I started this blog on the 7th of April 2014, exactly two years ago. Exactly one year ago I did a retrospective post, so I figured it was about time for another. In that post, I mentioned how part of the aim of this blog has been to avoid self-indulgence – while I obviously write about things that I like and am interested in, using a style that’s enjoyable for me, I also try to make each article have some sort of external point and avoid writing about completely mundane and self-centred topics (no-one really wants to read your diary on the internet).
These posts have come to be my one exception to that.
Last time I made a list of my most popular posts. Because a foolish consistency is the hobgoblin of small minds, I’m not doing that again this year.1 Instead, I’m just going to air some general thoughts on this blog and how it’s evolved over time.
Most months, I post two articles. There’s no set schedule for when these articles are posted, and I make no effort to spread them out or bunch them together – some months I post each one two weeks apart, other months I post them two days apart. Sometimes I only end up posting one, and very occasionally I’ll go without posting anything for a whole month, but two articles per calendar month has become the schedule I try to keep to and usually do.
As things have gone on, my posts have got longer on average. I’d like to think they’ve improved qualitatively in tandem with that: ultimately you need to have a certain number of words in order to really discuss anything, and I believe a certain level of prolixity is required for things to be easily readable: no-one communicates in regular expressions.
But there’s a point of diminishing returns, and if there’s an unofficial standard where, say, all posts are at least 1.5k words long, a few things start to happen:
The scope of content grows in ambition. As an example, I’ve got a long post comparing a bunch of systems for making text games (like my last two). It details about five different systems, going into very varied degrees of depth on each, and leaving out a bunch more.
Schedules start to slip. The post I mentioned above has been a work in progress for around a year now. Probably I’ll be very happy with it when it’s finally done, but it’s just as likely to into an all-consuming monster of textual overindulgence and grow ever longer and more gnarled without ever actually getting done.
It’s difficult to post stuff that hasn’t been in drafts for at least two months. Everything you’ve ever read online is woefully reductive and over-simplified to the point of incorrectness. For proof, just take a look at a few comments sections on Hacker News. And so as a writer, it’s natural to want to not be simplistic or reductive: to have all the facts, to include all the caveats and acknowledge all the positions. But thinking like that makes it very difficult to write an article in a sub-geological timeframe. On some level, that’s good: the internet hardly needs more ill-informed “informative” thinkpieces and one-sided diatribes. But if you always have to get all the facts, all the perspectives and a pre-emptive reply for every possible response and nitpick before you can call a piece complete, you’ll never finish anything.
So basically I’m going to be trying to write more short, light pieces, and more split-up serieses like the text editors one, because I enjoy writing blog posts and need to reverse a trend I see in myself before my drafts folder gets too much bigger. Even so, don’t expect the massive wall-of-text posts to go away – that monster post I referred to earlier isn’t the only one of its kind.2
In the last year, I’ve extended the Ghost blogging system this site runs on in a couple of ways and I’ve gone from using an off-the-shelf theme to my own painstakingly handwritten HTML and CSS. I’ve also added a search feature after some requests (that magnifying glass in the top right).
Over the next year, here are some features I’m considering implementing:
Comments: Though this is less and less fashionable with every passing year, I’m of the opinion that blog comments can be a real value-add (and a good way of outsourcing article fact-checking to internet pedants). I’ve already written most of a commenting system similar to the one I described here, to the point where most of the effort in rolling it out to this site is deciding whether or not I actually want to.
Blogging web application → static site generator: This is more fashionable and modern that comments, but it’s not just about being cool. From everyone’s perspective but mine, this site is static: you go to a page, you read some text. That text isn’t going to change according to what you do on the website – you don’t have a way to log in as a reader, and the only input field is the Google-handled search. So it’s a bit wasteful that whenever you request a page, Ghost takes a bunch of templates, makes database calls to populate them and generally does a whole lot of dynamic processing for an ultimately static result.
From your perspective, the results would be the same if my blog just served you a
.html page containing exactly what you see in your browser’s
View source. So why not do that? It wouldn’t be a noticeably faster, but it would still be faster and it would be simpler and it would cut out a whole class of security vulnerabilities.
From my perspective, I wouldn’t have an admin panel anymore, which could make it a bit more of a pain to edit drafts on different PCs… until I set up a few simple things with
git to achieve much the same effect with a number of benefits I don’t have now, such as seamless local editing – the Ghost blog editor is fantastic, but nothing beats the comfort of your favourite text editor.
So I might just go for that. If I do, you should ideally not notice.
And thus ends year two of my vanity blogging project.
- If you must know, this (from last year’s list) and this (which just missed last year’s list) are my most popular posts by a mile, to the degree that places three through five don’t even matter. ↩
- “Text editors (III)” is coming up, and it’s going to be about Emacs. There’s a lot to say about Emacs. ↩