Also: great API for creating own plugins.

Kudos for solving this. And yes, this is too much for a job interview.

I don't think this an unreasonable puzzle for a role that involves problem-solving or security.

You can read the code, figure out what primitives you have, translate the problem into something simple (I like card and board games), then map it back.

Problem: two processes each do this ten times: - read the counter - do any number of other things in other processes - write back what you read, plus one.

Game: You have two piles of cards. Each card costs 1 move to play, but lets you discard any number of cards from other piles at the same time.

Solution: play one red card, discarding everything but one blue card. Play the single remaining blue card, discarding the leftover red cards.

Back: process 1 reads, process 2 runs 9 steps, process 1 writes, process 2 reads, process 1 finishes, process 2 writes.

Love the note on rejecting ai-generated code and about the alleged authors who don't understand what the code does.

I think most of the language implementers know about it.

Here is a relevant line from Go, even with a comment that it's about overflow:

https://github.com/golang/go/blob/19e9231/src/sort/search.go...

It is really hard to believe that no one defends antirez&co as protecting themselves from greedy overpriced SaaS businesses who were parasiting others hard work. What about them saases breaking a social contract of giving reasonable prices adequate to the effort put into a product (or a wrapper around the product)?

It's possible to hold both positions at the same time.

In the world of legal consequences it is not.

Or please explain.

Erm I am not sure if I was understood correctly - I meant legal consequences from the perspective of redis guys on one side, and saas guys on the other side..

One simply can't admit that they both were doing a valid thing

Lots of people defend these licenses on exactly those grounds. I, too, find it pretty mystifying that this seems to be a minority perspective here. At least this seems to be the case among the people who speak up when this comes up here. It's hard to know if there is a less noisy majority who feels differently about it.

I think the "silent majority" is just fine with this change. The OSS "extremists" will certainly be the noisiest about this - everyone else will continue on with life as usual, especially if they have an understanding of the concept of business pressures.

That's what I think too, but it's hard to know for certain.

If you just call SaaS "hosting" does that change your view of things? Because I remember when commercial hosting for free software was actively encouraged.

It's not that weird. With a FOSS license, I know I can use a program for anything I choose to and the terms will never change. I might launch a startup with a friend tomorrow using a package and complying with its license terms, and a million dollars later we're still legally entitled to use it.

Adding constraints later is a non-fallacious slippery slope. The restrictions Redis added aren't directly onerous to 99.999% of users, but now the world is divided into people who are allowed to use it at will, and people who can only use it in certain ways (or not at all). That's an enormous change from a FOSS license. And if they change it today specifically to lock out Amazon, will they change it again tomorrow to lock out my startup? The thing is, now that they've established precedent, there's no way of knowing what'll happen later.

Again, I know I can use GPL or MIT or BSD software for anything I want as long as I play by the exact same rules as everyone else. There's nothing that says you can use it but I can't. The difference between Free to non-Free is vast, even when I'm not affected (yet).

Your answer is "it's not that weird what if I wanted to launch a startup competing with azure and amazon"?

Well, Id say in that case youve got waaaay bigger problems than a license.

>Adding constraints later is a non-fallacious slippery slope.

No, it's very much a fallacious slippery slope. Constraints are the very essence of licenses and contracts.

There was no crypto stuff during the floppy times, and that's a lot of extra code

Oversimplifying warning:

Must of the problems described here come from a cultural setup that you can't tell or even suggest your managers that they are stupid and/or someone above them is even more stupid.

I noticed that this is especially painful in the US companies, we in Europe seem to be much more lucky with telling people that they do stupid sh*t and not lose the job after telling that. But YMMV

Being CGo-free is very desirable when you can't/don't want/don't bother to link C library deps for some reason. On example when you don't want to do that is cross-compiling for another architecture and not really being able to control what runtime libraries are installed on the target.

So, it simplifies build pipelines, at the cost of being a bit slower in the runtime.

This is also nice to read about the topic: https://dave.cheney.net/2016/01/18/cgo-is-not-go

Yeah this gentleman is Jan Mercl and is quite known in the Go community. I think his most known project is the sqlite driver in pure Go, as opposed to mattn's driver.

Modernc.org is kind of umbrella of his projects.

Once I got fired because commented when boss smoked weed at work and liked to be intoxicated at workplace by other means as well. Who can beat that