No, ML works fine with large output spaces too.

The problem is that the solution space grows super exponentially, and if you need to find an exact one, then the number of samples too.

Thanks a lot for this.

I never understand why Mozilla doesn't allow to enable this in the settings.

Apple should not have that kind of power, though. They have been wrong before.

Instead, the scrutiny and investigations should be public so that the developer can defend themselves.

> Apple should not have that kind of power, though.

Oh, but I’m happy they do

Also, notifications go via apple’s servers, you can’t really force them to deliver everybody’s notifications for free and without any discretion

Theses marketplaces require transparency, accountability, right to appeal, adjudicators, and so forth.

In other words, the rule of law and a fair impartial court system.

> few site reliability engineers are looking to add extra ways for the site to be taken offline

If your company has anything close to "reliability engineers", then you already have legal and finance teams too that can sort terms out.

The discussion is about companies that do not even have departments to begin with.

Kind of, yes.

Nowadays (for almost 20 years actually), Windows supports and encourages programs to deploy their dependencies in their own folder.

It is trivial to do for developers and works perfectly for users.

Some people claim security updates are a problem, but many apps do not deal with untrusted input for starters. For those that do, like browsers, servers or your office suite, you really should be using one that is supported and updated automatically.

> So as firefox creates threads, I would think memory could get tight with a statically linked FF.

Threads have nothing to do with that and don't increase memory usage due to dependencies.

Why? Using different Python versions is a solved problem and no containers are required for that.

I guess you are trolling, because a native app can print "hello world" in around 100 bytes.

That's a difference of more than 6 orders of magnitude :')

It is true that a web app needs a browser. But a native app needs an operating system (OS). Something that I find fascinating though are apps that you can boot into, eg. they do not require an OS. But if you want graphics, mouse support, play sound etc, it becomes quite complicated. Compare that to a web app where you get all that functionality for free, with no libraries needed. And in between we have the shared lib utopia with Debian et.al and dll-hell with Windows. Each stack have their advantages and disadvantages. I however don't see bundle sizes on the web as a huge problem, at least not yet. With service workers you can download the app once, then even use it offline. Everything is just one click away and can be reached via an URL. And if you stick with "vanilla" it can also be fast and lightweight.

Wait, what?

So you think a kernel is large and has to be counted in the total, even if a browser also requires a kernel.

Yet you think web apps don't need libraries, so you are not even counting the browser and its dozens of libraries?

Casting away constness in C++ is something it is virtually never done, and has to be done very explicitly.

> https://github.com/search?q=const_cast&type=Code

We must have very different notions of "virtually never": https://github.com/search?q=const_cast&type=Code

> has to be done very explicitly.

That doesn't make it visible to the caller.