Why I moved my blog from Wordpress to Ghost, and why I turned off comments along the way

27 February 2014

In May last year I saw a tweet about a Kickstarter campaign for a new blogging platform called Ghost. I hadn't heard anything about Ghost up to that point, nor had I ever pledged money to a Kickstarter campaign. Nevertheless something about it intrigued me, and before I knew it I'd backed the project. The Ghost project is led by John O'Nolan and Hannah Wolfe, both of whom have been involved with Wordpress in the past.

Having worked with Wordpress (both the Wordpress.com hosted version, which I was using for this blog, and the Wordpress.org host-it-yourself version, used at In the Library with the Lead Pipe) I immediately understood what O'Nolan and Wolfe were trying to do. Wordpress started as a simple blogging program, but after many years of development it has picked up a lot of bloat and has turned into a full blown CMS. When you just want to blog and maybe have a couple of static pages, there's a lot of useless features cluttering it all up. Ghost is an attempt to build a back-to-basics, clean and efficient blogging system. It is, by design, 'just a blogging platform'.


Whilst I now know a little more about how Ghost is built (on the relatively new node.js, rather than the more mature PHP base of Wordpress), the thing that originally excited me was Ghost's Markdown editor. The Ghost team realised from the start that WYSIWYG editors are great in some situations, but lousy in others. Instead of a WYSIWYG editor, Ghost uses a split screen with Markdown input on the left, and an instant preview on the right. Markdown is fairly new to me, but was actually created nearly ten years ago by John Gruber. Like so many awesome things on the internet, Markdown's development also heavily involved Aaron Swartz. Markdown is designed to be easily read as plain text, but losslessly translatable to Rich Text or HTML. It can be thought of as the reverse of a markup language, hence the name.

Markdown has recently gained traction among some academics, and projects like Markx (with a similar split-screen editor) look promising for academic publishing. It has the potential to finally break the defacto academic formatting standard of "MS Word exported to PDF" and replace it with something flexible and web-native. Anyway, as you can see I'm pretty excited about Markdown. I have actually been writing my blog posts in Markdown (using iA Writer for about six months, then pasting them into Wordpress and manually editing them back to WYSIWYG. Now I can just dump them straight into the Ghost editor and hit 'Publish Now'. Ghost's Markdown implementation doesn't currently allow for footnotes, but given they are in the original Markdown spec I'm sure this is coming in future releases.


One of my goals for 2014 is to become more fluent in HTML and CSS, and to learn a bit of Javascript. Conveniently, Ghost runs on node.js, a fairly new server-side environment entirely based on Javascript, and to change anything in Ghost in terms of site design or functionality requires CSS and HTML knowledge, (or, failing that, really good Google skills). In this respect the fact that Ghost is so new and undeveloped is a good thing - it forces me to know how to do things by hand. By using Ghost when it's still only at version 0.4 I've been forced to learn more about how it actually all works in order to get my blog looking and working the way I want. Already I've discovered the bickering siblings SASS and LESS, and Handlebars - although admittedly I still don't have a firm grasp on how Handlebars really works. This is a great opportunity to be in at the start and participate in Ghost's evolution, whilst learning a bunch of stuff along the way.

All going well, I hope by the end of the year to reach the point of building one or more Ghost themes or maybe even a plugin. I'd love to be able to build something (however small) that's useful for scholarly publishing. Having a particular goal in mind rather than just 'I want to learn to code' is definitely going to help me to get there.

I'll explain some of what I've learned and how I've fitted it all together, in my next post.

Discussion, not comments

A significant change that comes with the new platform is that there are no longer comments (and old comments are gone as well). This was not an easy decision to make, and it deserves some explanation.

Ghost does not (and apparently will not) support comments natively. That is, there's no in-built comments functionality. This by itself is not a particular problem. There are a few comments tools around, notably Disqus and the intriguing but less well-known Moot. What Ghost's lack of commenting functionality did, however, is force me to think about how I wanted comments to work, which system to use, and what blog comments are for. When you have to actively decide to include a commenting function rather than whether to turn it off, if forces you to think about why we have them in the first place. The more I thought about it, the more convinced I became that in the context of my blog comments are actually counter-productive.

I've been interested in Matthew Reidsma's approach to comments ("do it on Twitter") for a little while. Thinking it through further, I decided to also abandon comments altogether. This has some key advantages:

  • it doesn't add bloat and sluggishness to the site like a comments widget would (Disqus is particularly notorious for this);
  • I have a reasonable number of readers via RSS rather than direct traffic, and most of them won't see the comments in their feed anyway;
  • it encourages more considered engagement with my arguments rather than just a reflex instant-reaction comments;
  • it potentially widens the conversation - if discussion of a post is held on Twitter and through dialogue between blogs, a lot more people have the opportunity to see and participate in the conversation than if it was simply sitting on my blog where not that many people interact with it;
  • I can use this environment as another tool to encourage other librarians to write publicly about the profession (i.e. start your own blog and respond that way, rather than just leaving a comment on mine).

These last two points are crucial and the main reason I decided not to set up comments. The post I've had more hits on than any others (by a long way) is A 3D printer will not save your library. The reason it's been read so many times is that a few people read it and then wrote responses on their own blogs. When one of those people is David Lankes, suddenly you've started a 'new debate' and a dozen more links and opinions are published. If Lankes had merely left a comment on my blog, that wouldn't have happened. The result is that librarians had a robust international discussion about 3D printing, the nature of information and the purpose of libraries, and I also got a lot more readers. Whilst the extra readers are nice for my ego, it's really the robust and broad discussions I'm after.

I'm nobody special, just someone who has committed to write and share his thoughts about his profession. I'm also not selling eyeballs to advertisers, so I have no particular reason to keep people coming to my site by ring-fencing the discussion. I'd prefer to have dialogue as equals, not a relationship of publisher-author and the peanut gallery throwing out comments. That's why my comments section now reads:

Discuss Use Twitter mentioning @hughrundle, or why not write a response on your own blog?

I'd love to read your thoughts on this approach - but not here. Here's three publishing platforms:

Now go and start a blog.