Quote from: TelegramSDice on March 22, 2018, 06:33:00 PM

TELEGRAM DICE

[...]

Visit https://telegram-dice.com for faucet and live data.

[...]

Provably fair
🤖
The random number is simply created from last Bitcoin transaction , we extract the last 4 digits from last Bitcoin transaction hash
Bets are instantly verifiable

The described system is premised on a severe misunderstanding of how Bitcoin works, coupled with incorrect claims about provably fair systems. It reeks of pure snakeoil.

Foremost, learn the first rule of applied cryptography: Almost anybody can design a system which he himself cannot break. That appears to be what youve done hereassuming good faith on your part: You made a system which you dont know how to break, then assumed that it could not be broken by people who have far more knowledge than you do.

As RGBKey correctly states, there is no such thing as a last Bitcoin transaction. The whole purpose of the Bitcoin mining system is to create a Byzantine fault-tolerant ordering where no order otherwise exists. If you are drawing off the last transaction which your nodes mempool happened to see (or you claim to), then you can easily cheat. Furthermore, an attacker could easily influence your mempool. You are just waiting for some smart person to clean out your funds!

Forget about using transaction hashes: Even a block ID is not secure for this purpose, since it can be influenced by miners. (I should add: The hashes of confirmed transactions within a block are also easily influenced by miners, since a miner is the one who chooses which transactions to include in the block.)

Quote from: gmaxwell on February 09, 2018, 08:57:28 AM

The hash the last block's ID approach can be biased by miners. Without knowing what the the result would be used for you can't argue that they wouldn't do it... if they could make themselves win a 100 BTC lottery for sure, ... it would be totally reasonable to orphan and throw out blocks to pull it off. The earlier proposal to use "the last 64 blocks" doesn't help, the last block is sufficient-- it already commits to all prior blocks anyways.

But that is not the only place where your design is insecure.

You say, we extract the last 4 digits from last Bitcoin transaction hash (boldface in the original). But it would be trivial to create a transaction with fully customized, artificially chosen last 4 digits of its txid! By tinkering with the nonce used for signature generation in a non-Segwit transaction, I can create any last 4 digits I want with an average of only 2³² work. Creating a valid Segwit transaction with predetermined last 4 digits to its txid would be a bit trickier; but off-the-cuff, I can think of a way to do it. Either way, you have created a system which allows anybody to spend some CPU power and fully determine the outcome of a bet.

So, your system has multiple severe security flaws.

Moreover, you destroy your own credibility is when you throw FUD on sites which are actually provably fair.

Quote from: TelegramSDice on March 23, 2018, 07:04:33 PM

the other dice can cheat and change the seed at any time

Wrong. A site which uses a properly designed commit-and-reveal system cannot cheat, and is guaranteed by the laws of mathematics to be not cheating. If a site claims to be provably fair when it isnt (as you are doing), then that is simply fraudulent false advertising.

Quote from: TelegramSDice on March 22, 2018, 10:50:14 PM

I respect your opinion

It is not a matter of opinion in any arbitrary sense. Either a site is in fact provably fair, or it is in fact not. This is a question of mathematics and cryptographynotoriously objective subjects which have no respect for opinions in the colloquial sense of that word.

Quote from: TelegramSDice on March 23, 2018, 09:38:20 AM

So claiming that you know the truth and writing it in bold and very big , is a little snooty coming from somone who are promoting another dice in his banner.

I have no affiliation with any existing dice site, although my PGP keys are paying an awful lot for my signature space. I know that RGBKey is correct in what he says, because I have a technical understanding of how Bitcoin works and also, of how provably fair systems work.

What RGBKey has explained to you is very basic Bitcoin knowledge, plus a dash of Applied Cryptography 101. It is a matter of factual correctness, not of opinions; and your ad hominem argumentation does not change the facts.

To OP, you should not foist such an insecure site on the public with incorrect claims that it is provably fair. It provably isnt. If you want to design your own dice system, then you need to either engage in long study of the secure design and implementation of cryptographic protocols, or hire a competent expert to help you.

To others reading this thread: Avoid this site.