Re: Proposal: allow @, $, !, and ? in Lua identifiers

[Sean Conner <sean@...>]

It was thus said that the Great Coroutines once stated:
> I'd like to discuss allowing @, $, !, and ? in identifier names.  I
> chose these explicitly because they can't be confused with an operator
> at the end of an identifier.
> 
> >From what I understand, it's common practice in Ruby and Lisp to write
> methods like this if you intend they return a boolean:
> 
> empty?(some_table) -- returns true if some_table is empty
> 
> Usually is_empty() is what I see people use because it's clear that
> the function is testing for emptiness and not emptying the table.
> 
> The problem is you sometimes wind up writing this terrible thing: if
> not is_empty(tbl) then ... end
> 
> Ruby also makes it convention to put ! at the end of the function name
> if if modifies its `self'.  The non-! version returns a modified copy.
> 
> I feel like these 2 added characters alone would make it much easier
> to express semantic in function names.
> 
> @ could be used if a function creates an object that internally
> references the object it was created from, so you'd know whether to
> copy or deep-copy later.
> 
> I'd propose using $ if a function has side-effects, or performs some
> form of I/O -- but I'd love to hear other ideas.  I've never seen $
> used except as an operator in Haskell -- the expression on the
> right-hand side gets evaluated and passed to the expression on its
> left-hand side.  What could it mean as part of a function name?

	function panic!!!!(message)
	  print@(screen.width/2,screen.height/2,message)
	  sys$wait("5s")
	  why_must_i_die?("now")
	end

  -spc (If you want to enforce a convention, make it part of the langauge)