Nato Boram

There’s also lots of people who made an account in multiple instances before realizing that you don’t have to do that

What’s even the point then?

The point is that you can enable each separate extension you want running on your code editor or uninstall them if you’re unsatisfied. This makes it as light as you want it to be - or as heavy as you need it to.

I was doing fine with just vim and tmux

VSCode is like vim without vim controls and in a browser. Seen that way, it makes more sense. With Vim, you have to hunt for obscure Github repositories and follow arcane installation instructions for hidden extensions that you may or may not need and you have to learn a whole-ass keyboard-shortcut-based programming language just to use any of it.

With VSCode, you click on Extensions, search what you want and it’ll probably be there unless it’s a toxic ecosystem like PHP/C# or some niche ecosystem that no one heard about.

“An ex-Netflix engineer’s take on piracy; in a YouTube drama near you”

So hard to understand ><

Trying to find good subs

Clients can work around it by making a search on the home instance that filters by community id and submitter id. Something like this.

You can’t have content addressing because it’s mutable. On the other hand, UUIDs are made for that. There’s even multiple types of UUIDs made for distributed computing with namespaces and such.

Amazing. One feature that is desperately needed on Lemmy is to open a post in another instance, not just a community or a user.

Well, that reminds me that Mastodon has huge, unresolved problems, such as tags being part of the post’s body like Twitter rather than being a separate field like Tumblr.

Reading tweets with a hundred hashtags at the bottom seem really thirsty for attention, which is bad because Mastodon wants to fundamentally work with these, yet doesn’t have good in-post integration for them. It makes interactions less genuine, more performative.

Rome wasn’t built in a day, and Mastodon won’t be good tomorrow either. In the meantime, you can vote to make it better on https://github.com/mastodon/mastodon/issues/10743.

Cats are all the rage on Lemmy at the moment

Tumblr is a blogging experience that’s similar to Twitter, but more focused on the user itself than on the central feed.

• You have your blogs and you post there. Yes, you can have many blogs.
• There’s global feeds with posts from all users, potentially including yours.
• Posts can have non-intrusive hashtags, meaning they are not #partOfThePost, but in a separate, smaller, dedicated section of the post.
• You can’t post stuff to someone else’s blog, but you can comment on their posts. Comments are tiny next to the post.
• You can quote posts, but that makes a duplicate in a blockquote rather than linking to the original post like Twitter

It’s still like that with programming languages like Go and Rust. Job offers are exclusively for senior staff engineers with 5 years of language-specific experience.

Yep. And clients would be able to participate to the seeding.

Servers software developers would still have a massive amount of work to do to implement IPFS integration, but it’s doable. IPFS also has work to do here to make IPFS work natively with cloud storage protocols (like Amazon S3), but it already exists.

One issue with open source software is that you often have to pick the least-effort solution to avoid burning out your free labour. Free time is limited, and if IPFS takes slightly too much work to add, then it’s off the table.

And also the importance of APIs. It’s the only reason why there’s so many Lemmy clients compared to Kbin.

Right? Like, my app is definitely not ready yet its source code is available for all to see. And since I’m currently inactive, you could even fork it and get a bigger following than me if you wanted to.

These people just think too highly of themselves.

That only-one-ignore-without-premium thing is really asshole design, though

A mod? It’s not like changing instance will do anything in that case

It’s already happening on Pixiv…

A job shouldn’t force you to modify your body in any capacity

Yeah, the discovery process is shite on IPFS. You kinda have to cheat it to get it to work with something like .

Idk if it’s inefficient with large data, but it’s inefficient with compressed storage, as it does block-level deduplication, which is very cool.