basicneo

> It was super awesome when we were able to delete a huge chunk of the GCS Connector for Hadoop

Is this public?

Do you copy the entire 50GB on every backup, even if you've only changed 1MB?

As functionality increases, configuration knobs cause a combinatorial explosion in testing and compatibility concerns, which disincentivises them.

As an alternative, I choose to use stable platforms that are designed with modularity, strong abstractions and internal programmability, so I can choose when to keep a particular workflow/UX, and when to adopt a newer style.

That platform is Emacs.

Yeah, I meant that Jane Street aren't doing this to find a solution that'll improve their own tech.

I expect in a constrained environment, someone on Kaggle will beat Jane Street's solution.

By a similar argument to https://danluu.com/sounds-easy/ , no one will beat Jane Street in a weekend.

Jane Street's hiring standards exceeds FAANG's.

This is a hiring/branding strategy. Good luck to them.

In a post about Rust adoption, it seems on topic in a threaded forum that commonly discusses pros and cons to discuss pros and cons.

This is an opportunity to educate/advocate. Someone who wants to do that might link to a good article, rather than just attempt to shut down the discussion.

I think you're being downvoted because this is a bit passive aggressive.

That said, I sense a bit of Rust bias on HN.

The cost to learning Rust might be low but the opportunity cost might be high, since there's many other ways someone can improve their career.

"Good luck with" is slightly negative, but overall I don't see the post as having the traits you describe.

They're replying to a line about a Java comeback by saying there's a lot of big data libraries/systems used by Java developers that it'll be difficult to compete with.

I don't know the big data space well enough to judge.

What's the significance of it being RISC-V?

That sounds great!

+1. It's reasonable for projects with large backlogs to do a quick prune with lower quality, either via manually or via tooling.

If something is particularly pressing, it will re-appear. It's not ideal, but neither is a large backlog.

Can you explain a bit more about the usecase remote supports?

I agree, FWIW.

No, it's all degrees of risk.

The only accurate way is to quantify the risk, rather than blindly assert "extremely obnoxious behavior".

Guix and NixOS are a generation ahead of other distributions. It's great to see Guix's focus on trust and reproducibility.

If they can grow enough to be able to support LTS, they'll be a compelling choice for enterprise too.

When someone drives a car, they impose their estimates of safety on pedestrians.

Is that obnoxious?

Even if it were true that a hobby project is just a temporary reprieve from burnout... so what? They're free to do so. And maybe we'll all get something good out of it at the end.

The author has a history of delivering quality OSS projects: i3, Debian Code Search, RobustIRC, gokrazy.

Sounds like Arch is still x64 only, if you have to use a derivative to use other architectures?

https://www.caradvice.com.au/658919/australia-could-become-a... might explain some of it.

I'm not sure what experience you've had with Emacs, but many Emacs users think of Emacs as a platform (with unique benefits) rather than a text editor.

I use Emacs as a platform, can see the benefits of using Emacs as a window manager, but consider the disadvantages (in light of Emacs hangs or crashes) as too much, versus using i3.

I'm also a bit disheartened by the direction of the web.

Emacs has been a decent haven for me in my search for privacy, interopability, programmability: my user agent.

That sounds like a great idea.

+1. For many years I heard strange things from Emacs users about how they wanted to do everything in Emacs. It's just a text editor!

Now I'm one of those strange users. It's not just a text editor.

If the author had suggested learning a few key bindings or setting up CI, sure.

But a homelab that's cost 10-100s of hours to setup and maintain to "sparing me the headache of dependency conflicts and TCP port collisions"...

I'm not saying I've proved it has poor payoff. I'm saying I don't believe it has good payoff, similar to all the other alleged "helps me work better" usecases for homelabs.

Homelabs are fun, and a place to learn/experiment.

Disclaimer: I have a homelab that I spend time working on.

Homelabbers justify their work wrt productivity: "sparing me the headache of dependency conflicts and TCP port collisions".

But let's face it: tons of productive developers manage without having spent 10s-100s hours on a VM-per-project homelab setup.

Homelabbing is a hobby, and that's okay. Researching parts, building computers, automating provisioning, self-sufficiency, etc.