TomBombadildoze

20 years working on Python. Two things have driven its success.

First, it's easy to learn and reasonably powerful. You can make a lot of useful things really quickly, which was true even 15 years ago. It's only gotten easier since, and the ecosystem has continued to grow.

Second, ML latched onto it. That gave Python a lot of credibility for people coming through CS programs and into industry, especially as AI use cases were picking up momentum. Think Tensorflow starting 10-ish years ago, then Pytorch, and so on.

[opinion time] Unfortunately, ML has contributed to disaster. The quality of software engineering in ML applications lags badly behind that in other segments, and there's where most Python development is happening now. The feedback loop of training Bad Python into LLMs is only making it worse. The skill floor was dropping, and now it's practically zero.

It also wasn't helped much by language design decisions. Gradual typing was a huge mistake. It would have been much better to go the Typescript route and define a superset of the language that compiles to native Python. The grafted-in typing features that rely on runtime behavior and never got an actual spec for a type system haven't helped.

Beats Opus 4.6 at what exactly? It certainly isn't code.

I use a combination of a Claude Max subscription and local inference, including qwen3.8-27b, 4bit. I have found qwen to be absolutely useless at anything but very specific, surgical code changes. In my experience, for anything even remotely nuanced, a frontier model is required.

Unpopular because it's so flagrantly ignorant. How much do you know about the "egregious" expenditures by rank-and-file federal employees just doing their jobs?

> A butler has always been a person of authority, expertise, and responsibility. Why "houseboy"? Because he's Asian?

facepalm

The author describes Marc Andreessen's Gilded Age slash Roaring 20s lifestyle and attitudes, and then suggests that a hundred years ago, this person may have been referred to as "houseboy".

Marc Andreessen is the hypothetical person from a hundred years ago. It's a scathing criticism of Marc Andreessen.

Cartels disrupting law enforcement, and/or vice versa.

Pulumi, cdk8s.

Until you need to update it, or apply it to multiple environments, or any number of tasks where the concepts of configuration parameters and dependencies become important.

Edit: I'm not saying Ansible is an appropriate substitute (it absolutely isn't), rather that `kubectl -f` is not a scalable solution.

The lesson is that shockingly few people have realized is that the problem space is not, nor has it ever been, served by declarative solutions. You can describe the state you want a system to be in but there will inevitably be procedures involved to do some of the work. It is unavoidable. At the point people began developing templating around their configuration, ah, templates... which themselves have imperative language constructs awkwardly (and badly, cough Golang) embedded in them, it should have been clear to everyone with a shred of talent that it wasn't a good solution.

Declarative configuration is wholly expressible by imperative languages. The converse is not true. Declarative configuration is appropriate for declaring state (i.e., values), and not for describing procedures.

Stop using half-baked YAML crap and start using real tools.

When all you have is a hammer, everything looks like a nail.

"Moving on" and recounting a story can be, and often are, entirely different things.

Ruby/RoR values form over function. Over the past few (five or so?) years, as teams went through the transition from aggressive feature development to devoting more effort to maintenance and operations, they realized that relying on conventions as a guiding principle has a substantial cost burden that you don't pay when prefer rigor and specification.