I should remove the
Weekly moniker of these posts and emails. They are done when they are done. Enjoy!
You can also get these as a weekly newsletter by subscribing
Photo by Pushpak Dsilva on Unsplash
I love the idea of ADRs, but I still haven’t managed to convince teammates to use them.
Well, you don’t need
that many language extensions for most work. Just stick with
GeneralisedNewtypeDeriving , and
Don’t shoot your grads. It’s bad for the environment.
I have been following Cedric’s writing for a while. It was a somewhat interesting interview.
Luckily for Go developers, Phil could convince others that interfaces are enough to have generiticity.
A classic on
start small and get it working.
You can find my full review
I had managed to configure
ghcide on VS Code, but only if the version in Stack was exactly correct. This new VS Code plugin manages to figure it out (well, and add a lot of downloads to
.stack) and install always the correct version. Totally recommended.
Lego, what else do I need to say?
This is heart-warming to see.
Crystallising some things you may also see.
I didn’t fully get Applicative in
Finding Success (and Failure) in Haskell, this helped.
Short. Most can be found in other books by Drucker. Still useful.
I don’t have a backyard, sadly.
A “classic” John de Goes talk. A very thorough refactoring from buggy imperative code to bug-free functional (ZIO style) code.
Cedric’s blog is a trove of good content. I have been thinking of career moats (without having the concept itself) for the past 3 or 4 years (I tried to fomrulate some of it when I analysed
future tech trends here).
Another one from Cedric. Interesting idea: can we apply Kelly’s Criterion to
time allocation for our pursuits?
If you are part of
any online community (not only the ones I’m part of) you can probably think of someone like this.
Can’t argue with these. Although we might as well use Scala 3, which will have significant whitespace and has all of these already.
Singletons are tricky in multi-processing code. Async python is not
exactly multiprocessing, but things can happen at the same time so handling this issue is required anyway.
The demo scene is alive and well, and you can find the most approachable version in the JS1K competitions each year. Most contestants write their how-to, and this one is super interesting.
While the ideas are interesting (I already used some, and will write some tools for others for fun), the packaging (i.e. the story and the book itself) are not that great. I would still recommend it for the ideas on it, though.
Book of the year for me so far. Expect a long review some day.
The tool looks pretty useful (could complement a Wardley map) but this could be summarised in a moderately long blog post. Not even very long.
It was a dark and stormy web…
It’s a quick read, of not much use.
I want to do this.
This is pretty impactful. Must read if you code for a living or manage coders for a living.
There is an awesome underlying idea here that I think extends further than “product”. I need to think more about it.
This is some really obscure post. I miss Simcity 2000, that was
It’s always good seeing tips to optimise Haskell even if I don’t write Haskell (yet?).
An easy read on what topos are by John Baez. You need some knowledge of category theory, but not a lot.
Caveats and stoppers from GPT-3 and its (real) impact.
It’s not rocket science after all: people usually plan for harder things than a roadmap. Follow the guidelines here!
This is immediately understandable, great explanation of this family of algorithms.
Seriously, just write a Pandas backend for Scala 3.
I haven’t run ARMA or ARIMAA in… ages (last time may have been in R, and if you know me, you know how long ago that was), but they are a good tool in the toolbelt and this looks like the definitive McGyver in your pocket.
It timed out the first time I tried to open, so maybe not yet.
I would have expected the perfume to go rancid even with a perfect seal (due to oxidation). Now I can’t help wonder how it smells, the description is not enough!
Buy me a coffee