I'm new here..



  • @asuffield said:

    And if you aren't doing CS research, you probably don't want to be talking about Turing-completeness. It's really not a very interesting property. For example, as noted earlier, the pure lambda calculus is Turing-complete. However, it is also semantically null - no expression in the pure lambda calculus has any meaning. We have to invent meaning in order to accomplish anything with it, and this is remarkably difficult to accomplish; it took the best mathematical minds in the world many decades to find an interpretation that was sound and consistent. All practical work is done using impure variants, and the pure form is used only to prove that the impure forms are valid. You need a great deal more than mere Turing-completeness to do anything useful.

    I wrote a factorial in LC once. It was horrifying.



  • @asuffield said:

    The real reason why XML is not Turing-complete is because it doesn't include a semantic model at all, and Turing-completeness is a property of semantic models. Asking whether XML is Turing-complete is therefore as meaningless as asking whether yellow is square.

    Interestingly enough, asking "Square is yellow?" (or other variants like "Is square yellow?") make sense in an illiterate sort of way.



  • @too_many_usernames said:

    @asuffield said:

    The real reason why XML is not Turing-complete is because it doesn't include a semantic model at all, and Turing-completeness is a property of semantic models. Asking whether XML is Turing-complete is therefore as meaningless as asking whether yellow is square.

    Interestingly enough, asking "Square is yellow?" (or other variants like "Is square yellow?") make sense in an illiterate sort of way.

    Try telling that to a formal programming language. :P



  • Hello there, ehird. *huff puff* Sorry, I'm a little late to this recently-unearthed thread to wish you welcome.

    I like your avatar. I was waiting for someone to throw in an appropriate "anonymous/legion" reference.



  • A square can be yellow, but that isn't certain.
    But certainly though, yellow is kinda square. 

     Green is much prettier
     

    #################


    Bah it strips the HTML/CSS.

    i meant #A1E826, lovely colour.



  • Wouldn't, by that definition, grayish white be the squariest color of all? #EEEEEE after all...



  • I agree however, green is a lovely color.

     



  • @PSWorx said:

    I agree however, green is a lovely color.

     

    is that really how you see this thread? 



  • Something tells me he used user CSS, or maybe a greasemonkey script, to do it.


    Maybe he just used JS in the address bar for that one page load for the screenshot.

    But you probably knew that.



  • Almost. Firebug actually, for those who want to know :p
    I'm actually glad though, no one suspected I just used paint...



  • Why is half of this topic an argument about what defines code?

    By the way I recently mastered Javascript, I've known basic for a while and am working on learning C++ and PHP, so that clears up the "Can I be considered a programmer" argument.



  • @Mal1024 said:

    Why is half of this topic an argument about what defines code?

    welcome to the internet and geek/nerd culture.   



  • @asuffield said:

    And if you aren't doing CS research, you probably don't want to be talking about Turing-completeness. It's really not a very interesting property. For example, as noted earlier, the pure lambda calculus is Turing-complete. However, it is also semantically null - no expression in the pure lambda calculus has any meaning. We have to invent meaning in order to accomplish anything with it, and this is remarkably difficult to accomplish; it took the best mathematical minds in the world many decades to find an interpretation that was sound and consistent. All practical work is done using impure variants, and the pure form is used only to prove that the impure forms are valid. You need a great deal more than mere Turing-completeness to do anything useful.

    I find that the more interesting property of 'Turing Completeness' is whether someone is touting it as a feature.  Because when I hear someone say "Oh yeah, and QVXZ is Turing Complete!"  a giant Red Klaxon starts  howling in my skull to warn me that that is a euphemism for "You can technically write anything in it, just like in INTERCAL, in fact, it will probably be about as readable as INTERCAL, and if you are forced to use this tool in this abusive manner, seppuku is imminent!"

    This is what "It's Turing Complete!" means.   So don't say that.  You'll make your beloved technology radioactive in other people's minds. 

    I mean, perl regular expressions are Turing Complete on their own:  http://99-bottles-of-beer.net/language-perl-737.html 

    Just try and figure that out. 



  • @PJH said:

    @ehird said:
    Well hey it looks like this is the "young programmers introduction thread" :P

    Hi. I would say "I'm 11", but I ceased being 11 a month ago, thus putting me in the "young programmers" group instead of the "very young programmers" group.

    I know Scheme, Python, Ruby, PHP (unfortunately), C, and a little bit of Java, and a handful of random languages not worth mentioning.

    Also, I'm a mac loser and proud. Nyaah. BSD-derived in your face.

    You're 12, and you managed to compose a whole message using (I presume) a 101+ keyboard without using txt spk, and with punctuation and capital letters? There's something wrong with you... 

    On the intertron, nobody knows ur 12 if u dnt wr1t liek a tard. 

    Of course, I've seen people a lot older than twelve write using sms speak.  

    Oh my poor English Language!  How they butcher you so!  Oh French People!  I understand you now!  I UNDERSTAND!

    Ok, done freaking out. 



  • @phaedrus said:

    I mean, perl regular expressions are Turing Complete on their own:  http://99-bottles-of-beer.net/language-perl-737.html 

    That's just a regexp constructing some perl code, doesn't count. Perl regexps aren't Turing-complete unless (?{ code }) is used (which is cheating). However, the construct "s/re/bar/ while 1" is Turing-complete. I believe that works with any kind of regexp.



  • @asuffield said:

    @phaedrus said:

    I mean, perl regular expressions are Turing Complete on their own:  http://99-bottles-of-beer.net/language-perl-737.html 

    That's just a regexp constructing some perl code, doesn't count. Perl regexps aren't Turing-complete unless (?{ code }) is used (which is cheating). However, the construct "s/re/bar/ while 1" is Turing-complete. I believe that works with any kind of regexp.

    Ah.  That's the trick.  I hadn't figured out what they were doing--my eyes kind of glaze over staring at that much line noise.  However, that would mean that sed is Turing complete, which is just as bad, so s/perl/sed/; s/perl-737/sed-1087/ my comment, and  it stands. 

    (Now, mind you, I love sed, but sed has it's place.  If you are finding yourself using multiple lines and loop tags, for the love of ${DEITY}, switch to perl or awk+bash+sed.) 



  • @phaedrus said:

    However, that would mean that sed is Turing complete, which is just as bad

    sed is. As a general rule, any reasonably sane language that can construct conditionals and loops will be Turing-complete, and anything that can't won't be. Regexps alone can't loop and perlre doesn't add anything to change that, so they aren't.

    Pathological languages like Malbolge require special analysis, but for anything you might want to use you should be able to rely on that test.



  • @phaedrus said:

    However, that would mean that sed is Turing complete, which is just as bad

    You jest, but... 


  • Discourse touched me in a no-no place

    @Random832 said:

    @phaedrus said:

    However, that would mean that sed is Turing complete, which is just as bad

    You jest, but... 

    There - fixed it for you. 



  • @Random832 said:

    @phaedrus said:

    However, that would mean that sed is Turing complete, which is just as bad

    You jest, but... 

    And here's the actual proof: an implementation of a Turing machine in sed.



  • @asuffield said:

    @Random832 said:

    @phaedrus said:

    However, that would mean that sed is Turing complete, which is just as bad

    You jest, but... 

    And here's the actual proof: an implementation of a Turing machine in sed.

    Oh, Intertron!  How did you know?  You'll notice in my post above I included the 'for the love of ${DIETY}' in my call for people to use something else.  Some people actually do write stuff in Brainf*** and the like.  It doesn't matter how kinky or what kind, someone, somewhere is doing it and posting proof on the 'net. 

    Of course, what you haven't seen is someone take sed and write a sed script to compile itself into machine code and then write an operating system using sed. 

    I'm asking to be struck by lightning, aren't I?  One of you is going to post a link...
     



  • @Random832 said:

    @phaedrus said:

    However, that would mean that sed is Turing complete, which is just as bad

    You jest, but... 

    And it's more entertaining than it should be.  z and x move the paddle back and forth.  Hit return for it to frame refresh.
     



  • @phaedrus said:

    Of course, what you haven't seen is someone take sed and write a sed script to compile itself into machine code and then write an operating system using sed. 

    I'm asking to be struck by lightning, aren't I?  One of you is going to post a link...

    There's a toolkit around somewhere to bootstrap a crude C compiler on any Turing machine. Stack that on top of the script I posted and then bootstrap GNU, it's basically the same thing. That's kinda the point of Turing machines.


Log in to reply