It sounds to be wasting a lot of time proving things about programming, when the spirit of Lisp is usually much more test-driven and not proving things. The stated goal is so that you can go about your normal business, but PROVABLY! I'd say this is much more in the spirit of ML than the spirit of Lisp.
A good read! Tarver writes extremely well. And while it echos the voices of what sounds clearly like a cynic, it's pretty convincing and concerning. Sad thing is those concerned few are usually those who don't play by the rules and you end up with little might. But what the hell, someone's gotta fight.
I don't like the case-sensitive symbols. Nor do I like the implicit quote on symbols...it doesn't make much sense to me but that might because there aren't many good code samples on the site.
Is it too much to ask for a nice looking website with lots of good code samples? And a reference link for all the functions that work in the language would be nice.
As much as some people dislike newLISP, it at least has a good clear/concise website: http://newlisp.org/ The purpose is clearly stated, and right away you know on what it can run and where to look for comparisons to other Lisps.
Not much of a competitor to Arc in my opinion. But pg, please show us some more Arc code :P
I'm no lisper but it seems to me that the implicit quote and the resulting need to use square brackets to delineate lists would make it more difficult to write macros.
It depends on how he's defined the implicit quote syntax. You could make it so it doesn't have any effect in a def-macro call or something but that increases complexity just a tiny bit for a feature that doesn't do much for the developer.
Off-topic-ish: The square brackets are kind of neat but not for that purpose. If instead he allowed the interchange of square brackets and parenthesis, i.e. (+ 2 [ 3 2]), that would be neat and useful...it would allow someone to differentiate between nested lists slightly. And of course, it's easier to click [ instead of Shift+[ to get ( :p
I suspect the need of special syntax for list construction came from the introduction of currying. If you want currying you need to know arity of procedures, so you lose variable arity procedures such as 'list'.
Any competitor is a good thing. It will either cause Arc to move faster, or just be better than it. The pattern matching simplifications seem quite slick.
I agree, ML style pattern matching seems so elegant. My favorite pattern matching system is actually Mathematica. I wish more programming languages had 'em :)
Yes, "bad marketing". It's been possible for years to significantly improve the look and feel at the scripting level, but they never bothered. Also, the fact that it took them so long to figure out that Gnome and KDE were where Unix was headed and get with the times is not a good sign that they have the marketing skills necessary to stay current.
