I've been trying to write something every day, but last night I was out late and fell asleep as soon as I got home. My first instinct is to make up for it by just posting twice today, but the problem with that is that it lessens the instructive value of having failed. I think it's important to be reminded occasionally that success isn't something you achieve once, it's a process that requires constant maintenance. Suceeding at writing every day, then, is something that can go away fairly quickly.
But I do gain a lot of value from feeling like I am trying to keep up a particular pattern, and it feels good to have a long chain of daily posts stretching back. The most important thing is that the habit has significance, and that it's something I take seriously. Eventually I arrived at the compromise decision of posting twice, but making one of those posts about the failure itself.
I think the strategy that would have prevented this is writing before I left home, or alternatively being prepared with something I could write quickly as soon as I got home regardless of how late that was. Doing those things obviously won't help this failure, but it should make one less likely next time.
I recently realised how nice Catenary's source looks on Github. I think it's mostly a coincidence caused by Literate Coffeescript happening to be a subset of Markdown and Github having really nice Markdown rendering.
Honestly looking at it kinda makes me want to write everything using this particular Markdown-as-literate-programming style, even non-Coffeescript projects. It'd be lovely if there was a general tool that let me use any language embedded in Markdown and extract it out in a build step. Though I suppose part of the difficulty is that language would need to be flexible enough that it doesn't limit your ability to write your thoughts in whatever order makes sense.
I know Knuth's original literate programming had a whole metalanguage that let you break completely away from the conventions of the language but in the immortal words of Sweet Brown: ain't nobody got time for that. It feels much cleaner to have something that still looks (mostly) like the original language and keeps its structure.
This is a cover I made of Pain Fade Down for a presentation today at SydJS. Pain Fade Down is a song from the Darwinia soundtrack, and you can find the original on trash80's website.
It turned out to be surprisingly tricky to recreate a lot of the sounds! The original used a text-to-speech engine for the vocals which I tried to recreate with vocoding to questionable success. Also the bleepy bloopy synth was really tricky to match tonally because of the delay effects. If you're reading this and are a musician, I'm happy to take suggestions!
It's built using my Tabletone library, which you can also find on Github.
I've been trying to get a bit more of a handle on my finances lately, so I figured I'd try to get all the data into the time-series database I've been messing with. Actually getting data out of the various banks turned out to be a nightmare of badly formatted pseudo-CSV and export systems that give you slightly less than 6 months of data at a time.
Still, I've got it going now. I've written scripts to clean the CSV up and other scripts to load it into my database. I've got scripts up to my ears. And it kind of occurs to me that I still don't necessarily have all the information I need. The pictured graph is how much I spend each week. Ideally I'd be able to break that down by what kind of things I'm spending it on and filter things out and so on, but that's still a ways off.
I'm pretty surprised that none of the banks seem to be working on analytical tools like this given how useful they are. Plus they'd easily be able to beat any third-party offerings because the crappiness or straight up nonexistence of their APIs and export tools would make competition impossible.
In certain sciences like astrophysics it's often impossible to measure things directly. You can't really fly out with a giant galactic tape measure to figure out the size of the Milky Way, instead you measure stars that pulsate in proportion to their energy and use that energy and how bright they appear to figure out how far the light must have travelled. Similarly, you can't measure the mass of a black hole, you rely on the effect its mass has on nearby bodies. In fact, you basically can't detect black holes at all except by their effects on things around them.
It occurs to me that sometimes people have the same property. You can't really measure someone's intentions, personality or abilities. Instead, you have to measure the indirect effects those qualities have on the world around them. There's no perceivable difference between an amazingly intelligent person who never says anything and an amazingly dumb person who never says anything. Someone who would do amazing things if they ever had the chance but never gets that chance has still achieved exactly as much as someone who never tried at all.
A possible (depressing) conclusion to that line of thinking is that you don't exist except as measured by the opinions of others and the results you achieve. But those measurements are so unreliable. In poker they talk about a bad beat: a hand where you do everything right but lose anyway, or vice versa. How can you possibly be happy with your decisions if all you care about is the win? And how can you even have an identity if it's just what people think you are? There is, however, one (and only one) person your internal state does make a difference to: yourself.
I think to be able to function effectively in the face of adversity or indifference you have to be able to measure yourself on your own terms. If you can define without reference to the outside world the shape of good and bad, of success and failure, then you're also free from trying to control those external factors which are, fundamentally, beyond your control.
Measuring yourself by your effects on the things around you may be simple, even easy, but the end result is being dependent on those things for your happiness, and even your identity itself.
