« first day (1045 days earlier)      last day (3951 days later) » 

user20683
1:56 AM
@JimmyHoffa This is why I've no interest in PHP
 
5:48 AM
@YannisRizos Let me know when you get a chance to look over the kind of tailor made parser combinator stuff I did for you, if it makes sense or what you think
 
user20683
6:19 AM
@JimmyHoffa I'm drinking the cool-aid, mono has been installed. MacRuby has also been installed and is vaguely and I mean vaguely working
 
user41796
3:26 PM
0
Q: Does software reuse preclude process repeatability

GlenH7TL;DR - Does the ability to reuse software prevent process enhancements that come from repeating a project? For example, the construction of a bridge relies upon the accumulated knowledge of the bridges that were built before. Does software reuse prevent that accumulation of knowledge? I was...

 
user41796
^^^^ My attempt at bringing in some more "big picture" type questions to the site.
 
4:04 PM
@GlenH7 Good. we need more questions like this.
 
 
3 hours later…
7:26 PM
@WorldEngineer What coolaid is that? Is MacRuby run on mono?
 
user20683
@JimmyHoffa no but C# and F# are
 
user20683
;)
 
Drinking the .NET koolaid?
 
user20683
@JimmyHoffa better tasting than the Java cool-aid
 
I've heard it said that F#'s future is open source, and that the open source F# compilers and such are as quality as MS's
 
user20683
7:28 PM
@JimmyHoffa here's a question for ya, what is the ML equivalent of Scheme?
 
user20683
I did figure out my issue with MacRuby
 
@WorldEngineer ML equivalent of Scheme? Uhh... ML? I don't exactly know where Scheme and LISP are different other than I understand Scheme is a stricter type system, beyond that; (most) LISPs are imperative, is Scheme also? If so, then SML is your standard
 
user20683
Scheme is more minimalist
 
user20683
SML is probably closest
 
7:30 PM
Right, but at it's heart it's pretty much the same outside of the few type system dissimilarties
All ML's have arbitrary depth lexical closures, All MLs have a strict type system (well for the most part; I'm sure some dynamic MLs have been made but I'm sure they're esoteric) All MLs have the HM type system which LISPs don't have (though they can of course use macros to fake them), other then that there's really not a lot of similarity between MLs and LISPs...
MLs take a much more declarative approach "This scenario means that", LISPs take a much more applicative approach, the fact that pattern matching isn't an in-built part of (most) LISPs alone tells you they have different goals
They're both declarative compared to purely imperative languages, but SML (and other MLs) and LISPs are both imperative as well, though LISP has a much more do-this-then-that feel, the tree structured approach is what keeps it declarative at all when MLs put much more focus on the defining of a scenario so you can declare the outcome of that scenario
But yeah, in the ML world, SML is your minimal (So long as you don't count Haskell as an ML)
Haskell is likely the most minimal language I've ever seen without question, the fact that it's so amazing is just an effect of the particular set of features the language allows even though the language itself is as minimal as they get
You've seen the qbasic in Haskell blog I'm sure, creating a Scheme using template-haskell would be stupid easier than even the qbasic one, because Haskell is minimal enough to be a close match in it's own way. You could take a Scheme script, remove the outer parentheses and put parens around each function to force RPN, and it would probably become completely valid haskell
aside from the defun syntax for defining a function
Only thing I don't appreciate about the F# community is there insistence that Haskell is stupid for real world applications, this is a common meme in the F# community which is annoying, Haskellers see F# as a perfectly fine language if you want the .NET framework available but for some reason F# folks are certain Haskellers are just arsehats. But I guess when the monad is strong with you, you get used to people thinking you're off your nut...
 
user20683
I think it's Haskell's learning cliff that makes it difficult
 
Of course
 
user20683
it's the programming language equivalent to EVE Online
 
user20683
massive learning curve but it can ANYTHING
 
user20683
and it's a conceptual learning curve unlike say C++ where it's more of a usage learning curve
 
7:45 PM
Learn F# well and maybe you could get a job with that; that would be an absolutely awesome first gig and a great step towards the data analysis/research type work you want in the eventual
 
user20683
@JimmyHoffa probably be an awesome second job
 
@WorldEngineer yeah, the fact that I was able to wing some parser combinators in PHP for @YannisRizos last night off the top of my head was a clear realization thinking about it that there are conceptual things I had to grasp in climbing Haskell which I'm rather certain most people would find more than foreign.
But that's ok, people don't need to go that route, I've worked with plenty of great engineers and currently work with a couple who I take as wiser than I in many ways who create great designs but have no sense of Haskell. It just adds some tools like the 55-degree-angle-ratchetting-hammer that almost nobody would know what to do with, but aren't necessary. Yannis's brainfritz interpreter and parser is quite good without any of that extra alien nonsense
 
user20683
I find myself understanding objects far better than many I see here just from exposure to Programmers in general
 
Imagine how well you'll learn rather than the smart people being on an online website, but being in a cube right next to you working on code with you every day. That's why joining somewhere with good engineers is key; especially at the start
 
user20683
8:09 PM
@JimmyHoffa
 
user20683
new Haskell tutorial
 
I love the graphical approach, that's one of the coolest things about Haskell people don't know a lot about; Haskellers for some reason tend to be into graphics stuff, as such there's tons of quality libraries for various graphical uses
Which flies directly in the face of everyone with there "Haskell isn't for real application stuff"
 
user20683
@JimmyHoffa xMonad for instance
 
xMonad get's too much air time in my book, it's a classic cathedral project and if you try to use it you'll think haskellers are crazy for starters
 
user20683
8:16 PM
yeah
 
Not that it's not amazing quality and all that, it's just esoteric to the max. I'm referring to the fact that there's a hand full of quality OpenGL libraries for Haskell for doing things at various abstraction levels, and a variety of good libraris over uhh SGT? What's that 2d drawing lib I can't remember the name of? S-something-something, as well as some QT stuff and the SVG libs that tutorial uses etc
Haskell's cross-platform support is also super-star
native everywhere, but the compilers are kept so perfectly in line as well as the packages. There's rarely a slight IO difference related to kernel differences from Win to Lin
 

« first day (1045 days earlier)      last day (3951 days later) »