devRant - A fun community for developers to connect over code, tech & life as a programmer

Why would anyone defend the British government right now? The UK has turned into an authoritarian nightmare. A refugee can rape a kid and get 8 months in prison while a native British writes one social media post and gets 2+ years!

The UK government wants civil unrest. They want people upset. They want people to riot, so they can enact a crackdown and further draconian measures.

America has a lot of problems, but we have the freedom to speak and think here. Neither of those rights exist in the UK.

Stay vigilant. There are tons of bullshit "protect children" bills in the US right now. They want to destroy the small forum (like devrant) and only allow communication over large state-sponsored (DARPA and Peter Thiel funded/created Facebook; look up Digital LifeLog) where your government identity is tied to all your online actions.

I use Traefik, but yea there's Caddy and there are even Lua scripts for nginx that handle LetsEncrypt.

It looks like devrant is using LetsEncrypt, so whatever renews the cert must have just thrown up. I've had a misconfiguration Traefik that couldn't get certs on one of my sites for a while.

> grant me (the owner of the project) a permanent, unrestricted license to use, modify, distribute and relicense your contribution under any terms I see fit.

This feels like it defeats the entire purpose of open source. This is why I hate CLAs in general. I could never contribute to a project that has one of these.

You hold a normal, balanced, middle of the road (slightly libertarian) opinion.

People aren't allowed to have normal middle of the road opinions. The psychological operation that is "the news" is designed to make everyone cheer for one political sports team and disown everyone who likes the other team.

The most radical view to take in America is to reject both political parties.

You're fired and the AI that generated your slop is fired too.

Honestly most developers should use the weighted random word generation machines .. not for anything except small isolated tasks they can read over, or things that need to be written once and thrown away/not maintained.

I thought it was 2 because you can't start with a number, and the $ makes it not start with a number. I forgot about anonymous inner-classes as mentioned by others. This just creates a variable named "$50_00," right? Nothing special about it?

@retoor It's a fork of Gitea by former developers after Gitea started to offer commercial services

I dunno. I think we should jump between pages. I miss jumping to page 50 on search engine results just to see what was down there. Now you literally can't make it past page 10. Duck, Google and Bing just stop. There's nothing else.

You can have pagination and not having n+1 queries. We did it at my last company. Even when you search, you might need to go through pages 1-3 a lot of times.

I typically refuse to do any "homework" assignments.

I haven't used my Github in years, but my Gitlab is pretty active: https://gitlab.com/djsumdog .. although I really want to move off of Gitlab to self-hosted Forgejo at some point.

The vast major of leaders in the EU and US likely have some form of sexual blackmail on them. I'd venture to guess that less than 1/3 of the US congress hasn't been compromised (on both sides of the fake left-right spectrum).

The type of psychos that aspire to those leadership positions are also the types that take no issue using anyone, including children, for their gratifications. It's likely also a part of a lot of ritual to filter out people to those insane echelons of power.

To quote George Carlin, "It's a big club, and you ain't in it"

macOS went to shit after 10.6. That was the last great release. Everything Lion and beyond has been total shit.

We've been in a decaying trend for over a decade. It's hard to see the collapse of great empires when you living though them.

That's like ... a year from now? 70% is a pretty big jump ... and nobody likes developing in PHP. Seems like two clear steps up.

Lambdas don't really make any sense. They're non-trivial to deploy, difficult to debug and just annoy the hell out of me, and yet they're required to do AWS things like autoscaling. I had to write some to scale up/down Jenkins worker instances at an old job.

Microservices are their own beast. The better shops I've worked at only had a few services (<12 for our core team) and you'd work on a different one every few sprints to lean the entire domain. I'm currently at a shop where they build tiny handlers and microservices for everything. Our manager tried to get one guy to built out 10 microservices. We finally convinced him to let the guy do it in one Monolith "for now" that we can "split apart later."

There's a bad "skeleton" anti-pattern I've wanted to write about; tons of empty directories and TODOs and very small modules. I did write a post about five years ago on Microservices and biological systems:

https://battlepenguin.com/tech/...

At my last company, we didn't have devs accessing production. For support tickets, we had a Slack bot written where you'd apply to it and it would grant you production on a time limit.

But the CI/CD system did push directly to production. Your tests had to pass and you had to make it past code review, and our code review process was not trivial. We even had a complex code owners setup so if you touched things form other teams, you'd need their approval.

We also used Github's "Merge Queue" tool which would attempt a virtual merge, run all the tests, and only merge/deploy if that passed.

We probably deployed 10~20 times a day in the busy season, with dozens of developers. But we also had 100s of tests (it took 15 min to run the tests distributed across 8~12 nodes. It was almost impossible to run the tests locally unless you ran them overnight. One guy tried and it took 8 hours)

Parts of it were probably way too complex, but it was amazing how smooth the whole process was

@wojtek322 I thought that's what OP meant at first ... oh god vertical scroll .. which only makes sense on mobile .. and still doesn't make much sense.

modern? cmake is ancient by now. There's ninja, meson, etc. etc.

She's going to fuck uk. us is going to get fucked when the AI-generated appendix bursts.

It's nice when you can trust your team.

@torbuntu It's honestly about the same. There's more general information/tutorials for Docker, so there's always a chance you run into a podman specific issue (you likely will at some point).

Podman also has the advantage of being able to run containers as a regular unprivileged user, if that matters to you (but if you end up do needing certain CAPs, you'll have to run it in system mode anyway).

It is good to help support and give feedback to the Podman team though. Having multiple mature container runtimes is a good thing for the opensource eco system in general.

@donkulator Stop noticing! Noticing is anti-septic.

Back when I was running stuff on podman, it's API isn't entirely Docker Engine compatible. Little things like being unable to handle gzip image streams got me once:

https://stackoverflow.com/questions...

Using docker compose with it back then was also hit or miss. It would mostly work with warnings.

I'm pretty familiar with the Docker API at this point. I've written my own build tools against it:

https://github.com/sumdog/bee2

It's pretty complex. There was a native FreeBSD implementation of Docker that used ZFS+Jails but it went unmaintained a decade ago.

I'm sure podman has gotten a lot better. It's been years since I've seriously used it. But I guess all the comparability issues have made me stick with Docker for a while.

This works just as well and is way shorter and you can run it in bash:

:() { : | : };

@donkulator I need to add these to my nginx config.

Confluent Cloud, hosted Kafka, puts a hard timeout at 30 min, even if you're actively doing stuff in the WebUI. It has a VS Code plugin .. which also has a hard 30 min limit, and requires to you re-auth, but the link/launcher in VS Code is perpetually in various states of broken. It's all so tiresome.

At my last time I broke user stories down into individual tech cards and lead the estimation sessions on them. I did it all the time. I thought they were pretty well detailed, and I was pretty good at them.

@asgs It depends on the manager and management style really. There are certainly managers who are still involved in coding, but those who do a lot of it are usually delegating and trusting people.

There are other types of managers who are better at big picture, process, building trust, etc. It takes a lot to properly design a system and trust your team to implement it. This is the first shop I've been at that requires manager code reviews, not just your co-workers.

Coding based management roles have really moved into what's now called "Principal Engineering" which is a big mix or architecture and planning.

I could see a lot of good potential process changes at the place I'm currently at. They are really behind in some ways, and I've worked at a lot of shops big and small. I guess we'll just see where things go.