Re: [ANN] Lua 5.2.0 (work3) now available

[Tobias Kieslich <tobias@...>]

On Thu, 20 May 2010, Alexander Gladysh wrote:

> That being said, my strong opinion is that if there would be
> deprecated version of ipairs in the 5.2, it must be 5.1 ipairs, not a
> new variant. No sense in requiring user's to adapt their code twice.

I agree. Same behaviour or get rid of it. The only place I'd see it
being useful was the previously implemented __ipairs metamethod. While
there a re straightforward replacements with __len and __index, it did
open a few nice backdoors (which all could be declared as feature abuse)

so the numeric loop is fine with me.

	-T