Apparently I’ve got a thousand posts under my belt, so I thought it’d be fun to write down my current format. Sometime in the future I can look back on these notes and compare to see how it has evolved since.

Length: My target length has remained 300 words, but I’ve become a lot less stringent about it. 300 words is enough for a beginning, middle and end to a story. It is also about the right length to describe a problem, list the constraints, and explain why I made the decision I did. Sometimes I could get my thoughts out in 250 words, and that’s fine. When something goes long, I usually try to cut them into multiple ~300 word installments, but sometimes splitting up doesn’t make sense. I forced it a few times and they read poorly in hindsight, so if I run into it again (like this post) I just let those pieces run long.

Always Have A Featured Image: When I started writing I paid little attention to images, because the original focus is to have a written record I can search through. As it turned out, the featured image is really useful. First: it allows me to quickly skim through a set of posts just by their thumbnails, faster than reading each of their titles. Second: making sure I have at least one picture attached to every story is very helpful for jogging old memories. And sometimes, what I thought was a simple throwaway image became a useful wiring reference. I now believe pictures are a valuable part of documenting. Today’s cell phone cameras are so much better than they were four years ago, it only takes a few seconds to snap a high quality picture.

Still figuring out video: While images may have been an afterthought, video was not a thought at all when I started. Right now I’m in the middle of exploring video as an supplement — not a replacement — for these written records. It is another tool to use when appropriate, and cell phone camera improvements helps on this front as well. The only hiccup today is that I can’t directly embed video because VideoPress is only available to higher WordPress subscription tiers. As workarounds, short video clips are tweeted then embedded, and longer video clips are uploaded to YouTube and embedded. I expect video usage evolve rapidly as I experiment and see what works.

Use more tags, fewer categories: I started out trying to organize posts in categories, and that has become an unsatisfying mess representing a lot of wasted effort. When I want to find something I wrote, I go for the straight text search instead of browsing categories. And if I want to relate posts to each other in a search, I can use tags. It has advantage of arbitrary relations free of constraints imposed by a tree hierarchy.

Yet to stay with consistent voice: This is my blog about my own work, so I usually say “I”. But sometimes I slip into talking about “we” because in my mind I’m talking to my future self.

Keep up the daily rhythm: Scheduling a post to go out once a day, every day, is the best way I’ve had to keep the momentum going. I tried going to slower rhythms, like every other day, and it never works. If I stop for a single day, I’m liable to stop for multiple days that drag to weeks without a post. Usually there’s a good reason like a paid project that is consuming my time, but sometimes there isn’t. I’ve learned it is very easy to lose my momentum.

If it was interesting enough to take time, it’s interesting enough to write: I now describe tasks that took time, multiple searches, and multiple tries, before I found the solution. My original reasoning for not writing them down is the that since I found all the information online, my blog post won’t have anything new that people can’t find themselves. But there have been a few episodes where I forgot the solution and had to repeat the process again, and I was unhappy I didn’t write it down earlier. I’ve learned my lesson. Now if it took a nontrivial amount of time, I’ll at least jot down a few details in my “Drafts” folder for expanding to a full blog post later. Some of these are still sitting as a draft, but at least in that state they are still searchable.

I just learned WordPress puts up a special milestone notification when a blog site has one thousand posts, because I triggered that notification with yesterday’s post about vaguely attainable somewhat humanoid robots.

It’s pretty common for a personal blog to have only a handful of posts — sometimes just one — before it goes dormant. My first attempt ended after less than a dozen. The second attempt had more than a dozen, but not by much. Fortunately for me, they have stopped taunting me as they have been erased by no actions of my own: both of them were hosted on small startup blog hosting services that have since gone out of business. Maybe fragments have survived in Google caches and what not, but I haven’t felt inclined to go searching for them.

I had no reason to expect the results would be any different with this third attempt, so again I started with the free tier of service. Except this time I started with a more established host: WordPress.com, the commercial hosting counterpart whose revenue helps support the free open-source blog software available from WordPress.org. When I felt that I’ve found my groove and can keep this going, I upgraded to the “Personal” plan so I can have my own domain and remove WordPress ads.

So far I have felt no need to upgrade beyond the Personal tier. Most of the higher tier features are tailored to people trying to make money in one way or another but I have no revenue goals for this blog. This is mostly documentation for my own aims, and if my notes are useful for someone else, that’s just a happy coincidence. One way I’ve described this site to friends is “a diary with zero expectation of privacy”. My content is not tailored to maximize traffic and, in fact, is the wrong medium to do so: consumer traffic (and corresponding ad revenue) are migrating towards video and away from text.

But I want text. I like to read and learn at my own pace. While I’m glad YouTube (and other video sites) have implemented ability to adjust playback speed of a video, having to go and change that setting is still a hassle. And finally: as documentation for myself, I want to be able to search through my notes and that’s a lot easier with text than video.

But there are some things more suited to a video than the written word, and for them I’ve shot video footage and created a New Screwdriver YouTube channel to host them. Right now I see the YouTube channel as roughly analogous to my first few aborted blogging efforts: an exploration into the medium looking for a way to make this work. Hopefully it won’t go dormant, but the YouTube channel certainly won’t be my focus for the foreseeable future.

One thousand posts is a good milestone, and I intend to keep things going. But as things will continue to evolve and change, it’s a good time to write down the current state for future comparison.