Huge -1 on this, %% as an integer division operator is totally unclear.

It's one of the few sane syntax choices IMO, I'd prefer that over \\, div or intdiv.

Although it has no potential for name clashes, I agree that with % being the remainder, having %% return the division is unintuitive at best.

Sure, but I can't think of a better choice. :/

I don't like %% too much either. Also, what about gmp? I don't see this patch updating gmp to support the new operator.

Also, tokenizer tests are unhappy.

On 16 Jul 2014, at 18:36, Stanislav Malyshev notifications@github.com wrote:

I don't like %% too much either. Also, what about gmp? I don't see this patch updating gmp to support the new operator.

Somehow I managed to think of doing it for my bigints patch but not for the gmp extension, I’ll do that, my bad.

Andrea Faulds
http://ajf.me/

@smalyshev I'm unable to reproduce the tokeniser issue on my machine.

@TazeTSchnitzel still not the reason to break CI :)

Sure... I'll try and figure out what's up.

@smalyshev Fixed it. It "didn't happen on my machine" because I compiled with --disable-all and somehow missed that the test was in ext/tokenizer.

@datibbaw I'd argue it does make some sort of sense, it's just rather backwards: % returns the remainder, %% returns the quotient.

If only we didn't use \ for namespaces, we could use it here like some other languages do...

Yeah, Visual Basic does that and it's a good choice. // is used by Python. Maybe we could use \\? (Is that a terrible or a brilliant idea?)

No, \ is already used in namespaces, we can not use it without creating an ambiguity.

Right, but would the double-backslash work? Can you have a blank namespace?

You can not use \ for namespace, but I don't think it is a good idea to reuse namespace syntax for completely unrelated operator.

Yeah, that's the problem. %% is weird but I think it's the most sensible (and I suppose it makes sense as a % counterpart in a twisted way) of the available options.

I don't like this %%. ........... -1

RFC was rejected, so this request is closed.