[carterparks]

I can't understand why anyone would choose to use this language. Clearly it is unreadable and I highly doubt this scales. Can somebody give me a good reason for using this on my Facebook clone instead of a more mature language such as PHP, Python or Java? Also, I heard somebody was developing a web framework, Trolls under Bridges. Has anybody heard about this?

[lkrubner]

I don't think this is a language people use in real life. It's more for testing. When I went for a job interview at Alexander Interactive http://www.alexanderinteractive.com/ one of the tests they gave me was to write a short bit of code in Brainfuck to reverse a string. The string they gave me was "Hello World". I was allowed to look at the Wikipedia page regarding Brainfuck. They had an interpreter running, into which I could type my code and see the results. I had 15 minutes to make it work.

I think that is where these kinds of languages are used. I think the idea is to test how fast a programmer can learn something new. (For the most part, I'm doubtful about these tests.)

[xnxn]

My friend, I believe you've missed the tongue-in-cheek nature of the parent comment (and, in fact, that of Brainfuck itself).

[lukejduncan]

Sounds like either they were messing with you; you are messing with us; or Alexander Interactive doesn't respect their applicants

[burgerbrain]

Impossible, or seemingly impossible, interview questions are actually quite useful. It can be quite informative to see how the applicant response to such a situation, and what their "Oh shit, wasn't expecting this" emergency problem solving techniques are.

[Helianthus]

Which is another way of saying, messing with people is a respectable interview tactic--see how they react to a little nudge.

[RickHull]

> Which is another way of saying, messing with people is a respectable interview tactic

I don't think so. Rather it's that technology jobs (and others, surely) often confront frustrating, confounding, and even downright impossible tasks. The ability to identify a situation as impossible or otherwise -- to drill down to the point of conflict -- is prized, as well as the ability to keep it together mentally and keep plugging away.

[VladRussian]

beautiful! 12 minutes :

>,[[>,]]<[.<]

edit: of course it should be

>,[>,]<[.<]

[redthrowaway]

">,[>,]<[.<]"

    pStr++;
    *pStr = getc(stdin); //or wherever
    while *pStr {
      pStr++;
      *pStr = getc(stdin);
    }
    pStr--;
    while *pStr {
      putc (*pStr, stdout)
      pStr--;
    }

It'll print the reverse of a string, but it will also print everything before it until it hits a null byte. I'd add

    <[-]

to the beginning to null out the preceding byte and hope it wasn't important.

As silly as brainfuck seems, I could see something like it being used in nanobots (should such a thing ever come to pass), where memory would be at a severe premium.

[meatmanek]

Brainfuck doesn't really save more memory than any other Turing machine language. However, it could save you silicon space in your hypothetical nanobot because everything is an increment -- incrementing can be done by a chain of half adders, which require fewer gates than full adders.

[VladRussian]

>It'll print the reverse of a string, but it will also print everything before it until it hits a null byte.

exactly! The brainfuck execution starts with the pointer at the first byte of zero-filled array:

http://www.muppetlabs.com/~breadbox/bf/

So the first ">" in my program moves the pointer to the next byte, and this skipped first byte would serve as a stopper when pointer moves back.

[redthrowaway]

Ah. I wasn't aware of that part of the spec. In that case, then, the code would work.

It's an interesting language, and one that I'm not convinced is so entirely without practical application as is commonly stated.

[ashleymoran]

I think this is a case of premature optimisation. Trolls under Bridges is struggling to hit v1.0 due to the limited memory capacity of trolls, but someone is hard at work on a Node.js implementation of the interpreter. This will allow spawning a new worker process, or "troglodyte", for each request, and will be totally webscale (there will be no unnecessary trolling for events).

In addition, there's also a JVM port of TrollScript in the works, which will add a static type system to the language for troll-safety. This should alleviate any concerns that the language would be unmaintainable for large systems. JTrollScript will also have type inference to maintain the terseness of the original language.

[_3ex7]

Re: Node.js implementation

my buddy has been hacking on this and he says the benefit of being able to troll on the server the same way you troll the client is a huge benefit.

[reinhardt]

True, but at the same time it's no match to the TIT (Troll In Time) compiler in cases where every ounce of performance counts.

[wtracy]

The various derivatives of Brainfuck are all joke languages.

That said, Brainfuck originally had an at least interesting purpose: The idea was to create a language that could be handled by the smallest possible compiler.

AFAIK a Brainfuck compiler can be implemented in less code than a compiler for any other Turing-complete language.

Still not practical, but interesting nonetheless.

[tombell]

PHP, Python and Java don't have a 'lol' operator!

[brodie]

There's Like, Python, which is arguably even more readable than plain Python! http://staringispolite.com/likepython/

[Semiapies]

Sorry, meant to vote up.