Luke Palmer wrote:
Rod Adams writes:I see that as preferable to saying "we had it in 5.10, we dropped it in 6.0, then added it back in for 6.2."
C<study> is an odd sort of function. AFAIK, it's the only optimization hint that we have.
Will the P6RE even use this information, and is it worth keeping?
My gut feeling tells me that it will be useful again around 6.2, and we should keep it around until then as a potential no-op.
Ummm... yeah, keep a function around if it's not currently implemented.
I don't think so.
Imagine the guy who has a case where it makes a huge difference, and in porting his code, he wants to make it where when C<study> works again... he uses it. But until then, he has a no-op to hold it's place.
I wouldn't suggest this if C<study> actually changed semantics in any way. Semantically, it already is a no-op. Removing it makes no difference to your output, only in how fast patterns are matched.
-- Rod Adams