Design Patterns In Ruby

• Russ Olsen
• Programming

Deploying Rails Applications

• A Step-by-Step Guide
• Ezra Zygmuntowicz, Bruce Tate and Clinton Begin
• Programming

Everyday Scripting with Ruby

• for Teams, Testers and You
• Brian Marick
• Programming

Amazon Web Services

In my iPhoto application I now have over 6,000 photos. Some I've taken with my Canon digital camera, but the majority I've scanned from negatives and added to iPhoto. With over 120 rolls of 36 exposure film to scan that little task took quite a while. But that was back in 2004.

My backup drive is a LaCie firewire drive that’s partitioned. One section is for a full iMac backup using SuperDuper!, the other, at the moment just contains a Time Machine backup. Just as a little aside, I was thinking of getting a Drobo and having loads of backups across lots of discs, but I've since gone off the idea.

So, the worry has been “What if my Mac and backup drive are destroyed by fire, power surge or are stolen?” I don’t have an offsite backup. I did think of just burning the photos to a couple of DVDs and putting the discs either in the garage or in my Dad’s safe, but that isn’t exactly ‘hi-tech’ for this day and age.

Reading around the internet I decided to try Amazon S3 (that’s Simple Storage Service) a part of Amazon Web Services. You are basically charged a per GB fee for storage per month, per GB transfer in and out in a month and for PUT, LIST and GET requests. It sounds more complicated than it is but at the moment my bill for the month is running at $1.73. Which with the dollar/pound exchange rate is very little indeed. That is after uploading nearly 11 GB of photos. The actual transfer in fee took the most of the total at $1.08.

To get my photo folder to S3 I used a set of Ruby based programs called S3Sync. I have been using rsync to download backups from my web-server, so I knew what to expect. After setting up a web services account with Amazon I added S3 to my set of connections in Transmit. It was this I used to create my first ‘buckets’ (a folder by any other name) on S3. This was fine, copying files and folders just like normal FTP. The only snag being that Transmit won’t, at the moment, let you delete any buckets that you've created. It just doesn’t happen. It even struggles deleting folders. I sent Panic an email and haven’t heard back, so I presume they know about it and are working on a fix for the next release. The only way to delete buckets, that I've found, is to install the S3 Organizer for Firefox, this works like a charm.

The majority of the tinkering took place last Thursday. Since then my iMac has been awake, during the day at least, and securely transferring photos to S3. It’s worked so well that I've set-up another bucket to take backups from my web-server directly to S3. I used to use the rsync command that I mentioned to download them to my iMac but from now on I don’t think I’ll bother.

So far the whole Amazon Web Services and S3 experience has been quite enjoyable. I just have to think of other uses for it.

Redesign

Yes, I felt that a little redesign was in order.

I wrote a little Ruby script the other weekend to scrape my Digg RSS feed so that recently Dugg links could be listed on my blog. All it does is read in the XML, cut-out the titles and links and writes them out to a dugg.php file as list items. The file is then included in the sidebar. This is done, more or less, every hour for the hours in the day that I'm awake and could be Digging. All done on the web-server with a cron job. Of course Digg have now released an API which I will no doubt tinker with at some time. For the moment this way works fine.

This is the first time that I've written the stylesheet from scratch. The last one was based on Kubrick, the Wordpress default and so had lots of elements that I didn’t need or use.

A sidebar has been added which either has a double or single column based on the content. That was to make it easier to navigate without scrolling down to the links in the footer.

Each theme is still generated by a little Ruby program that I wrote. The switching is done with a change to the Wordpress stylesheet selector which switches the theme based on the current week.

I wanted to be able to generate the shades of a particular colour.

This site has many themes with one main colour for the links and lines. But I wanted to be able to input the hex colour, that I had picked from the banner photo, and find a lighter shade for the menu section at the bottom.

Eric Meyer has written a Color Blender and I thought that I could do something similar using Ruby.

The program just takes the colour in hex (e.g. #ff0000 or even #f00), the shade number that you want and a number for the scale. Positive shade numbers make the colour lighter, negative numbers darker. Scale tops out at 100 and defaults to 10.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

def colour_shade(colour, shade, scale=10)
  # default the scale to 100 if it is greater...
  scale = 100 if scale > 100

  # if the shade is out of the range of scale then reset shade to scale...
  if shade > 0
    shade = scale if shade > scale
  else
    shade = scale * -1 if shade < scale * -1
  end

  colour = "##{colour[1,1]*2}#{colour[2,1]*2}#{colour[3,1]*2}" if colour.length == 4

  # Convert the colours to decimal...
  colour_split = [(eval "0x#{colour[1,2]}"), (eval "0x#{colour[3,2]}"), (eval "0x#{colour[5,2]}")]

  # Calculate the shade...
  new_colour = "#"
  colour_split.each do |c|
    case
    when shade <  0: colour_shade = (c - ((c / scale) * shade.abs))
    when shade == 0: colour_shade = (c)
    when shade >  0: colour_shade = (c + (((255 - c) / scale) * shade))
    end
    new_colour << sprintf("0%x", colour_shade.to_i).slice(-2..-1).upcase
  end
  new_colour
end

Hover the mouse over the bars below to see what parameters were used to create the shades.

Download the source code.

Agile Web Development with Rails

• Dave Thomas, David Heinemeier Hansson With Leon Breedt, Mike Clark, Thomas Fuchs and Andreas Schwarz
• Web Programming

Out with the old

My little web-site has had a quick re-design over the weekend.

The ads and badges have gone from the left sidebar and it’s been dropped to the bottom. The links sidebar on the right has disappeared altogether. It did seem to get out of hand, with every web-site that I thought was good getting a mention. The links are still in WordPress so they might reappear as a separate page, especially if there’s a link that I want. It’ll be interesting to see if my Google rank or my site hits per month drop off the chart.

The scrolling idea was pinched from poocs.net, oh you haven’t found it yet?

I did contemplate switching to a Ruby on Rails blog engine after looking at the code behind the quotes WordPress plugin. It’s all in PHP, obviously, and I couldn’t figure out how to just list 250 quotes on a page. At least I know more about Ruby and Rails now so I could write something in that instead. I looked at Typo again and downloaded a few of the themes. The Lucid theme really caught my eye by Jamie. I tried to get Typo working locally so that I could import my WordPress stuff but kept having trouble even connecting to the admin pages. I never have trouble with my own Rails applications I just run webrick in the directory and that’s it. I also had a look at Mephisto and couldn’t get that to work either. I'm way too busy with my own little Rails application to look into it further.

Programming Ruby

• Dave Thomas with Chad Fowler and Andy Hunt
• Programming

Ruby on Rails

I've been busy learning, or at least trying to learn, Ruby on Rails. For people who don’t know this is a new programming language/application which allows programmers, like myself, to create web applications really fast.

At work we have one of our applications written in ASP.NET and since the beginning of last year I have been the unfortunate soul who has to debug and maintain this piece of software. At first I did think that it was pretty good. Towards the end of last year I had an idea to write some web pages linking to a database so that we could store client information, contacts, which clients have what software, what platform they are using etc. Just, more or less, everything that we need to know, all in one place. I felt that this was a good idea so I started writing it. I set-up the database and the tables in SQL Server and started to code the pages having copied chunks of code from our web application. Of course it has never been finished.

Let me warn you that if you work for a software company and one of the managers, who we’ll call Scouse Boy No. 1, buys a Rolodex to store contact information then leave immediately. Just run as far and as fast as you can and don’t look back.

This year I've had to convert our ASP.NET application so that it works with the new version of our main software. Even after having ASP training last year it was a slow and tortuous task. The fact that one of the pages doesn’t render correctly in Firefox, that it’s all tied into Microshites® web server and probably doesn’t validate to any W3 standard makes me think that it’s not going to be a possible career path. Maybe if we had spent £500 on Visual Studio things may have been less painful.

So, Ruby on Rails, is just the kind of thing that I've been looking for. A development environment written by programmers to make life easier so that more time can be spent writing clean, well structured, code. I ordered the Agile Web Development with Rails book, downloaded the necessary software for Tiger and started to work through a great little Ruby tutorial by Chris Pine. When that was done I started on the book proper, reading through the first section before promptly coming to a grinding halt on page 57. This is when I had a MySQL database and tables created and was about to see the power of Rails for the first time. I must have spent at least 2 days, on and off, trying to figure out why the views of the table weren’t being created. I even un-installed the latest version of Xcode and and re-installed the old version from the Tiger DVD, all to no avail. It just wasn’t seeing the database for some reason. Last Saturday afternoon I found the solution. In the config/database.yml file specify the username, in the book ‘dave’, that you granted database access for. Then re-run the ruby script/generate scaffold Product admin and you should see app/views/admin pages being created. Since then I've been working through the Book Store application. Everything is so easy, well not exactly easy, but just a joy to use. The scaffold that is generated by Rails knows what the database table looks like so you don’t have to key in lines of HTML with the different types of input code embedded in it. You don’t have to give each field a name, as this is already done for you from the field names. Links for edit/list/show/delete are already created and all the pages behind them. With one Rails command I generated a set of table maintenance pages. This is something that would have taken me at least a few hours using ASP.NET. And all without writing SQL statements.

I think to get the most out of the book you do need to have your own application to work on yourself. It would be really easy to just blindly get the demonstration application working without understanding how the principles can be put into practice. It hasn’t all been plain sailing. Just last night none of the controllers for my application would display in Safari. I found that emptying the cache and deleting any localhost cookies did the trick.