Totally stole the idea for the title from Better Software Conference. They are live as I'm writing this and I've just watched a talk by Casey Muratori. It was fucking awesome. Around the same time I've come across his post on Semantic Compression which I'm yet to read but quite excited to. Like a movie you've been waiting to watch. That's how exciting I find programming, any opportunity to steer my journey through it in the mastery direction I will take it. The talk was about the history of OOP and the direction it could have taken decades earlier. It has its merits but not to the extent that it's peddled by most. Luckily for me all that thinking ahead and planning was too much for my lazy self, so I've somehow managed to dance around it. Skipping it mostly. The only time I recall was at one of my previous workplaces I had to write the same feature in two separate codebases one after the other. One of them I was done in about four days and the second one took about a week. The first one only took four days as I had no idea what I was doing. The second was the OOP one. Although I'll admit apart from a lot of abstraction I don't know if it was better or worse. That's the extent of my experience with OOP.
I'm looking forward to reading the paper on Sketchpad tomorrow. To be honest I'm not sure I can build that software myself, so I'm half tempted to give it a try. Sounds like fun. It has been a great week for me in terms of finding programming gold mines. Last week I read this paper on Scalability. What I'm mostly finding out is that just try the stupid sounding simpleton idea, measure how it performs and let that inform your decisions about what to do next. Keep it simple stupid. So this week at work I had set up caching for some of our APIs so I decided to get a benchmark first, see which APIs needed the most attention, understand why and then make any "improvements". I'll admit I'm not very good at writing good software but I'm trying. I'm hoping one day I will. My biggest struggle is building things on my own with minimal help. Exploring a solution to the problem that is not very well documented. I know I can do it. This is my main point of focus now. The compiler is almost done thanks to a lot of hand holding from The Compiler Book. I'm glad I did it though. After a while I stopped getting any help with the code, I just read the concepts and then built it. It is definitely way more fun to figure things out as you go. After I would go back and look at the Author's implementation and keep the one I preferred. To my surprise I'm happy with most of my decisions. The only way to build confidence is I suppose by walking the path. There is no shortcut. I'm falling asleep so not going to be proofreading this.