 |
lua-l archive |
|
|
||
|
On 10/04/2011 10.45, Dirk Laurie wrote:
On Sun, Apr 10, 2011 at 08:37:25AM +0200, Lorenzo Donati wrote:Yes. I know that. But I avoid like the plauge to redefine standard libraries functions (even in short scopes - it's about readability).Python does the various sorting options neatly: a.sort() sorts a in place, sorted(a) returns a new array containing the sorted version without disturbing the original. OK, Python is a fat API, but that sort of distinction aids clarity. Your argument against Sean's local redefinition of table.sort() does not apply to making a new table.sorted().
Of course that would be a little less intrusive change, but I still refrain to pollute standard namespaces with my tools.
Please bear in mind that I didn't look for an alternative solution, (the solution is rather easy anyway). I, at the risk of being "heretic" :-), just suggested a small change in the API because I think it won't cause any problem and will add clarity (by avoiding needless temporaries) and some flexibility.
I like Lua's approach, so I'm not a fan of language or API changes just because "Java/Python/C++/Snortlescript has it!" :-). The only objection to my proposal (which I think is in Lua's philosophy - but maybe I'm wrong here) was that of Luiz, but as I argued before, I'm not convinced by that alone (maybe there are other technical reasons for that choice?).
I repeat: the simple difference between "in-place" vs. "new table", whereas important, is already stated extremely clearly in the docs and there are almost no sources of confusion with other API functions.
The argument of Luiz, while in general is really soundAs arguments of Luiz and Roberto tend to be :-)
Definitely! Almost all posts of the "dynamic duo" (pun intended! :-) covering technical matters end up tagged as "important" (the ones I understand) or "interesting" (the ones I'd like to have more time to understand! :-) in my mail client archive!
Dirk
cheers -- Lorenzo