Keyvan Nayyeri

It was quite a long while that I had decided to migrate my blog from Graffiti CMS to a new engine, and today I finally completed this task.

There is a long story to tell about this process and my reasons to choose my own blog engine after over 4 years of using Telligent products (Community Server and Graffiti CMS), and I try to talk about them shortly in this post.

Why Migrate?

Old followers of my blog know that I’ve been using Telligent products since the beginning of this blog for a very long time. First I was using Community Server as a single-blog and then migrated to Graffiti CMS as a more appropriate option for my blogging style. I used to be an active community contributor to Telligent products, ran popular open source projects for both products, and praised both these products as they really deserved it at the time of being. I used to be a Community Server MVP and a prominent name in Graffiti CMS development, and could find great friends inside Telligent and via this company and its products. Many of my old readers had come here from my initial activities on this community and suddenly we became great friends.

However, as all of you already know, this economic crisis had a huge impact on the way that Telligent leadership were managing the company, and after many layoffs during the last year, they restricted the activity around their products. The main product that was negatively affected by this movement was Graffiti CMS. There has been a long time that Telligent hasn’t contributed any update to this product and this has been the source of many complaints by customers and community members recently.

Here I’d mention that Graffiti was a great product as was, and it’s still a thorough product for its own whether as a CMS or a blog engine. However, I don’t want to deviate from the main point of this post by talking about Telligent and its products, and why many of the community members including myself believe that Telligent had to drop new builds in the past months.

But the reason that convinced me to move off Graffiti and write my own engine was the fact that most likely I won’t have time to spend on maintaining the technical side of this blog in the next few years, and I thought that it’s better to have something simple and easy to maintain that I have full control on. Graffiti was written for ASP.NET 2.0 and IIS 6.0 and there was no source code available, so it could cause difficulties in the future to manage this product.

Destination Options

It’s a matter of a fact that after more than 4 years of blogging, I’ve realized that the type of a blog engine and the technology behind it are not a big deal whether you’re a technical developer or not. All an author wants is a tool to go, and almost all the current blogging engines are satisfactory in this matter specifically for me as a blogger who has stopped categorizing and tagging his posts. I really need a plain blog engine that works for me.

I had many destination options to use. I was very interested to use an engine powered by ASP.NET MVC, but neither Oxite nor AtomSite were what I wanted. Both these options add an extra complexity that doesn’t have any value for my needs.

There were also some great engines written by ASP.NET WebForms such as Subtext or BlogEngine.NET but less or more, they couldn’t satisfy my requirements, too.

Looking around, I was seriously going to migrate to a PHP option like MoveableType or WordPress, and even used Jon Sagara’s tool to export my posts, but somewhere in the middle of the process I disliked the quality of my imported posts in the destination engine. It was breaking all my URLs and couldn’t carry all the data that I had.

Why Write My Own Engine?

Having this in mind, I thought about writing my own blog engine. To be honest, I dislike writing blog engines as I think that world is saturated by various engines available out there. I’ve contributed to many open source projects but never wrote a blog engine myself because I don’t believe in its necessity. When there are many engines available, why should we start a new project, drop one or two releases, and then stop the project after a while to make difficulties for our users?

However, I was locked in the middle point of a process where I couldn’t satisfy my needs with any of the existing blog engines. My blog is old and there were over 1000 posts (mostly lengthy with source code samples) written by me with many feedback. It wasn’t easy to find something suitable for this amount of data and type of content. I was looking for a simple blog engine that can import my posts and maintain the huge number of incoming links to my blog that have become a reference for the community over the years.

Furthermore, it was a while that I wanted to use ASP.NET MVC to power my blog engine, and I had promised some of our book readers to build a blog engine similar to WroxBlog that I’ve written for Chapter 18 of the book that lets them learn about the backend part of the blog.

Therefore, I decided to write a simple blog engine for my own that meets my goals and doesn’t make things complex

What is Behistun?

After spending much time on finding an alternative, I started writing my simple blog engine and called it Behistun. Behistun is the name of an ancient UNESCO World Heritage Site near Kermanshah, Kurdistan (Iran) where I was born.

I wrote Behistun in 7-8 hours then spent 4 days to end up with this simple theme and wasted 7-8 days to move the images and other files of my old blog to a single location to make them consistent!

As you see, Behistun is pretty simple. It’s written with ASP.NET MVC 1.0 (C#) and applies SQL Server as its storage. I used LINQ to SQL to interact with the data storage that consists of four tables only. I also used WCF for some parts such as RSS feed generation and XML-RPC service hosting.

Behistun is as simple as following features that are all built for single-blog functionality.

• Post management
• Comment management
• Trackback and Pingback (the latter one is not enabled yet)
• Monthly archive
• MetaWeblogAPI
• Sitemap
• FeedBurner support

The first and foremost advantage of Behistun is its simplicity, but it’s also a cool engine as it’s built by ASP.NET MVC, generates neat URLs, and loads pretty fast!

In order to build Behistun I used some third party tools and components that I list here:

• Ninject
• RhinoMocks
• CKEditor
• Ryan Lanciaux’ HtmlHelper for Gravatar
• Eran Kampf’s RSS Feed Action Result.

There is only one minor issue to inject my XML-RPC services built with WCF that I’m working with Ninject team members to resolve.

The theme for this blog is a simple one as the engine is. I inspired some ideas in coloring and fonts from Posterous design.

Major Changes

During the migration process I applied some major changes in my blog:

• Now I’m using Google Custom Search as the search engine for my blog. The search engine for Graffiti was pretty slow for this number of posts, and I saw no point in reinventing the best wheel that is ever built for the search on the web!
• There is no excerpt for posts anymore. You see the full content of the posts in the homepage.
• There is no About or Contact page anymore. You see a short biography along with a few links in my sidebar that you can use to contact me (don’t forget that I’m not that type of guy who accepts any friends request on social networking sites).
• After a long while, I restored my monthly archive with tons of blog posts that shows the power and age of this blog!
• Once again I migrated my feed to FeedBurner! This time just to make you happy and let you laugh a lot!

Having these changes in look and feel and functionality, I’m also going to apply major changes in my blogging style in a smooth manner. You’ll be witnessing a transition from my current focus on Software Development and .NET-related stuff to other topics that I like them more than these.

New License

The very important change is that I changed the license of my blog to a copyleft GNU Free Documentation License. I’m a big advocate of openness in software (don’t look at my Microsofti face!), so I thought it’s worthwhile to open all the doors for anyone to use this resource of content about .NET and Software Development and spread the word. There are some content stealers that steal my posts every time I publish them, and there is nothing I can do. At least, I can help them do a legal job!

Behistun as a Public Project?

The short answer is no, not at the moment! As I said above, I don’t like to add something to this mess of blogging engines and stop supporting it after a while.

However, the good news is that at the moment I’m using an Alpha version of Behistun and once I complete working on the first version as a stable engine, I’ll publish it as a disclaimer project with a very open license, so anyone can adapt the code and use it with no support on my side.

Having this said, if I can find 1-2 contributors who are really interested to get Behistun and make it a public open source project, I’m ready to cooperate with them.