Moving the blog to Jekyll by Mark Seemann
ploeh blog is sorta moving
For more than four years ploeh blog has been driven by dasBlog, since I originally had to move away from my original MSDN blog. At the time, dasBlog seemed like a good choice for a variety of reasons, but not too long after, development seemed to stop.
It's my hope that, as a reader, you haven't been too inconvenienced about that choice of blogging engine, but as the owner and author, I've grown increasingly frustrated with dasBlog. Yesterday, ploeh blog moved to a new platform and a new host.
What's new? #
The new platform for ploeh blog is Jekyll-Bootstrap hosted on GitHub pages. If you don't know what Jekyll-Bootstrap is, it's a parsing engine that creates a complete static web site from templates and data. This means that when you read this on the internet, you are reading a static HTML page served directly from a file. There's no application server involved.
Since ploeh blog is now a static web site, I had to make a few shortcuts here and there. One example is that I can't have any forms on the web site, so I can't have a contact me form - but I didn't have that before either, so that's not really a problem. There are still plenty of ways in which you can contact me.
Another example is that the blog can't have a server-driven comment engine. As an alternative, Jekyll provides several options for including a third-party commenting service, such as Disqus, Livefyre, IntenseDebate, etc. I could have done that too, but have chosen to try another alternative.
There are several reasons why I prefer to avoid a third-party solution. First of all, in case you haven't noticed, I'm into this whole blogging thing for the long haul. I've been steadily blogging since January 2006, and have no plans to stop. Somehow, I expect to keep blogging for a longer period than I expect the average third-party content provider to last - or to keep their service free.
Second, I think it's more proper to keep the comments with ploeh blog. These comments were originally submitted by their authors to ploeh blog, which implies that they trusted me to treat them with respect. If I somehow were to move all those comments to a third party, what would become of the copyright?
Third, while it wasn't entirely trivial to merge the comments into each post, it looked like the simplest solution, compared to have to somehow migrate comments for some 200+ posts to a third party using an unknown (to me) API.
What this all means is that I've decided to simply keep the comments as part of the body of the post. Comments, too, is static content. This also means that comments are included in syndication feeds - a side effect that I'm not entirely happy about, but for which I don't yet have a good solution.
So, if comments are all static, does it mean that I no longer accept comments? No, I still accept comments. Just send me a pull request. Snarky, perhaps, but the advantage of that is that if you want, you'll have a lot more flexibility if you want to comment: you could include code, images, tables, lists, etc. The only checkpoint you'll have to get past is me :)
Actually, you are welcome to send me pull requests for other enchancements as well, if you'd like. Did I spell a word incorrectly? Send me a pull request. Is a link broken? Send me a pull request. Do you know way more about web site design than I do and want to help? Send me a pull request :) Seriously, as you may have noticed, ploeh blog is currently using the default Twitter Bootstrap design. That's not particularly interesting to look at, but I hope that the most important part of the blog is the content, so I didn't use much energy on changing the look - it probably wouldn't have been to the better anyway.
Since the content is the most important part of the blog, I've also gone to some lentgth to ensure that old permalinks to posts still work. If you encounter a link that no longer works, please let me know (or send me a pull request).
Finally, as all RSS subscribers have probably noticed by now, switching the blog has caused some RSS hiccups. While the old feed address still works, the permalink URL scheme has changed, because I didn't think it made sense to have static files and URLs ending with .aspx.
What was wrong with dasBlog? #
Originally, I chose dasBlog as a blogging engine because it had some attractive features. However, the main problem with it is that it's no longer being maintained. There seems to be no roadmap for it.
Second, the commenting feature didn't work well. Many readers had trouble submitting comments, particularly if they wanted to include links or other HTML formatting. I also got a lot of comment spam.
Finally, the authoring experience slowly degraded. The built-in editor only worked in Firefox, and while I don't mind having Firefox installed, it was a cause of concern that I could only edit with one out of three installed browsers. What if it also stopped working in Firefox?
It was definitely time to move on, but whereto?
Why Jekyll? #
Before I decided to go with Jekyll, I looked at various options for blogging sofware or services, including WordPress.com, Blogger, FunnelWeb, Orchard, and others. However, none of them could meet my requirements:
- Simplicity. Everything that required a relational database to host a blog was immediately dismissed. This is 2013. A RDBMS for a single user blog site? Seriously?
- Easy hosting. Requiring a RDBMS significantly constrains my deployment options. It also tends to make hosting much more expensive. Since January 2009 I've been hosting ploeh blog on my own server in my home, but I've become increasingly annoyed at having to operate my own server. This is, after all, 2013.
- Backwards compatible. The new blog should be able to handle the old blog's permalinks. Redirects were OK, but I couldn't just leave hundreds of external links to ploeh blog with a 404. That pretty much ruled out all hosted services.
- Forward compatible. This is the second time I change blog platform, and it's not going to be the last. I need a solution that enables me to move forward, taking all the existing content with me to a new service or platform.
I seriously toyed with writing my own blogging engine to meet all those requirements, but soon realized that Jekyll was a pretty good match. It's not perfect, but good enough for now.
While I like the move of the blog post to GitHub, the overhead of creating comments here is just too high. I like having some level of entry barrier to keep the quality up but this is more to the point of completely detracting additional useful information from being posted in comments.
An option that comes to mind could be possibly using the Issue tracking system in GitHub for comments or something that gives access to the GitHub markdown features like the pages themselves.