Re: Proposal for a standard way of defining custom operators in Lua

[Lorenzo Donati <lorenzodonatibz@...>]

On 25/07/2013 14.03, steve donovan wrote:
> On Thu, Jul 25, 2013 at 1:51 PM, Lorenzo Donati
> <lorenzodonatibz@tiscali.it <mailto:lorenzodonatibz@tiscali.it>> wrote:
>
>
>     Yes, choosing good names is more an art than a science, but choosing
>     good symbols sometimes is akin to witchcraft!
>
>
> Oh yes!  Where overloading does make sense is implementing generalized
> arithemetic operations (e.g. classic 'operations on a field' + - * / ^
> with corresponding 'zero' and 'one' meanings) and there we have all the
> operators we need.

Yes, really what I also had in mind.

> A (great) exception is LPeg which uses operators to create a powerful
> notation.

Although I know LPeg only superficially (yet - I hope ;-), still its 
notation mimicks strongly those already established in the CS field (if 
I'm not mistaken), so this wouldn't contradict my point: operator 
overloading is great when you already know an operational notation and 
the underlying semantics (fields operation, EBNF, etc.)


> Even in C++, the standard library makes very sparing use of operator
> overloading, + for strings, << & >> for stream reading/writing, [] for maps.
>
> I agree with Lorenzo that otherwise an impressive string of symbols is
> an attempt to make 'code' (source) look like 'code' (a secret cipher).
> Not a good fit for a DSL intended for non-professional programmers!


--
()  ascii ribbon campaign - against html e-mail
/\  www.asciiribbon.org   - against proprietary attachments