Category: til
-
Some notes from reading about the UK’s plan for clean power by 2030
For me at least, one of the most politically exciting things that happened this year was seeing the UK finally get rid of a conservative government and elect one that had to extremely ambitious climate promises in its manifesto. More specifically the goal to reach clean power by 2030 really caught my attention, and coincidentally…
-
TIL: Using consul-template to test local rendering of template files before using them in Nomad jobs
I’ve been using Nomad for a while, as an alternative scheduler to Kubernetes. It’s good, and I generally think I’m in good company when using it. I learned a handy trick with consul-template today and I think it’s useful enough to share here for others, and my future self.
-
TIL: Searching shell history in Atuin
I’ve been using Atuin recently to manage shell history. I really like it though it does take a bit of getting used to when it takes over the up arrow key in a terminal by default. Fortunately you can switch the default to make the up arrow filter commands to ones run in the current…
-
Why talk about monopoly in the context of climate change?
This came up in a recent conversation about sustainability in the digital sector, because the answer isn’t immediately obvious. To understand it, I think you need to understand why we have antitrust or competition laws in the first place. When they were first brought about bout in the first half of the 20th century, there…
-
Does the EU AI Act really call for tracking inference as well as training in AI models?
I’m sharing this post as it I think it helped me realise something I hadn’t appreciated til today. I don’t build AI models, and to be honest, while I make sparing use of Github Co-pilot and Perplexity, I’m definitely not a power user. My interest in them is more linked to my day job, and…
-
TIL: training small models can be more energy intensive than training large models
As I end up reading more around AI, I came across this snippet from a recent post by Sayah Kapor, which initially felt really counter intuitive: Paradoxically, smaller models require more training to reach the same level of performance. So the downward pressure on model size is putting upward pressure on training compute. In effect,…
-
TIL: Keeping your hair when upgrading Django 3.2 behind a Caddy server
This is a quick note to my future self, on how I recently got Caddy and Django to play nicely after an upgrade from Django 3.2 to a something newer, like 5.0. The problem After updating to Django 4.0, every request that sent data like a POST (the kind you use to login) was no…
-
TIL: buildpacks can rebase to save needless compute work
CNCF buildpacks have a nice trick that if you work with dockerfiles a lot you might appreciate.
-
Comparing what is spent on share buybacks vs the cost of decarbonising the grid
When doing some research this week, I came across what I think is a pretty eye opening stat. Tech firms spent twice as much money buying up their own shares in the last 18 months, as it would cost to buy out and shutdown every coal fired coal plant on earth, and replace them with…
-
TIL: Using Caddy with Django apps instead of Nginx
As mentioned before, I’ve been doing a bunch of work recently with Caddy, and a number of web apps. These are my notes for using Caddy, and why I think I prefer using it over nginx from now on. I also share a sample Caddyfile config for showing how to use it instead of nginx.