2023-09-12

Blue specks

12/09/2023

I didn’t paint again today. I have to certainly tomorrow, I don’t want to keep delaying.

Instead I decided to implement read only volumes into faasd. I dived head first into Go. Having never done anything with it before. Learned how to do everything I needed to this evening and have started the proposal in the faasd repo to implement the change. I’m very happy. I want this for cAdvisor, as I don’t think it’s a good idea to mount the whole filesystem onto cAdvisor. What if the container crashes? What if it sudo rm -rf? I don’t know how it works or what google are up to. All I know is they advise to use readonly volumes and now that is possible on faasd. Quite happy with myself to be honest.

I did that this evening instead of eating so it’s going to be a really quick and late dinner.

I think I’ll park the functions platform for a little bit, at least until I get a response on my issue for faasd. To summarise my progress so far I’ve learned how to set up faasd on hetzner (terraform in scope, not done), how to set up grafana, loki, and promtail to do all of my monitoring dashboards, I’ve learned how the inside of faasd works, and I’ve deciphered my previous workerd work, to add the next level of the platform. I think it’s quite good progress. I just really want to pick usedteslas back up again.

One thing that all the buzz has been about is Bun, as they released v1 this week. I was going to add a note about it yesterday but yesterdays entry was long enough. I need to give it a go, because I like the idea of a decent speed increase in my js projects. Astro actually doesn’t have the warning on the Bun website about node specific API’s so that’s promising. I also wonder if Bun will work form my workerd stack. That’ll be a next week exploration.

Outside of all that I’ve been considering functional design aesthetics and how this has influenced the way people talk about design, especially in the tech world. I have a big problem with this, that should be a blog post at some point. The crux of the issue is that there is a presumed dichotomy of form and function. Your design can be highly functional and fulfil all of it’s requirements to the best of its ability, or it can be cute and colourful, playful and interesting. There is a school of thought that seems to pursue the former, giving rise to minimalism, brutalism, and function, as a sought after aesthetic. It’s followers dismiss design for form’s sake as, even though it’s aesthetic qualities are at the forefront of it’s intention and purpose, they seem to prefer their design to seep through the gaps of function, presenting itself under the surface of something that does what it is supposed to do. They want beauty in design subtly rather than explicitly. They like it when their designers say “I just wanted to make the most useful and bestest tool for the job, it’s a coincidence that it looks great” so they all pretend that they only care about function and not the form. I blame Dieter Rams, who is the most commonly quoted designer in this school of thought, harshly too, I believe he didn’t intend for it to result in this. Dieter Rams is a truly talented artist at design. He put technology in the background so that the experience that the user could have was in the foreground. And that is exactly what all of those utilitarian function over form designers misunderstand. They read the 10 principles of good design and ignore the subtext of it all. The assumed variable. The thing that breaks the dichotomy. That emotion is the third factor in design. That the feeling of the user should be the first thing considered in the design process, and carefully balanced to produce form and function that creates the emotion in the user. We need websites to look beautiful because this puts the technology in the background, shielding users from the cognitive load in understanding how the technology works, and lets them be free to feel what you want them to feel. I don’t know if I start giving examples I’ll be here all night. I’ll leave it here for now.