YOU and ME and HER or Doki Doki Literature Club moment.

Well, they didn’t mention Misskey, *omas, Gotosocial, Lemmy…

The author doesn’t mention any of the observations that are considered to be better explained by dark matter than modified gravity, then concludes his colleagues are delusional. That’s a smear job, not science.

For dark matter scepticism that’s intellectually honest, I’d suggest for example https://backreaction.blogspot.com/2017/01/the-bullet-cluster-as-evidence-against.html?m=1

The “sandwich problem” looks like just a lack of a traverse function to me.

try leftypol.org next

Nothing in Git requires reversing hashes (which requires a lot of computing power), there’s only forward computation (much cheaper), and there isn’t any other method of establishing global consensus either (like proof of stake). This means no one needs an expensive rig to create git commits and fork whole repos, but on the other hand the hashes only prove the integrity of one set of commits; there is no global consensus on any repo and there has to be some sort of access control to prevent someone from force-pushing or replacing whole histories, central maintainers to decide what gets merged, etc. At best, to check if the upstream did something fishy, a mirror can be compared to another, or to a local copy.

edit: or perhaps, it wouldn’t be technically incorrect to call a git repo a blockchain, as the distributed global consensus without human maintainers may not be part of the strict definition, but this seems pretty anachronistic and pointlessly confusing to me.

Most functional languages, like Haskell, ML family (SML, OCaml, F#). They’re usually not using an enum keyword though.

It doesn’t try to replace trust with computation, which is a crucial difference.

Interesting that szmer.info has 65 active users monthly, I think I’ve seen like 5.

I think the idea is that downloadable binaries are merely a convenience, but that might be a holdover from a more innocent time.

I think the recent allegations (whether true or not) that period tracking apps in USA (and other counties where abortion is illegal) might make users’ data available to goverments could make people reconsider.

I don’t want to sound jaded, but I have a feeling that eventually someone will use it to run Electron apps in Flatpak on Linux compiled to WASM running on Chrome, all because portability is hard and Qt’s licensing is unpleasant.

Anyway…

I’m not entirely sure about it, but this issue: https://github.com/hyperium/hyper/issues/2420 and this linked one: https://github.com/seanmonstar/reqwest/issues/1162 seem to be resolved.

This seems to be a 1 year old bug that appears to be already fixed, am I missing something?

Higher-kinded polymorphism without having to use the GAT workaround, dependent types, comptime like in Zig instead of proc macros, fast compiler.

Don’t know much about the slur filters myself, but it does make me wonder about the type of characters that would fight so hard to allow the use of more slurs.

Some of the complaints are that the filter is too trigger-happy, and that it doesn’t (or didn’t until it was patched) respect languages other than English. Problems with this sort of filtering are known since at least 1996, when AOL banned the entire town of Scunthorpe. It’s one thing to consider this a reasonable price to pay for keeping nazis out, but it’s disingenuous to say that everyone who doesn’t agree must be a nazi themselves.