In the race to build credibility with the development community, Ruby on Rails wins the battle to be cool hands down – look for a new web 2.0 application, or hunt down a quintessential garage-dwelling developer and it’s a safe bet they live in Ruby. That doesn’t change the fact however that when it comes to true enterprise use, it is .NET that has the biggest adoption level – indications that I’ve heard suggest that, who tracks API calls closely, have noticed that by far the biggest proportion of calls they receive come from .NET applications.

This then raises some opportunities for companies looking to aid in the creation of .NET applications but on cloud infrastructure, and without the concerns about managing underlying infrastructure. While it is true that Microsoft itself is going “all out” (or was it “all in”?) to get with the cloud – it’s probably fair to say that Azure hasn’t been the stellar success that one would have expected from Redmond.

This is where App Harbor comes in – quite simply, AppHarbor is Heroku for .NET – in the same way that Heroku allows Ruby developers to concentrate on code and forget almost entirely about servers, so to does AppHarbor aim to offer the same value to .NET developers. Founded only last September, AppHarbor had a prototype in December and pushed out a public beta in January of this year. They’re not shy about stating their case – the AppHarbor website says in no uncertain terms that they are “Azure done right”.

AppHarbor was founded by three Danes, Rune Sørensen, Troels Thomsen and Michael Friis. I spent some time talking with Friis in advance of a product announcement from them. Firstly we had a bit of a chat about the rationale behind, and motivation for, App Harbor. After that we delved into the product annjounement itself.

Friis wasn’t shy to tell about the deep relationship they have with the folks from Heroku. he even admitted that when development opportunitites, roadmap priorities nd other questions become apparent – AppHarbor makes their decisions using a simple test – “what would Heroku have done?” More than simple role-modeling however, it’s safe to say that Heroku (and it’s impressive exit to salesforce) showed the world the way to build a successful PaaS ecosystem – AppHarbor could do plenty worse than follow Heroku’s lead. Like heroku, AppHarbor gives developers infrasturture management through the following process;

  1. The developer pushes code
  2. AppHarbor builds the site
  3. AppHarbor runs the unit tests
  4. AppHarbor deploys and scales the site to multiple servers

AppHarbor are today announcing another move that follows one taken by Heroku – they are setting up an add-on program that closely resembles hat of heroku. In fact they are even going so far as to emulate the Heroku add-on program API so that add-ons can work with either PaaS. That’s super useful for developers who want to use tools such as MongoHQ, Redis and Cloudant. It’s also an oppoturtnity for .NET specific add-ons such as RavenDB to become part of an ecosystem that is aiming to change the world for .NET developers. The add-on gives the following;

  • A self-service portal for third-party service providers
  • A single, easy-to-use interface
  • Easi integration into the applications built on AppHarbor platform

AppHarbor is doing exciting things – it’s not a stretch to imagine a nice exit for the founders in the imminent future. That’s be my pick anyway.

Ben Kepes

Ben Kepes is a technology evangelist, an investor, a commentator and a business adviser. Ben covers the convergence of technology, mobile, ubiquity and agility, all enabled by the Cloud. His areas of interest extend to enterprise software, software integration, financial/accounting software, platforms and infrastructure as well as articulating technology simply for everyday users.

1 Comment
  • Interesting that the most API calls come from .Net. I wonder if the Salesforce for Outlook adapter which would be the most common API use case is included in that

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.