Re: concerns about Proc,lambda,block

On Monday 01 March 2004 09:28, David A. Black wrote:
> and building everything else up from that.  I admit I never understood
> why that was rejected.  It seems to involve fewer twists than having
> Proc objects choose their semantics based on their context.

There's a really long thread in comp.lang.functional right now about 
functional programming in non-functional languages(1), in which there's some 
discussion about language features that change their fundamental behavior, or 
interpretation, based on context(2).

There's a very real threat of obfuscation for these language features, and I 
agree with you, David, that these are language pitfalls.  Indeed, I can't see 
how anybody could convincingly argue for their POLS-ness, from any 
perspective.  Some people really like the fact that a single line of code can 
mean and be interpreted in different ways based on the context it is in; in a 
way, this is just an extension of polymorphism.  However, even people who 
support the "worst" form of polymorphism, overloading, agree that it should 
be used carefully -- sort of like goto, guns, and nuclear power, it lends 
itself easily to abuse.

I'd argue that context sensitive language features are a step beyond ad-hoc 
polymorphism (a.k.a overloading).  They provide 80% of the confusing language 
behavior while providing only 20% of the features, and I again agree with 
Dave that Matz should seize the opportunity to simplify this aspect of Ruby 
for Ruby 2.0.




(1) Titled "Functional programming in mainstream languages"
(2) 
http://groups.google.com/groups?dq=&hl=en&lr=&ie=UTF-8&safe=off&threadm=c1urha%2442e%241%40news.oberberg.net&prev=/groups%3Foi%3Ddjq%26as_q%3Dcomp.lang.functional

-- 
### SER   
### Deutsch|Esperanto|Francaise|Linux|XML|Java|Ruby|Aikido|Haskell
### http://www.germane-software.com/~ser  jabber.com:ser  ICQ:83578737 
### GPG: http://www.germane-software.com/~ser/Security/ser_public.gpg