eska

I don’t know zig syntax, but wouldn’t it be possible to put this common pattern into a macro and simplify it to mostly a lambda on V?

He probably meant that free users are more dickish. That’s my experience

The submitter isn’t the library author.

I also stopped reading the article there. This is major version 4 of a library about providing higher level db actions, i.e. transactions are paramount to make them atomic

If I was going to pretend I’m sick I’d just say I have diarrhea. There’s no way for the doctor to check it. So yea this is theater.

In Germany there’s private insurance which can only be chosen by people of higher salary and good health because the rate depends on their condition when they sign up. So this data would be biased towards sick people, people doing jobs that are more likely to make you sick, and people who are financially less well off.

People would love to build their own servers (and run AI or other workloads on them) and kids would love to build their first PC, but big tech is buying all the hardware and stuffing it into data centers. You make it sound like they asked for it.

This was so sad to read. I quit because I didn’t want to end up this dead inside.

At that point does your keyboard even still have arrow keys?

Sorry to be blunt, but if you don’t want to spend effort on touch typing (and therefore avoid arrow keys), learning vim motions is rather pointless and you might as well not bother.

I just looked up the compression rate of FFV1 because I never thought about this. It’s apparently 4x. More would be possible, but increase computational requirements.

Another use cases seems to be archival of historical footage.

I don’t think that you guys should be debating compressed vs uncompressed, but lossy compression vs lossless compression. Your math seems to derive from a naive storage format.

It should be illegal to have others purchase what you as a company only licensed and therefore aren’t legally allowed to sell.

I’d still use a byte slice for that. Some formats may mix encodings, or have a text header and binary payload. For those cases one would need to use memchr for the first byte, then compare the remaining few bytes. So I don’t think it would be a huge performance impact

Even if you don’t use Facebook they have a “shadow profile” of you. One of the things that is different on the modern internet.

For context, initial WoW was developed until 2005, and up to roughly 2003 Blizzard was going to release it only for the US and Korea, because they thought “Europeans only play racing games” (source: WoW dev diary by Staats)

Even Riot’s rootkit “Vanguard” has reduced requirements for Windows 10.

As both a lawyer and a software engineer, I don't understand why big tech and EC want to scan messages, if they actually want to combat online abuse.

They don’t want to combat child abuse, they want to improve their advertising fingerprints.

Exit strategy for after their political career. Compare with Gerhard Schröder

That is still simple AI and commonly used.

1. Make individual agent AIs that can act on their own.

2. Make squad AIs that can influence their agent AIs.

3. Add even more AIs on top, like a scenario AI.

Each of these AIs can be a state machine, behavior tree, rule system or goal system. They’re exactly equivalent and can be translated into each other mechanically. So the whole hierarchical AI is equivalent to one big state machine.

I started to learn Japanese 30 years ago, and in my experience the people who try to be smart and build systems almost never get decent. It’s procrastination while thinking they’re actually productive.

To add insult to injury this article hasn’t discovered anything new, makes it sound way more complicated than it is, and in the end still requires you to just remember which verbs are of the eru/iru group, and which are not (which was posed as a problem to solve in the intro).

Just make cards and mark the stem, learn it along with the verb. No need for heuristics. If you ever forget, you’re bound to remember the masu-form and can reverse engineer the stem from that 100%.

I changed my opinion about parens when I stopped formatting like C, and used indent rather than parens to denote blocks. That is, a large amount of them at the end is totally fine.

I mean that mostly in the sense that there is huge variance in idiomatic code. So your optimized C/Rust code might be 100-1000x faster than two idiomatic versions of writing that code

I appreciate that you first tried to optimize the original Python code. Idiomatic Python is unfortunately disappointingly slow and not so interesting to compare to.

Aren’t the generated tests superfluous since they just test whether the framework implements validation correctly? This should already be tested at the framework level and add no value?

Does meta have any offline products? Their business model is to connect everything to the cloud (e.g. multiverse with ar glasses before)

There is no “zero” immigration policy, this is a strawman. Controlled immigration is different from opening the floodgates.

Also weird to admit that no country has reversed its birth rate problem, but still insist upon massive immigration being the solution.

Ad hominem. Skilled programmers != old-timers calling others unskilled. Who does fuzz testing, automated integration tests, etc already, that LLMs need as a test harness? It’s not the vibe coders saying “nobody reads assembly anyway” as an argument.

He’s referencing practices at Meta and probably others

I keep hearing this argument on HN. Yes, if performance is something you care about you totally look at the disassembly. You don’t even need to write assembly, just be able to read it. By now I assume that programmers who argue they don’t need to understand and review what the LLM generates for them are simply not that skilled and don’t realize that skilled programmers do this.