Comments on: Holy Grail: The quest for THE programming language http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/ Computational Biology, Epidemiology, Infectious diseases, Open Science Mon, 02 Nov 2009 14:07:57 +0000 http://wordpress.org/?v=2.8.5 hourly 1 By: Around the web - March 8, 2008 : business|bytes|genes|molecules http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/comment-page-1/#comment-4348 Around the web - March 8, 2008 : business|bytes|genes|molecules Sun, 09 Mar 2008 06:42:47 +0000 http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/#comment-4348 [...] Tiago searches for the Holy Grain of programming languages [...] [...] Tiago searches for the Holy Grain of programming languages [...]

]]> By: tiago http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/comment-page-1/#comment-3947 tiago Mon, 18 Feb 2008 17:57:47 +0000 http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/#comment-3947 Joel, I would (and I like Mathematica)... if it was free. Joel,

I would (and I like Mathematica)… if it was free.

]]> By: Joel Klein http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/comment-page-1/#comment-3946 Joel Klein Mon, 18 Feb 2008 16:47:46 +0000 http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/#comment-3946 Consider Mathematica: coming from the Lisp family, it has the "everything is an expression" concept that allows meta-programming, and its pattern language allows Prolog-type backtracking. With a gazillion mathematical symbols and forms it allows you define your own interpretations for nicely-typeset domain-specific notations. It has J/Link for interfacing with Java (see the Wolfram Workbench as a decidedly non-toy example of that interface), as well as language bindings for C, .NET, and Python. Of course it has graphics plotting libraries. Bio* libraries, I'm not sure, but it's worth taking a look. This is coming from a Wolfram employee who is a programming language buff. Consider Mathematica: coming from the Lisp family, it has the “everything is an expression” concept that allows meta-programming, and its pattern language allows Prolog-type backtracking. With a gazillion mathematical symbols and forms it allows you define your own interpretations for nicely-typeset domain-specific notations. It has J/Link for interfacing with Java (see the Wolfram Workbench as a decidedly non-toy example of that interface), as well as language bindings for C, .NET, and Python. Of course it has graphics plotting libraries. Bio* libraries, I’m not sure, but it’s worth taking a look. This is coming from a Wolfram employee who is a programming language buff.

]]> By: tiago http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/comment-page-1/#comment-3928 tiago Sun, 17 Feb 2008 18:59:29 +0000 http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/#comment-3928 Raphaël, I am (very pleasantly) impressed with Groovy, but if I find a roadblock JRuby is my plan B, it is good to know that at least some kind of support is expected in the future. For things like auto-completion, a standard mechanism of declaring parameters is important, also for debugging and maintenace, IMHO. Jyha-Pekka, I do believe that having a "Turing complete" language is fundamental, especially in an environment where not all is know about the problem domain and extensibility and knowledge upgrade is fundamental. It is not clear for me that making a generator is easier (I have actually done a couple of them, for web form and html generation - nothing fancy) than an embedded DSL, as long as the language underneath helps. Raphaël,

I am (very pleasantly) impressed with Groovy, but if I find a roadblock JRuby is my plan B, it is good to know that at least some kind of support is expected in the future. For things like auto-completion, a standard mechanism of declaring parameters is important, also for debugging and maintenace, IMHO.

Jyha-Pekka,

I do believe that having a “Turing complete” language is fundamental, especially in an environment where not all is know about the problem domain and extensibility and knowledge upgrade is fundamental. It is not clear for me that making a generator is easier (I have actually done a couple of them, for web form and html generation – nothing fancy) than an embedded DSL, as long as the language underneath helps.

]]> By: Juha-Pekka Tolvanen http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/comment-page-1/#comment-3919 Juha-Pekka Tolvanen Sun, 17 Feb 2008 09:51:29 +0000 http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/#comment-3919 If you really want to raise the level of abstraction closer to the problem domain (away from the format compiler/interpreter), perhaps you could consider non-embedded DSLs. This allows you to give language that maps to the problem domain and then use generators (much simpler to make than compilers) to the preferable target language. This gives the natural possibility to map to other implementation languages than just one. Some of the technology alllows you also to satisfy, at least partially, the evolution side: you can change the syntax and semantics of the language without loosing your existing specifications. For cases on variety of domans see www.dsmforum.org/cases.html If you really want to raise the level of abstraction closer to the problem domain (away from the format compiler/interpreter), perhaps you could consider non-embedded DSLs. This allows you to give language that maps to the problem domain and then use generators (much simpler to make than compilers) to the preferable target language. This gives the natural possibility to map to other implementation languages than just one. Some of the technology alllows you also to satisfy, at least partially, the evolution side: you can change the syntax and semantics of the language without loosing your existing specifications. For cases on variety of domans see http://www.dsmforum.org/cases.html

]]> By: Raphaël Valyi http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/comment-page-1/#comment-3916 Raphaël Valyi Sun, 17 Feb 2008 04:55:16 +0000 http://tiago.org/ps/2008/02/16/holy-grail-the-quest-for-the-programming-language/#comment-3916 Hi, the lack of optionnal typing in JRuby is definitely an issue. Still, core devs of JRuby are now working in hard to provide a JRuby to Java typed compiler. Meaning that you would be able to specify Java types using an optionnal syntax within JRuby code. Then, when using this code from the Java language, types will be enforced. How nice is it? I suggest to have a look to this in a few weeks or months. I don't think it's targeted for JRuby 1.1 unfortunately, but soon later hopefully. Cheers, Raphaël Valyi. Hi,

the lack of optionnal typing in JRuby is definitely an issue. Still, core devs of JRuby are now working in hard to provide a JRuby to Java typed compiler. Meaning that you would be able to specify Java types using an optionnal syntax within JRuby code. Then, when using this code from the Java language, types will be enforced. How nice is it?

I suggest to have a look to this in a few weeks or months. I don’t think it’s targeted for JRuby 1.1 unfortunately, but soon later hopefully.

Cheers,

Raphaël Valyi.

]]>