Quote from: Crestington on April 24, 2015, 03:13:03 PM

Quote from: Griffith on April 24, 2015, 02:26:22 PM

Quote from: rainly on April 24, 2015, 09:35:13 AM

we will support ECC forever. Wink

i wish there was an easier way to unit test for staking, but i dont fully understand what the kernel is doing to actually get the result of the scrypt blocks ( i guess i could mimic inputs now that i think about it, damnit) so really what is taking so long is i am syncing the blockchain for the 5th? time to see if my changes will work (50k) blocks to go. and that is what is taking so long.

actually i could do input output testing now that i think about it. why did i not do that earlier...... mistakes were made

Curious what you mean by the unit testing for Staking, please elaborate.

unit testing is one of the basic steps of software engineering/development. you test your code units (functions and methods) with static input data sets to see if it matches the expected outputs. which if it does, means your code works. if it doesnt then you know you messed up somewhere. you are supposed to do this for all units when coding a new program.

for staking what i should have done (and didnt but its kinda too late now) is take the kernel that is used to create blocks out of the wallet code and made it a standalone program. had it accept input from a command line ( which ideally would be a block that has already been generated so that you can compare the output) then pipelined the output to a text file. i would then read the text file and compare it to the block in the chain. if it matched for 5 instances (arbitrary number i chose) then the staking would work and id copy the code back into the wallet.

i didnt do this. partly due to lack of time, partly because i grossly underestimated the difficulty level/possible issues that could possibly arise from the staking portion of the code.