Comments on: Some Disadvantages of Typing a Stack Language http://cdiggins.com/2007/01/21/some-disadvantages-of-typing-a-stack-language/ Christopher Diggins Fri, 16 May 2008 04:51:12 +0000 http://wordpress.org/?v=2.5.1 By: spc476 http://cdiggins.com/2007/01/21/some-disadvantages-of-typing-a-stack-language/#comment-19 spc476 Fri, 26 Jan 2007 05:35:15 +0000 http://cdiggins.com/2007/01/21/some-disadvantages-of-typing-a-stack-language/#comment-19 There are always pathological cases, such as: 1 5 do 5 random dup 1 = if drop "hello" else dup 2 = if drop 2.2 else dup 3 = if drop true else dup 4 = if drop "foo.txt" openfile else drop nil then loop 3 peek Third element on the stack, but no idea what could be there until runtime. There are always pathological cases, such as:

1 5 do 5 random dup 1 = if drop “hello” else dup 2 = if drop 2.2 else dup 3 = if drop true else dup 4 = if drop “foo.txt” openfile else drop nil then loop 3 peek

Third element on the stack, but no idea what could be there until runtime.

]]> By: sam http://cdiggins.com/2007/01/21/some-disadvantages-of-typing-a-stack-language/#comment-18 sam Tue, 23 Jan 2007 08:29:45 +0000 http://cdiggins.com/2007/01/21/some-disadvantages-of-typing-a-stack-language/#comment-18 Now I remember. I was thinking about Oleg's paper because I got there from a link discussing dependent types. Like in Epigram (http://en.wikipedia.org/wiki/Epigram_programming_language) Now I remember. I was thinking about Oleg’s paper because I got there from a link discussing dependent types. Like in Epigram (http://en.wikipedia.org/wiki/Epigram_programming_language)

]]> By: cdiggins http://cdiggins.com/2007/01/21/some-disadvantages-of-typing-a-stack-language/#comment-17 cdiggins Mon, 22 Jan 2007 17:14:36 +0000 http://cdiggins.com/2007/01/21/some-disadvantages-of-typing-a-stack-language/#comment-17 Hi Sam, I think your WWHD approach is solid. As for Oleg's paper, it describes an interesting feature (type parameterization using integers) which I am interested in. Right now, I don't see a direct link between it and the problem I describe, but there are some links between the two problems. I will definitely give this some more thought. Thanks for sharing your thoughts! Hi Sam,

I think your WWHD approach is solid. As for Oleg’s paper, it describes an interesting feature (type parameterization using integers) which I am interested in. Right now, I don’t see a direct link between it and the problem I describe, but there are some links between the two problems. I will definitely give this some more thought.

Thanks for sharing your thoughts!

]]> By: sam http://cdiggins.com/2007/01/21/some-disadvantages-of-typing-a-stack-language/#comment-16 sam Mon, 22 Jan 2007 08:55:31 +0000 http://cdiggins.com/2007/01/21/some-disadvantages-of-typing-a-stack-language/#comment-16 I have been spending a bit of time reading your various writings on Cat, and often, when I try to figure out some weirder bits, I think WWHD (What Would Haskell Do?) For this I initially thought of using a list of the Dynamic type as the stack. But that is so obviously gross that I only mention it because it because it is such an abuse of Haskell to be laughable, and it totally misses the point of what you're attempting anyway. In the process of thinking of how bad my initial idea was I remembered coming across http://okmij.org/ftp/Haskell/number-parameterized-types.html . Which may or may not be 100% applicable but sounds like the road you're going down. (Note: I haven't quite read Oleg's paper yet, so I could be totally wrong.) I have been spending a bit of time reading your various writings on Cat, and often, when I try to figure out some weirder bits, I think WWHD (What Would Haskell Do?)

For this I initially thought of using a list of the Dynamic type as the stack. But that is so obviously gross that I only mention it because it because it is such an abuse of Haskell to be laughable, and it totally misses the point of what you’re attempting anyway.

In the process of thinking of how bad my initial idea was I remembered coming across http://okmij.org/ftp/Haskell/number-parameterized-types.html . Which may or may not be 100% applicable but sounds like the road you’re going down.

(Note: I haven’t quite read Oleg’s paper yet, so I could be totally wrong.)

]]>