| sphynx | oh, I've got it, soory for stupid questions :) |
| kc5tja | littledan: In fact, the language you'll end up with will look amazingly like Joy or Factor. :) |
| littledan | kc5tja, yeah, but with laziness and typeclasses; that's the gap |
| mutjida | i think continuity is just assumed for most programming language semantics applications |
| sorear | Haskell style is pretty much impossible to implement with anything other than a batch compiler. |
| kc5tja | Ditto for Joy, actually. |
| sorear | Forth is the epitome of nonbatch. |
| kc5tja | To get it to be interactive, you'll need to structure it more like Postscript. |
| monochrom | Yes, continuity is assumed. There is also proof that continuity = bounded nondeterminism. |
| mutjida | i think conitnuity is necessary for fix points. |
| littledan | Factor actually has what amounts to a batch compiler (though the interpreter is used fairly often) I think and, of course, you could redefine : to make a batch compiler in Forth! |
| monochrom | continuity is sufficient, but not necessary. |
| kc5tja | : is already a batch compiler. It just so happens that ; terminates it. :) |
| mutjida | how do you prove existence of fixed points without conitnuity? i mean continuity |
| monochrom | monotonicity is also sufficient. tarski-knaster or something. |
| mutjida | oh yes you're right. monotinicity is necessary. tarski's fixpoint theorem. |
| kc5tja | GAWD, this is so monotonous! ;D |
| monochrom | If f is continuous, it is a bonus, because f^omega(_|_) is the least fixed point. Without continuity, may have to replace omega by an even larger ordinal. |
| kc5tja | (sorry, it was just begging to be said) |
| littledan | where could i read about stuff like the theory behind fixed points? |
| monochrom | I recommend Davey & Priestley "introduction to lattices and order". You can also wait for me to write it up on haskell wiki and haskell wikibook. :) Right now I'm writing a Parsec example instead. |
| sorear | monochrom: even LARGER ordinals? ouch! |
| mutjida | wikipedia has some stuff, but perhaps not introductory. many graduate level programming language texts also cover it (e.g. pierce or mitchell) |
| monochrom | I can make it sound very simple. I just don't have time to write it up. |
| kc5tja | OK, back to my modeling. I was hoping to have new code out by today, but I'm wasting too much time on IRC. |
| mutjida | with just monotinicity, you can still construct a fixed point by starting with bottom.. right? |
| sorear | mutjida: you'll never get close enough |
| monochrom | Yeah, f^(huge ordinal here)(_|_). |
| littledan | monochrom, thanks, Google Books apparently has that in full |
| monochrom | W H A T ??!!! N I C E !!!111 |
| littledan | there's no PDF, though (which would be better) also, I'm too stupid to understand it |