Comments by "MrAbrazildo" (@MrAbrazildo) on "Computerphile" channel.

  1. What a moron!
    4
  2. It has been used nowadays as a lab for C++ dubious new features. When approved, they become C++ features.
    4
  3. I feel comfortable enough with C++. Even so, I'm ready to left it for a better 1. Problem is it doesn't exist, so far.
    3
  4. As long as I understood, the invasor must login in the target machine before attacking, right? It's not a remote access attack. If so, it's not that dangerous.
    2
  5. +3.14 Already evolved to C++.
    2
  6. From which console will this run? Don't say "Sega's veteran console".
    2
  7. "Bjarne Stroustrup (C++ creator) solved this problem, but the world continues to act as if it wasn't" - mr. STL, from Microsoft. C++ solution for this is so much cleaner, elegant and faster: write a tiny destructor in a class, and it'll automatically do the job for you. It just won't catch the lost of all references, if you aren't using a shared_pointer. Even so, there are classes implementations in its library, that already have all this done, so you not even have too write a destructor anymore.
    2
  8. +thepokkanome "engine" is another 1. In chess, it means the AI.
    2
  9. Calling a f() at each looping cycle costs about 60% performance - I measured it once. However, in C++ there's the 'inline' keyword, which makes the compiler write the f() inline, unrolling it.
    1
  10. Mr. Joe Armstrong, you are talking about Java, not C++.
    1
  11. That's why you should control your game character using the KeyPad, with your right hand: the keys fit much more accurately in your fingers.
    1
  12. In Mass Effect there was the concept of a Virtual AI, not truly inteligent. There's a tale, from Vernor Vinge, I guess from the 70's, figuring out a space ship of this kind. Maybe this is the way.
    1
  13. It's the best thing you will learn. But then C++ will be even better.
    1
  14. Well, I would call this "essential" only for professionals, because beginners use to solve this through a huge 'switch'.
    1
  15. If JVM prevails... it will be the end for mankind!
    1
  16. The end!
    1
  17. So, do you assume many people still have it?
    1
  18. It's better to write in C++, using a 'static' variable inside the f(), to "yield" its value.
    1
  19.  @MadocComadrin  "So many tools, so little time..."
    1
  20. In a project, I had a recursive f() in the main loop, which is disastrous for performance. I took it off, using several 'goto's. However, I discovered that it got slower! o-0
    1
  21. I already tryied to code that, with C99. It worked, for almost all OO features, but it was way worse than C++.
    1
  22. Nothing new. Plus, it could be optimized, making an array (not-linked), of data, and other 1 to memorize the "jumps".
    1
  23. I never heard about "Triple Ref Pointers", so I thought it was something new. And, by the way, I gave thumbs up for this video, precisely because it's a "quick tutorial".
    1
  24. ...with all due respect for C.
    1
  25. +Hopson Java is such an awful language exactly because of that (trying to solve problems which don't actually exist).
    1
  26. +Ondra I'm talking about "issues" that don't exist, like pointers, friend qualifying, destructors, and others.
    1
  27. +Maor Wrong! It's an evolution. C has retired after that.
    1
  28. std::map is slow. Better to use 1 or 2 std::vector.
    1
  29. +Tegan You can have 1 vector filled with "unordered" indexes from, let's say, pixels in screen, and the other (vector) with "orders" (string), and a lambda f(), to coordinate them. So, you say '[]("HIT THE MAIN CHARACTER")' , and the lambda finds this string in 1 vector, measure the index from the beginning, then returns the match pixel, from the other vector.
    1
  30. Haha! On my college, there was a dumbass that used to answer like that!
    1