Rails InstallFest happened on 5 Oct 2013, 9:30am at Silicon Straits CO.LAB.

At first, I was really worried about the turnout as it poured heavily in the morning on 5 Oct. But my fears were unfounded as we saw a total of 27 partcipants and 8 volunteer coaches. The particpants came from all sorts of background including product/business folks, developers, designers and students.

For our first Rails InstallFest, this is what the partipants have to do:

Get Ruby on Rails installed on your machine using the Rails Installer
Follow this blog post and get a blog deployed on Heroku

At the end of the day, all of the participants got Ruby on Rails installed on their laptops, and about half of them got a simple blog deployed on Heroku. Yea!

Thank You

reInteractive really helped a lot in making the first Singapore Rails InstallFest a success. Especially Chloe, who helped made sure I have all that I need - materials and vouchers! Without them, the event wouldn't have happened. Thank you reInteractive and Chloe!

We also required a venue for the event, and James readily offered his space for our use. Silicon Straits was really cosy and conducive for learning as remarked by a number of the participants in their feedback, and I am really grateful to James for loaning out his space! Thank you James and Silicon Straits!

And finally, all I can say is that the Singapore Ruby community is sooooooo full of love! After I annouced the event and asked for volunteer coaches, many stepped up!

Thank you Chong-Yee, Jeffrey, Jaryl, Kevin, Paul, Ray Chuan and Sean for volunteering your time on a Saturday morning as coaches!

Improvements

We collected feedback from the participants and the overall sentiment is that the class was a really useful session to help them get started in Ruby on Rails.

At the same time, most remarked that the volunteer coaches were friendly and helpful. Together with the conducive environment, they found learning RoR to be easy.

However, it would have been better if we did a brief walkthrough of the blog post exercise, so that the participants can better appreciate and understand what they were doing, and not just blindly follow the blog post.

In addition, there's also a suggestion to group the participants into small learning teams so that they interact and learn from one another within their small groups.

Overall, I enjoyed running the first Rails InstallFest in SG, and I am sure we'll continue to have more RoR training sessions in the future. =)

Time for some pictures!

OpenURI's open, Tempfile and StringIO

Wed, 02 Oct 2013 00:00:00 +0800

OpenURI#open has a dual personality. Consider the code below..

require 'open-uri'

zipfile = open("http://example.com/file.zip")

With this, zipfile can either be a Tempfile instance or a StringIO instance.

Apparently, that's how OpenURI#open works, and this StackOverFlow post explains it well. And as mentioned in the same post, to get OpenURI#open to always return a Tempfile, apply the initializer "hack".

require 'open-uri'

# Don't allow downloaded files to be created as StringIO. Force a tempfile to be created.
OpenURI::Buffer.send :remove_const, 'StringMax' if OpenURI::Buffer.const_defined?('StringMax')
OpenURI::Buffer.const_set 'StringMax', 0

Or, you can just be deterministic and create the Tempfile yourself.

Ruby - Reading A Remote Zip File

Tue, 01 Oct 2013 00:00:00 +0800

I need to access a remote zip file and this is something that works:

# In the console, gem install "httparty"
require "httparty"

# In the console, gem install "rubyzip"
require "zip"

# Get the contents of the remote zip file via HTTParty
# and write it into a temp zip file

zipfile = Tempfile.new("file")
zipfile.binmode # This might not be necessary depending on the zip file
zipfile.write(HTTParty.get("http://localhost:3000/file.zip").body)
zipfile.close

# Unzip the temp zip file and process the contents
# Let garbage collection delete the temp zip file

Zip::File.open(zipfile.path) do |file|
  file.each do |content|
    data = file.read(content)
    puts data
    # Do whatever you want with the contents
  end
end

The code is simple, but at the start, I kept getting an error when unzipping the temp zip file, and I thought I was doing something wrong.

End-of-central-directory signature not found

Did some debugging and finally figured out that the problem was with the remote zip file - because the file was not fully constructed even though I had a link to it.

The remote zip file link was actually returned by an earlier API call to an external service that also triggered the building of the remote zip file.

Moral of the story? Trust my code.

Anyway, RubyZip is poor in performance. Might want to try ZipRuby instead.