clacke2

> Making compilers slow and complex doesn't mix well with JIT compilation.

Funny, I was just thinking the opposite: Compiler-driven parallelism loses against CPU-driven parallelism because the CPU has live profiling. With a JIT the compiler can have it too.

The debugging problem on the machine-code level becomes less of an issue when most people write higher-level code too.

> For some reason, the BiiN project was terminated in 1988 and Siemens was not interested any more in it.

You mean it was biinned.

Of course. The vigorous policing is decentralized, so you can choose which community standards you would like your interactions to be guided by. There are safe spaces, there are crazy freezepeach places, and there are many places in between.

A US server may have content that is illegal in Germany, a Japanese server may have content that is illegal in the US, and an LGBTQ-friendly server may not allow subscribers from a server full of "gendercritical" people. You can pick where you feel at home, don't get ick thrown in our face, and don't get accused of ick for things that seem harmless to you.

Would they have even been there if not for the terrorism?

Precedent doesn't come into it (until it does), it's all about what's in your contract.

Work for Red Hat? You own your free software contributions, even those made at work. Work for Google? They own even the small lump of green putty you found in your armpit one midsummer morning.

Many companies will apparently put in grab-all clauses but back down (but Google won't) if you call them on it.

We didn't have to be first. Other primates might have picked up the rock before we did, but eventually lost the race. Maybe the fire-cooking, rock-throwing domesticating niche is just that competitive.

So it might be more like "we are the ones that remain", but yeah, either variant is a spin on the old weak anthropic principle, which is never quite satisfying.

I'm reading it on Firefox Klar/Focus in reader mode, no plugins, and it looks great!

I found shipit so amazing. I went there to order like 5 discs and the form came back saying that hey, actually the discs are pretty cheap and our main cost is postage per shipping unit so why don't you order like 15 more just in case? :-)

If it is in fact the case that only some redex models of `letrec` and `set!` were fixed, then it is indeed misleading that the release notes say "Racket's implementation" of them was fixed.

Unless you're the one who did the work, I'd rather trust the release notes than your unsourced assertion.

That said, just because a bug was fixed in an important function shouldn't be an indictment against the language. Racket is stable and productive, and formal modeling can reveal all kinds of corner cases which may not have been exposed in typical use, and may not have security implications. Maybe it was edge-case performance, maybe garbage collection, the release notes simply don't say.

I'd rather use a language whose authors improve their tools to discover and fix issues than one where this doesn't happen.

Friendica is still alive and can do this!

It's primarily microblogging-oriented, but you can subscribe to any RSS feed too. And users (including RSS feeds) can be grouped, so you could have one group "tech feeds", one group "cat accounts" etc.