Well, it's been a wonderful ride, nearly half a decade working with BlogEngine.net's great blogging software. But it's time to move on.
Orchard, it was very nice to meet you. You have a wonderful future ahead of you, and I was honored to have known you, even just a little. Unfortunately, you and I are each looking for something different.
WordPress, you are like a beautiful, sexy whore, tantalizing on the outside and known by everybody and his brother, but quite honestly I'm not sure I want to see you naked more than I already have.
I'm frickin' Jon Davis, I've been doing software and web development for 14 nearly 15 years now, and doggonit I should assume myself to be "all that" by now. Actually, blog engines should be like "Hello World" to me by now. I suppose the only reason why I've been too shy to do it thus far is because the first time I started building a complete blogging solution eight or nine years ago and stopped its continuance six or seven years ago the thing I built was proven to be an oddball hunk of an over-programmed desktop application that I had primarily leveraged to grow broad technical talents. It was a learning opportunity, not a proper blogging solution, and it smelled of adolescence. (To this day it won't even compile because .NET 2.0 broke it.)
In the mean time, I've moved on. I've been an employer-focused career guy for the last five or six years, having little time for major projects like that, but still growing both in technical skill set and in broad understanding of Internet markets and culture.
But I kind of miss blogging. I used to be a prolific blogger. I sometimes browse my blog posts from years ago and find some interesting tidbits of knowledge, in fact sometimes I actually learn from my prior writings because I later forget the things I had learned and blogged about but come back to re-learn them. Sometimes, meanwhile, I'll find some blog posts that are a little bizarre--thoughtful in prose, yet ridiculous in their findings. That's okay. My goal is to get myself to think again, and not be continuously caught up in a daily grind whereby neither my career nor technically-minded side life have any meaning.
Last weekend over two or three days I created a new blog engine. (Anyone who knows me well knows that I've been tinkering with social platform development on my own time for some years, but this one was from-scratch.) I successfully ported all of my blog posts and blog comments from my BlogEngine.net to my new engine and got it to render in blog form using my own NUnit-tested ASP.NET MVC implementation. I would have replaced BlogEngine.net here on my site with my blog engine already, were it not for the fact that as I used Entity Framework Code First I ran into snags getting the generated database schema to correctly align with long-term strategies. And as much as I'd be delighted to prove out my ability to rush a new blog engine out the door, I don't necessarily want to rush a database schema, especially if I intend to someday share the schema and codebase with the world.
And I never said I was going to open-source this. I might, but I also want to commercialize it as a hosted service. I'll likely do both.
But it's coming, and here are my dreamy if possibly ridiculous plans for it:
- Blogging with comments and image attachments. Nothing special here. But I want to support using the old-skool MetaWeblog API, so that'll definitely be there, as well as the somewhat newer AtomPub protocol.
- Syndication with RSS and Atom. Again, nothing special here.
- As a blogging framework it will be a WebMatrix-ready web site (not web application). Even though it will use ASP.NET MVC it will be WebMatrix gallery-loadable and Notepad-customizable. The controllers/models will just be precompiled. Note that this is already working and proven out; the depth and detail of customizability (such as a good file management pattern for having multiple themes preinstalled) have not been sorted out yet, though.
- AppHarbor-deployable. AppHarbor is awesome! Everything I'm doing here is going to ultimately target AppHarbor. Right now the blog you're looking at is temporarily hosted on a private server, but I want that to end soon as this server is flaky.
- Down-scalable. I am prototyping this with SQL Server Compact Edition 4.0, with no stored procedures. Once the project begins to mature, I'll start supporting optimizations for upwards-scalable platforms like SQL Server with optimized stored procedures, etc., but for now flexibility for the little guy who's coming from WordPress to my little blog engine is the focus.
- Phase 1 goal: BlogEngine.net v18.104.22.168 approximate feature equivalence (minus prefab templates and extra features I don't use). BlogEngine.net is currently at v2.x now, and I haven't really even looked much at v2.x yet, but as of this blog post I'm currently still using v22.214.171.124 and rather than upgrade I just want to swap it out with something of my own that does roughly the same as what BlogEngine.net does. This includes commenting, categories, widgets, a solid blog editor, and strong themeability; I won't be creating a lot of prefab themes, but if I'm going to produce something of my own I want to expose at least the compiled parts of it to others to reuse, and I'm extremely picky about cleanliness of templates such that they can be easily updated and CSS swappages with minimal server-side code changes can go very far.
- Phase 2 goal: Tumblr approximate feature equivalence (minus prefab templates). All I mean by this is that blog posts won't just be blog posts, they'll be content item posts of various forms--blog posts, microblog posts, photo posts, video posts, etc. Still browsable sorted descending by date, but the content type is swappable. In my current implementation, a blog entry is just a custom content type, and blogs are declared in an isolated class library from the core content engine. I also want to support importing feeds from other sources, such as RSS feeds from Flickr or YouTube. Tumblr approximate equivalence also means being mobile-ready. Tumblr is a very smartphone-friendly service, and this is going to be a huge area of focus.
- Phase 3 goal: WordPress approximate equivalence (minus prefab templates). Yeah I know, to suggest WordPress equivalence after already baking in something of a BlogEngine.net and Tumblr functionality equivalence, this is sorta-kinda a step backwards on the content engine side. But it's a huge step forward in these areas:
- Elegance in administration / management .. the blogger has to live there, after all
- Configurability - WordPress has a lot of custom options, making it really a blog on steroids
- Modularity - rich support for plug-ins or "modules" so that whether or not many people use this thing, whoever does use it can take advantage of its extensibility
- Richer themeability - WordPress themes are far from CSS drops, they are practically engine replacements, but that is as much its beauty as it is its shortcoming. You can make it what you want, really, by swapping out the theme.
Non-goals include creating a full-on CMS. I have no interest in trying to build something that competes directly with Orchard, and frankly I think Orchard's real goals are already met with Umbraco which is a fantastic CMS. But Umbraco is nothing like WordPress, WP is really just a glorified blog engine. If anything, I want to compete a little bit with WordPress. And I do think I can compete with WordPress better than Orchard does; even though Orchard seems to be trying to do just that (compete with WordPress), its implementation goals are more in line with Umbraco and those goals are just not compatible because WordPress is a very focused kind of application with a very specific kind of content management.
And don't worry, I don't ever actually think I could ever literally compete with WordPress as if to produce something better. I for one strongly believe that it's completely okay to go and build yet another mousetrap, even if mine is of lesser ideals compared to the status quo. There's nothing wrong with doing that. People use what they want to use, and I don't like the LAMP stack nor PHP all that much, otherwise I'd readily embrace WordPress. Then again, I'd probably still create my own WordPress after embracing WordPress, perhaps just like I am going to create my own BlogEngine.net after embracing BlogEngine.net.