I know I’m late to this but here’s my (probably insane?) take. We use Subject-Verb-Object in English right? So, hear me out:

dialog_create_tab(...)
dialog_open_file(...)
dialog_close_file(...)

Software optimization is mostly not a language-level problem. I’ll be dailying my 3-year-old OnePlus 9 Pro until it starts missing out on security updates, but it will probably still be “usable” long after that. Support/updates aside, my 6-year-old galaxy s9 can still run most normal apps. Hell, I got the most recent lineageOS running on a pixel 2 XL from the year before that and it straight up felt fast as long as I wasn’t playing some super intensive game or something. This isn’t an android vs. iOS problem, it’s a “developers of [insert flashy new app here] either not bothering to put effort in to optimize their code or being forced to push out a minimum viable product ASAP” problem.

Edit: fixed my hyphen use

The least expensive MacBook is still $1000, closer to $1500 if you spec it with reasonable storage/ram. It really isn’t that much of a stretch to add $100-300 for a 1080/1440p monitor or two at a desk.

I think blocking downvotes is an option built into Lemmy servers that can be communicated through the API. I know there are a decent amount of instances that don’t federate downvotes because of toxicity concerns.

In the grand scheme of things the difference between C, C++, and Python isn’t meaningful when operating over a network (edit: for a single-user system). It’s very likely that the difference for thread OP is just caused by weaker connections to specific repos.

We’re talking about a package manager, not a game, network server, etc. On a basic level the package manager only needs to download files from a network and install them (OS syscalls for reading/writing files, these are exposed C functions or assembly routines), or delegate to a specific package’s build setup (which will also likely be written in a compiled language)

That’s partially my point. You can never be 100% safe, but there’s a lot you can do to increase your safety besides just relying on intuition (edit: because intuition is usually the weakest link, see social engineering/phishing tactics). Anti viruses (when they aren’t just bloatware) are part of that.

Your second point about not meaningfully defending against backdoors and vulnerabilities is kind of against the point. You can totally defend against backdoors by not giving apps admin privileges, limiting network access, etc. so that damage can be limited even if an exploit happens. Then, if some backdoor or exploit is discovered, it’s only as dangerous as the permissions you give that app.

Linux gets viruses too (see recent xz-utils vulnerability that almost got into production environments) and its kind of a shame that corporate antivirus software like Norton and McAfee end up ruining the reputation of antiviruses. In theory the idea of having a software that can scan for common viruses is a great way to increase security, even if it shouldn’t replace common sense. I’m not too sure if there are any good FOSS antiviruses, but if there aren’t there should be.

I love not having downvotes federated 😎

I feel like squash and merge on GitHub/GitLab is nicer for that anyway though, it makes the main branch so much cleaner automatically

You’ll have to settle for naming your child after one of their public keys, but then your child will only be able to talk to them.

Oh true I didn’t realize the semicolons were missing (that’s what the compiler errors are for)

Could also be rust (no parens on ifs there either), kind of hard to tell with just an if statement and some function calls

Hi, it’s me, the intern refactoring the spaghetti .NET core backend. I’m not in a basement though. AMA