The Razor Templating Engine by Microsoft

By Aaron Ransley | Published: July 8, 2010

Microsoft has recently announced (and released, as part of WebMatrix) a new templating engine called Razor. Razor uses a parsing style which results in succinctly delineated template files. For example:

This will output:

  • the contents of the name variable
  • the current year
  • and a link to the details page for a particular product.

Flow control statements are similar, delineated intelligently with braces:

Multi-line statements look like this:

Many have expressed concerns of wheel re-invention. While a valid concern, I feel it pales in comparison to the promise of progress and novel techniques that come with 1st party backing of an alternate, less verbose templating engine. For example: full IntelliSense support for alternate view engines is practically non-existent in projects like Spark or nhaml.

Recently, Scott Guthrie commented on a feature request, stating the Razor templating engine is very likely to have it’s source code released on CodePlex. Yup, just like MVC, MS is likely releasing the source. This is excellent news, since we can all take a stab at branching and improving Razor as we see fit.

As an aside, please see this blog post highlighting recently released .NET 4.0 libraries, now open source and completely compatible with the open source Mono project.

I’ve also read that Razor has a distinct lack of dependencies. Since the engine is simply generating strings from particular blocks of C# code, there is discussion of having Razor handle mail-merges, small-time business logic scripting, and so forth. It needn’t be married directly to the ASP.NET platform, or even the Web environment.

Bring it on!

Posted in dork out | Leave a comment

Loving iFixit’s Website

By Aaron Ransley | Published: April 3, 2010

Check out: http://www.ifixit.com

There is something about this Web site that makes exploring fun. The content is laid out in a dense way, so your eyes are always “satisfied” with some bit of info or graphic. I could see some considering the layout crowded — “Woah man, where’s my whitespace?”. That’s not me, though. I dig the brainfood aspect of the experience.

It’s debatable whether or not every geek should explore the tools they use every day, but iFixit puts that question aside and just plows ahead. They end up going a long way to make you feel in control of your devices! From hard-drive assemblies to uni-body Macbook Pro laptops, instructions and pictures are abundant and all encompassing.

I personally was able to disassemble my laptop and install a new fan when needed, thanks to these exceptional guides. I’m all for the thrill of blind exploration, but when someone has already done it (and learned how not to break chassis mount 2B, for example) I’ll gladly take the free advice.

Check it out! Maybe that broken [xyz] you have lying around might be fixable… By you!

Posted in dork out | Leave a comment

Automated Testing

By Aaron Ransley | Published: February 9, 2010

In the spirit of sharing, I’m going to talk a bit about how automated testing has stolen my heart.

I’ve been working with some form of Web development for roughly 7 years now, but have had surprisingly little experience in maintaining a codebase over time. I’ve never liked this fact.

Due to the one-off nature of many freelance projects and personal experiments, there is often little that needs to be maintained or iterated upon after the fact. It’s like a software design utopia; nothing changes. I’ve always been aware of this absurdity and enjoyed it to the fullest, I assure you :)

Now, I work for a company and have a salary and have legacy code to maintain. I also write large amounts of production code, employing nebulous and ever-changing requirements. Sounds more like the software industry, doesn’t it?

This leads me to where I assume many are led: automated testing. I’m learning about unit testing now, but have also touched briefly on integration testing. Most interesting to me so far is the implementation and architecture required to set up test scenarios effectively. Interfaces and DI, oh my! It’s a perfectly well defined example of modular code which feels extremely natural to me.

Coming from a workflow that had me entering data and clicking widgets manually, I actually gain productivity. On top of that, the obvious benefits of overall runtime and structural stability lead to more “confident code”.

Sadly, I still feel resistance from higher-ups regarding unit testing being “too expensive” for small shops, so I am working on compiling citable studies.

Wish me luck :)

Thanks for reading!

Posted in dork out, i'm learnin!, web | 2 Comments

Humanized Output

By Aaron Ransley | Published: February 3, 2010

Shown here is a snippet of the woot! homepage. I enjoy this little widget particularly because of the way excerpts from the forums are prefixed. The human readability of “[user] would like to know…”, “[user] reports back…”, and “[user] had some troubles with a…” is significant.

Go organic with your Web application’s verbiage, especially in areas where users are likely to be in a hurry. “About two days ago” as a timestamp is wildly more helpful than “53 hours ago” or “2-1-2010 6:00 PM”. While it’s no replacement, widgets and primary listings are great places to use a more natural language.

And tomorrow, when I check my schedule and see my next meeting is in “about 3 hours”, I’ll remember. When I am in a hurry I can throw “[blah blah] 7pm tomorrow” at my calender app and it will just work. It feels comfortable, and it encourages smooth interaction.

Posted in dork out, web | Leave a comment

Website Launch: Girlboy Greetings

By Aaron Ransley | Published: January 31, 2010

I’ve just helped design a site for Girlboy Greetings!

Girlboy Greetings is a small greeting card establishment from Los Angeles, CA. The designs are quirky and are offered in a variety of colors. My favorite aspect of the cards is how they offer glimpses of everyday thoughts and places.

geek::facts();

I took a radically different approach with site conception and design in that I put the site together in Photoshop prior to any HTML development. This feels like a graduation from my previous design mentality, which had often sent me bouncing between divsdivsdivs and “oh-god-where’s my vector shape tool”.

The budget was small and time was tight so we went with a small 3 page, primarily static site. A Rails application with only 4 controllers runs the whole thing. The Paperclip plugin is in use by the Card model to handle semi-intelligent cropping of uploaded images. An implementation of a jQuery driven cropper is a to-do. On the front end, all is standard CSS with some basic jQuery effects. Fancybox rounds off the card display.

Posted in dork out, web | Leave a comment