Topic: [UNOFFICIAL] [VNL] Vanillacoin 0.4.1 | Instant ▱ Incentivized ▱ Innovative - page 274. (Read 433405 times)

It launched with consensus algorithm, so it's doesn't matter GPU, FPGA, ASIC, Alien Plasma Super Computer or whatever, they were getting the same amount of coins as a single thread of a crappy CPU. I know, you don't like Wolf0, but read history of VNL at least.

Can you provide more details? 

VNL was launched as cpu only. They were using the gpu kernel(s) long before it was made public.

You have been duped, scammed, ripped off..

Of course you're aware - I do believe smolen wanted to alert the rest of us!


That depends - if you mean this one (NSFW): https://ottrbutt.com/miner/whirlpoolxwolf-03162015.png -- then no.

I am, however, working on a WhirlpoolX miner I plan to release - I believe I posted about it earlier in this thread. I'm currently having several people test it, and working out the bugs they report.

Its current status is something like this (NSFW): https://ottrbutt.com/miner/whirlpoolxwolfpub-03172015.png

Non-optimal settings, as well as experimental code currently, but it mines. So, working on it.

GPU and FPGA only! While the currently used implementation of Whirlpool looks like hell for an FPGA, with those tables, Whirlpool is actually designed around a cipher they call W in the specification - a cipher similar to Rijndael (AES) with a block size twice the size of Rijndael's, among several other heavy modifications. Just like Rijndael, it can be implemented with tables, just twice the size and twice as many due to the enlarged state. Oddly, this is how EVERYONE EVER seems to do it - even the damned reference implementation uses precomputed tables! However, just like Rijndael, it has a complete form, consisting of four steps: SubBytes, ShiftColumns, MixRows, and AddRoundKey. Thankfully, the key schedule for W is very simple, nothing like Rijndael at all: each round key is created by running the key through a round of the W cipher itself. This is why you may notice in the miner's OpenCL file, there are two full Whirlpool rounds done inside the loop - the first generated the round key from the last round key (or the input key, during the first iteration), that is stored, and another Whirlpool round is executed, this time taking the data as input. Since the precomputed tables only do the SubBytes, ShiftColumns, and MixRows operations, AddRoundKey is done manually, by a simple XOR of the round key with the output of the round.

Anyways, because W can be implemented with ZERO table lookups, this sort of implementation would be nearly ideal for FPGA. You might be thinking about SubBytes, but that 256-byte lookup table can be emulated with logical operations. I know, because I did it. And before you ask, it sucked on GPU.

what problems with sending coins from the wallet to the stock exchange?
sending is not confirmed.

LOL - agreed - He pretty much trolls every coin I follow - and now this one.  I have him on ignore so do that and it is much easier to get the important information.

The first: Actually yes, it's not as well as I wanted it to.
The second: I'm ignoring the fact, that you are a troll and trying my best to answer to you as to a common folk.
The third and the last: VNL is not CPU only coin, so I don't care how many coins Wolf0 has already created or will create in the future, he has no connections with this project.

Either your english isn't very good or you are ignoring the obvious.

The answer to my last question is, ALOT. lol

Just look through Wolf0's posts and count the cpu only coins he and his collaborators have created.

If you are talking about GPU performance in the early days and now, it stays pretty much the same +20-30% maybe (don't know exact numbers), Wolf0 still has not published his optimized kernels yet.

This, like SPR and many others, didn't start that way.

How much VNL was gpu mined before the (deoptimized) kernel became public ?

Nice.  When master troll Hippie-Tech shows up you know things are heatin' up.

Running 24/7 - 754MB so far (in/out) in a month.

uTorrent as a reference 24/7 - 83.17GB (in/out) in a month.

Didn't get the second part, it's not a CPU only coin.

EDIT: uTorrent in past 31 days - 2.6TB, watched wrong section.

Indeed. So unique that it sucked back over 2gb (1.5gb upload/ 0.5gb download) of my bandwidth in less than 6 hours. Thats pretty freakin odd considering how only 41%/ 39mb of the blockchain actually made it to my HDD. WTF is going on here ?

And .. should I be surprised to see the shitalgo King preying on yet another shitalgo cpu only coin ?

Yeah, John got my maximum respect as a developer, rarely see anyone working nonstop on his/her project on a daily basis. Vanillacoin has also a unique codebase, so it's probably more time consuming.

What? Ew, ew, oh God, ew. Thanks for letting us know - I had no idea.

I fully support AGPL as a choice, corporate entities shouldn't just profit from closed source developments that are based on open source projects.

John as always, quite productive and amazing. Thanks for the update.

News 17/03

• backupwallet feature is finished, will be implemented in the next update
• private key import feature is being worked on (importprivkey), will be implemented in the next update
• the visual bug during staking is being monitored and worked on | Update: the cause of this and the stuck transaction bug was found and will be fixed in the next update
• ETA for the next update is less than a week
• Wolf0 will soon release an open source miner for AMD GPUs
• hardfork was successful, exchanges and pools are up to date, please update your personal wallets also in case you didn't yet
• tipbot is live on the IRC channel