Khoi Vinh’s group at the NY Times is thinking about embracing and extending the Firefox and IE7 feed icon. One of their designers proposed the following modifications:

Really very nicely done. What I don’t get is how the ‘RSS 2.0’, ‘ATOM FEED’, and ‘XML’ labels are going to make things clearer. And what the hell is ‘XML VIDEO’? Some experimental new XML-based video format?

The whole point of the new feed icon was to eliminate the confusion brought on by all these different acronyms. There’s really no reason to offer the exact same content in different feed formats.

The problem with the NY Times podcasts page is not that it offers the same content in different feed formats, but that it allows you to subscribe to the same content with different tools.

You can click to add the subscription to your default feed reader, you can click to add the subscription to your podcasts in iTunes and you can click to add it to your My Yahoo page.

Want to make this less confusing? Here’s how:

Subscribe (default)
Subscribe in iTunes
Add to your My Yahoo page

The goal for this new meeting is to have a few developers talk about the stuff they do with Rails. It’s going to be a fast-moving and informal meeting; talking time will be limited to 10 minutes and there will be lots of opportunity for asking questions. We’ll have both dutch and english language speakers.

What: 6 short presentations by Rails developers

When: Thursday, June 22nd, 2006, 14:30-16:30

Where: The basement of the Greenpeace International headquarters at the Ottho Heldringstraat 5 in Amsterdam

Registration required: Leave a comment with your email address to register for this meeting (click ‘also leave contact information’).

Mark started the naming game again, then Sam and Simon quickly joined in.

Why not just call it ’web’?

SanDisk is running a fake grassroots guerilla marketing campaign to convince us that owning an iPod is unoriginal and that iPod users are followers.

Good luck to them. I for one am not offended by images of animals with white airbuds. I think the chimps and cows and sheep are damn cool.

Canon LC210, Samsung SGH-P300 and Apple MacBook.

This week I wrote a small web tool in Camping but when I wanted to secure it a little bit I noticed that there was no default way to do HTTP basic authentication. So I wrote some code to do just that. Actually I wrote a lot of documentation and a little bit of code.

With just some downloading and three simple steps you can add basic authentication to you application.

Download the basic_authentication.rb file.

1. require 'basic_authentication'
2. Mixin the BasicAuth module:

   module Blog
     include Camping::BasicAuth
   end

3. Define how you want to authenticate users:

   module Blog
     def authenticate(u, p)
       [u,p] == ['admin','secret']
     end
     module_function :authenticate
   end

And you’re ready to start not letting people into you application. If anyone has a better idea on how and where to define the authenticate method, please drop me a line.

You can use this code under the same restrictions as the Camping framework. Now let’s just hope Why doesn’t sue me for using the Camping module (:

Update: Thijs found a bug in the header handling. The link above has been updated. I will try to give this code a home somewhere and some proper version information when I have some time.

I almost fell of my chair laughing when Avi skyped me just now and this popped up:

Yes, Avi does corporate imaging and information design. He’s also the one who came up with our new logo.

Last week we’ve started working on ‘Rhubarb on Rails’, a new version of a Greenpeace application for running letter writing actions. It’s going to replace a PHP based tool that was a quick replacement for an old Zope application that decided to roll over and die somewhere in the fall of 2005.

The nice thing about working for the Greenpeace development team is that they thoroughly get the Agile development methodology. Even better is that they do a lot of their work ‘in public’ as you can see on their Work in Progress weblog.

Rhubarb will be released under an Open Source license, so it can be used and improved on by other big NGO’s or small local organisations. You follow our work and download the source from our development site.

Update: The project has been renamed to ‘Write-A-Letter’.

I just found out that in Spain they’re drinking beer instead of coffee at Rails meetings.

Martin Fowler in ’Evaluating Ruby’ (emphasis added):

Too many things are hard to judge that way - hence we spend so much of our time on client projects being slowed down by technology that sounded good on a golf course. My preference is to make this judgment based on experience - find people who have a track record for delivering in the mainstream environments and who have tried using Ruby.

We recently received a two page REQUEST FOR PROPOSAL for a ‘Social Network Site’:

[…] has developed a concept that requires the development of a niche social network solution, incorporating some of the most innovative features of Web 2.0: RoR, Social Networks with extensive search features, Web Services, Blogs, Tagging, Photo Management, RSS, Internal Messaging, etc. Users are able to sign up, create a profile with pictures and personal data, set preferences and demographic information. This site differentiates itself from the other social networks by providing innovative features in a specific niche.

[…] would like the site to contain the following features.

• User Registration
• User Profile Setup
• User addition to a Network
• Photo management
• Ability to post personal blogs
• Ability to comment on a members profile
• Messaging capabilities
• RSS integration
• Ability to search the social network with varying search parameters
• External partner data feed and integration through SOAP/XML API’s
• Extensive niche specific features (further details in Phase 2).

Based on this information, we were asked to ‘provide a detailed estimate with breakdown by task (description of task, duration, total number of hours, etc.). ’ and to ‘a timeline for completion of the project (start date, completion date, milestones)’. Within 5 days.

Maybe I’m being arrogant here, but I don’t see how someone can realistically expect a developer to write a fixed time, fixed budget proposal based on this game of buzzword bingo. Can you?

I’m all for self-publishing, downloadable PDFs, no dead-tree version and all that, but why can I buy a single copy of Getting Real for myself only?

There are at least a handfull of clients I would love to buy a copy of this book for, but there’s no way to do that right now. And no, I don’t want to ‘encourage them to purchase their own’, I just want to give it to them and tell them they might like to read it.

A university that shall remain nameless wanted to provide its students with an easy web-based way to store, share and publish documents online. They’ve decided to go with a hosted solution based on Sharepoint Portal Server. This only works properly with Internet Explorer on Windows, so you’re out of luck if you own a Mac, if you run Linux, or if you like to use Firefox. It’s also slow, clunky and somewhat confusing.

Oh, and the TCO for this solution is €16 per student per year. So with 30000 students, this system will cost the university €480000 each year.

We created a simple demo application (QuickTime screencast) to illustrate how we think this application should work. Our estimated TCO for a full-featured version is €4 per student per year, so with 30000 students, the total costs will be €120000 per year.

The requirements called for 250MB of storage for each student, so a big part of our TCO estimate was based on a cool 15TB storage and a huge amount of bandwidth.

In the last few week we’ve been playing with the Amazon S3 storage service and we’ve created a version of our demo application that uses S3 as a storage back-end. Amazon charges $0.15 per GB per month for storage used and $0.20 per GB of data transferred. This translates to a rough yearly estimate of $45000 (that’s currently about €35600) for storage and bandwidth.

Now let’s combine this figure with a one-time investment of €40000 for the development work and €2000 yearly for hosting, support and maintenance of the front-end application. For the sake of the argument, let’s spread the costs our over 3 years. This gives us a yearly cost of about €60000 and a TCO of about €2 per student per year.

With S3 you only pay for storage and bandwidth you’re actually using. This means you probably never reach this figure because most students will never use their full 250MB and the system will only slowly fill up.

This is all just a rough estimate, so I am likely to have it completely wrong. But probably not so wrong that you ever need to spend nearly as much money as this university is doing now.