Topic: BitShares PTS (formerly ProtoShares) Mandatory Upgrade & Snapshot Announcement - page 48. (Read 218428 times)

Bear in mind difficulty! Back when years were only 365 days long it was easy, sure, but at higher difficulty when years are two to the umpteenth power days long it will be whole lot harder!

First you have to find two to the umpteenth power people to populate the mall with, too!

-MarkM-

+1

I have already described the algorithm as linked in my prior post. If there are any questions, I will answer when asked.

Anyone who is interested can code it and test it. I am not particularly interested to do so, because I am quite happy if bytemaster gets totally dependent on a non-CPU-only PoW. I have studied all the major PoW altcoins and none of them are CPU-only in my analysis. I don't have time to go proving them all, by implementing them all on GPUs. But I do suppose I will pay someone to do that soon (within months), so it benefits me to let these altcoins waste people's time so that they will then believe me when I later pay the someone like the author of cpuminer (if he is willing) to help me.

Indeed in my deep study, I have come to the conclusion that no coin can be CPU-only and have fast verification. Thus this presents a major problem for both verifying a long blockchain and for dealing with denial-of-service attacks (since the rate of determining if a received transaction or block solution is valid thus decreases). It is possible this is exacerbated for BitShares, because they attempt to do so much more within a block with their market clearing of bid/ask, but I would need to study their code to verify.

I would be happy to describe the algorithm for the GPU attack in more detail after I get back from eating dinner.

Any infos when the first exchange gonna accept PTS?

My client was apparently mining (getmininginfo showed 6.x hashes / sec all the time), but no blocks. Decided to restart it and got the message that it was 10 hours behind and downloaded blocks for about 30 seconds. WTF?

Restarted the client and it is 6 behind.

No blocks for 2 1/2 days now.

Having a cheap verify function is a must for a proof of work, otherwise loading the blockchain would take forever. BTW, I have yet to see any evidence or proof of concept of your claims. Ranting is ok, especially in this subforum, but bear in mind that for this very same reason, people in here are used to start taking things seriously only when an evidence is provided, not before.

So could an effective workaround for this algorithm be ;

a HD 5850 ATI hooked up to a PA system ? !

With a mic gaffe taped out the end?

A bad suite, a slick back hairstyle and a game show smile... ??

It shouldn't take that long unless you're incredibly socially awkward.

http://en.wikipedia.org/wiki/Birthday_problem

Yes - imagine going to your local mall and trying to find two people who share a birthday. It would take you a long time, but it could be verified that you succeeded very quickly.

bytecoin said "This new proof of work can require gigabytes of memory to solve, but almost no memory to verify and as a result the ideal ASIC is memory" in a post but didn't give any further details. I imagine it would explain in the white paper but I haven't read it

Thats great ! And great to hear ,  no one really knows why yet , but I think this discussion is healthy.

Its good that we have a focus on the vulnerability aspect this early.

It remains to be seen if there is an implementation. 

5 haha. 2.5 per core.

Went from 24 hours with nothing but stales blocks to 2 blocks mined within an hour on different computers.  Go figure.

New to mining in general. How many confirmations till a transaction is considered mature?

lol how many hpm you getting on that celeron ?

bytemaster apparently designed Momentum PoW and didn't use Scrypt for PoW for precisely that reason, but by making verification faster than finding a hash, I assert he opened a vulnerability that makes it easy to parallelize the finding of a hash and thus not CPU-only. See the link to my claim on the bounty and judge for yourself.

And when (if) someone does implement a faster GPU miner for this coin, please kindly acknowledge I was correct. You don't have to tip me anything, the acknowledgement would be more than enough.

Block 6048, rewards drop by 5%, from now 42.86875 PTS

am totally done with this stupid coin.
long time hashing, no blocks to be found.
Total 8 orphan on i7 980x 12GB, 1 ok.

Also nice to see a proof of work,
but I don't see a proof of payment.

its not worth it,
good luck with it

Genuine question about the protocol, correct me if I'm wrong anywhere: Bitcoin works well because it takes a huge amount of work to find a block, but only takes one hash to verify that a block is valid. So a wallet can verify thousands of blocks fairly quickly. From what I understand Protoshares works by the same concept, except one Protoshares hash takes many orders of magnitude more work to perform. Does the Protoshares wallet need to perform one hash to verify each block in the network? Seeing how most computers get ~5 hashes per min, wouldn't that take an outrageously long time to verify the block chain once it gets bigger?

I know very well you need fast verification time. I don't need to read your mind, I understand the design of a coin and what the tradeoffs are.


I gave you insight. You and I agreed from the start you would pay me whatever you felt was justified. I have explained a vulnerability. If you feel that is worth 0 and to be ignored, that is your prerogative.

However it doesn't justify calling me a troll and bringing up your spin on past events to deflect attention away from what I revealed.

You changed this from a technical discussion into a character assassination.


If you feel my insight is worth 0, then so be it. The community is observing.

I don't need the money. I wanted to help you recognize your weakness early on.


[redacted by Anonymint, as much I would like to return his accusation in kind given what I've been told, it is not appropriate]


I thought we were doing software engineering, not politics on this bounty.

Please proceed. I am done with you.