eonil

Your definition of can is different from most people are thinking. Usually people don't say can on illegal stuffs.

If you really think you can do something illegal because just you're able to do that, oops…!

Mozilla has Emscripten. I don't think they're really JS only.

U.S. itself is a big huge startup from Europe. Maybe that's why startup people are all in U.S.

Wow, this is real!

I am curious whether the tuition fee is same on international (non-EU) students or non-research courses.

Does U.S.(or any other country) government recognize this degree for working visa or immigration? (H-1B / green card)

I have heard of it. But the problem is data structure is not a pure-tree, so we need arbitrary referencing in a version of state, and FRP doesn't seem to provide a solution for it.

I was talking about pretty realtime applications like games. I'm sorry that was unclear.

Though I am not much familiar with functional programming, as far as I tried, one of the hardest problem in functional style is referencing to a mutating object.

Where an object is immutable, it has to create a new instance of itself for mutation. That invalidates existing references. (if the language support object referencing…) The only solution I can figure out is putting unique tags for each objects, and lookup them for each time. In other words, I have to reference them indirectly which introduces lookup cost. And this cost is usually unacceptable.

If the data structure is pure tree, functional style usually superior on simulations - easier writeup and easier debugging by retaining all the intermediate state. But most games are graph structured. They usually contains many links to arbitrary node which are very hard to be updated together with target object mutations.

If I am wrong, please correct me. I want to use functional styles on games, but I don't have a good solution for referencing problem.

Of course I welcome removal of GC syntax, but I am getting in doubt on when Rust language specification can be stabilized. Version 0.9 seems it must be stable now, but they're still putting big changes on the language.

I agree. But I also think the name shows the author's primary intention.

Pirating means they will not RESPECT the owner's rights including licensing like GPL or even crediting contributors.

Nice summary. Anyway, it seems he haven't tried Go seriously yet.

I also once had this idea. Because Go is mostly the only matured language which can describe processes and pipes in a program with solid flow control. But still, it doesn't feel nice to interact with shell…

Unfortunately, D's coroutine support isn't stable like Go's one. It has unclear stack size limit, it was simply impossible to me to make a robust D program which uses coroutine. (please correct me if I am wrong.)

I think that's absolutely reasonable that they remove your premium if you want to remove other people's safety just for your own fun.

IMO, that's because the AI is not yet proven to be actually safer than human. And it's pretty interesting what will be happen if the AI eventually comes up. Anyway, it doesn't seem to happen anytime soon.

Oh I don't know how it is a penalty.

Mozilla vs Google competition is always a lot better than Google monopoly on open-source OS market.

I don't get what's really cool. Viewport angle and camera path is fixed, than what's really better than just a video? Can let me know the point of this tech?

I tried pressing `c`, but it showed only some cracked images, doesn't seem really meaningful.

It's really hard to think the term nil is a value.

IMO, this is a bug in language design. If it's truly a value, I think they should choose a different term rather then `nil` as like they choose `panic/recover` instead of `throw/catch` because they're semantically different.

And also, I really don't know how should I detect nil reliably for an interface type.

I am waiting for the concept-lite feature support but I don't know why Clang people don't even mention it. Am I missing something?

This is even more interesting.

I had a big expectation on Boston Dynamics, but not anymore because Google purchased them.

Your assumption at first sentence is wrong.

He told employees. That means completely different with just people. Employees are fully informed of it, and even already have it for free. (further, they are being paid for it)

If someone doesn't want to use a product which s/he already have, and knows very well, isn't it enough to think that's a crap?

It would make sense to write an easier wrapper around current C++ standard library. Because C++ provides very strong inlining, just simple wrapping with better method names will not make any performance hit.

But IMO, naming is one of the hardest work.

I really wonder if those teens realize once dispatched data cannot never be forgotten.

Yeah, port system is fantastic. It takes longer, but just works. The whole system is designed to be robust, there's no broken shit if it's done successfully. *It reports error on failure at least.

Linux package managers are fast, but usually broken silently in some ways - which are critical on robustness - and I always discover myself manually doing what port system does.

CentOS packages are relatively more robust, but it's too outdated, and installing new packages nothing different from any other Linux distros.

You talk like an arm dealer. "The restriction only increases cost to get forbidden stuffs, and they will eventually take it. Sum's zero, so nothing beneficial."

And then, this article is hooking me with title…

It would be better to define HTTPS by default, and plain HTTP for exceptional cases. I would use HTTP only for strictly public and eternally cacheable big files.