Quote from: minerva on March 24, 2014, 07:24:56 AM

Quote from: vandokkum on March 19, 2014, 08:22:03 PM

Just do it.

well, in the past week I've downloaded winmerge and grepWin. I can't say I'm familiar with C but I can read it. (I'm a typical altcoin developer from the looks of it, not that Bitcoin or Dogecoin devs made multiple bad commits in the past)

It appears in order to create a new coin, I'll have to change various checkpoint and genesis block parameters around. Uncertain fully of the tasks required. I will likely release the code for public comment before releasing the client, and to counteract anyone from premining, I'll probably mine the first block and change the genesis block in the client (as well as the magic number).

I'm altering the original post in the next thirty minutes to simplify some things, and to explain my thoughts on hashes/pseudorandom number generators, as well as future plans with the coin.

Quote

I don't think any of the devs realise what the release of ASICs is going to do.

ASICs are a mixed blessing, they increase and decrease network security, they increase barrier to entry, and threaten the cryptocoin with a 51% attack by a sufficiently wealthy entity. Problem with litecoin is they had specific values for scrypt from the very beginning. Changing it will wreak havoc and encourage secret ASIC mining, unless the plan for changes is a year ahead of time and done in a slow transparent process.

Whatever formulas I make up, they need to be close enough to being adequate or correct now or before market capitalization reaches dogecoin proportions.

Naturally I welcome any help, but I'm not quite sure what I should specifically request.

Sounds like a fairly reasonable plan. Having experienced the DRK hardfork well after release, it was a bit of a mess and caused quite a few miners and users problems.

I'm not sure that every 660 blocks is small enough for your retarget - it's important to protect against multipool hopping, or whale attack. That can screw all your miners very quickly. All it takes is somebody to point a botnet at a pool and you've suddenly got 100MH+ that will most likely disappear as soon as the diff goes up, leaving all the rest of the miners with a stupidly high diff until the next 660 blocks have passed. You may want to look at DarkGravityWave (WIP, but looks almost 100%), or HeavyCoin's temporal retargeting. Both are max one-block retargets, and HVC's even starts retargeting before the previous block is finished. Better to get that stuff in before you start - hard forking later is a painful experience.

I have a fair bit of experience in C, C++ and Qt (amongst others) so hopefully will be able to provide some kind of sensible help for you. Fire me a message when you get a better idea of what's going to be required and I'll see where my skills fit best.