--- Log opened Wed Oct 06 00:00:27 2021
06:38 -!- jesseposner [~jesse@c-24-5-105-39.hsd1.ca.comcast.net] has quit [Ping timeout: 250 seconds]
09:47 -!- jesseposner [~jesse@c-24-5-105-39.hsd1.ca.comcast.net] has joined #secp256k1
10:03 < real_or_random_> is there a reason why we don't return after this line? https://github.com/bitcoin-core/secp256k1/blob/master/src/ecmult_gen_impl.h#L172
10:04 -!- ghost43 [~ghost43@gateway/tor-sasl/ghost43] has quit [Quit: Leaving]
10:04 < real_or_random_> it seems pointless to run public constants through a PRG to generate blinding values. blinding won't be effective anyway then
10:05 -!- ghost43 [~ghost43@gateway/tor-sasl/ghost43] has joined #secp256k1
10:06 < sipa> real_or_random_: i think the point is that passing in NULL resets the state to the default (without user-provided entropy)
10:08 < real_or_random_> sipa: right,. but why aren't the default not just the values set in this "then" branch?
10:09 < real_or_random_> we feed them into a PRG first to obtain the defaults
10:09 < real_or_random_> I think this is just wasting time during context generation. 
10:10 < sipa> ah
12:18 < real_or_random_> am I right?
12:20 < sipa> i will check tonight (currently in a car)
13:38 < gmaxwell> real_or_random_: it could be but then you'd create a performance benefit for not using the blinding.  Back when unknowndl blinding could be used, it was also useful to have a static 'blinder'.
13:39 < gmaxwell> seperately, multiple return points are a MISRA violation, and generally in this code base I've only tried to use them for things like the argcheck error handling.
13:39 < gmaxwell> (though the same behavior could be achieved without the multiple return by nesting the if, so that was more of an aside)
--- Log closed Thu Oct 07 00:00:28 2021