Week 47

Going to finish the week on Friday instead of Sunday and write these on the weekend, it’s way easier to find the time to do it then and leave Monday for getting organised for the week ahead. So, a short one.

Wrkbk

Finished off the schema generation code — it’s ‘good enough’ now, pretty modular, and almost understandable.

I’m using Typescript for the core libs, and it’s interesting how it changes the thought processes when developing. I find myself doing a lot of thinking up front to work out the types I need and how they interact, which takes a while, but the upside is that code is much easier to write once you’ve worked that all out. If it’s not easy that’s usually a sign you’ve got the types wrong!

There’s obviously an overhead with static typing, but it’s idiotic to pit statically typed languages against dynamically typed ones, it totally depends on what your use case is. I’m not in any hurry to change the frontend code to Typescript, but with the core libs the benefit far outweighs the cost.

RetroPie

Messing around with RetroPie on a Raspberry Pi B+ at the moment, it’s amazing how good the whole Pi ecosystem these days. Will definitely talk more about this later, trying not to fall too far down the arcade emulation/hardware rabbit hole, it is a particularly deep one.

Facebook is the new Yahoo@

Generally good article in The Economist about Facebook, they’re obviously at a fork in the road and they don’t seem to be going to The Good Place but it compares them to Yahoo! which I think is a wrong. When I worked there the fundamental problem was that Yahoo! didn’t know what it wanted to be. Every year it changed focus — Portal! Search Engine! Aggregator! Services! and on and on, changing direction every year, never having the nerve to really bet big on one thing. While they were vacillating the employees got bored and wandered off to other companies, the same companies that proceeded to leave Yahoo! in the dust.

Facebook don’t have that probem. They know exactly what they want to be, and have total focus on making it happen. Their real problem is whether the coming generation of savvy millennials want the service they’re selling.

Links

• A great overview of the design of Incredibles 2 from one of the designers, always interesting to see the amount of craft that goes into Pixar’s films.

In Progress

• Justice: What’s the Right Thing To Do — this week Sandel got deep into Kant and I think I understood it for the first time! Definitely a mark of how good Sandel is at explaining complex ideas. Really loving this book.
• The Good Place is so good, Season 3 haters can suck it.