Topic: Ripple or Bitcoin - page 41. (Read 34110 times)

The alt-coins may not be technically broken, but the market has lost faith. People no longer move to alt-coins when Bitcoin price buckles.
I doubt the market will pump up any more alt-PoW coins in a post-ASIC world. For a new idea to succeed it's going to have to be something major. Rightfully so, Bitcoin and Ripple have all the forward momentum.

Right. And, to be clear, I'm not arguing that means that these other coins are fundamentally broken or anything like that. I am saying that their attempt to "fix Bitcoin" has produced a system that is in fact worse. I do agree that there might be a small advantage in that mining may be more distributed.

OK. Your argument is a bit confusing, because you keep using a qualitative term (technological) in order to convey an advantage; and you convey agreement with my argument that the botnet would be up against the same amount of proof-of-work. I see where my reasoning went wrong: ASICs do the equivalent work for less cost, so an ASIC-based network can reach a higher combined proof-of-work while remaining equivalently profitable as the CPU-based network as a whole. This in turn gives the botnet a lower percentage of the proof-of-work of the entire existing network. Understood.

For every person who mines on a CPU just for the heck of it, there's an equal and opposite professional miner who is discouraged by the increased difficulty. So while the identity of miners might be more diverse, the proof of work will be the same. However, to mine on a level technological playing field for Bitcoin will mean you have to invest in ASICs. Anyone who tries to attack Bitcoin without ASICs won't even get out of the door.

The point is that ASICs do *much* more work, leaving anyone who tries to mine without them at a massive disadvantage.

No. To compete on a level playing field with miners, someone attacking Bitcoin has to invest in ASICs, otherwise they're at a massive technological disadvantage.

Yes, the total amount of proof of work will be the same in both cases. You won't have more mining because you don't need to invest in ASICs, you'll just have more casual miners and thus fewer professional miners.

If you really don't see it, consider two coins with exactly the same money spent on mining except one uses ASICs and one uses CPUs. Which one can you more easily attack with a bot net? Which one can you more easily attack with rented hardware?

Bitcoin got this right and people who don't understand security are mindlessly trying to fix what is not broken.

Well, the theory is that encouraging non-specialized computing encourages a larger number of miners, so that quantity offsets quality. You seem to suggest that total proof-of-work can reach higher hights through specialization? Since network power is largely determined by profitability, which is largely determined by difficulty, which is inversely proportional to combined proof-of-work of the network, I would think that the amount of proof-of-work will reach the same levels regardless of whether specialized mining is allowed or not; in that case, the specialized mining gains no advantage against botnets, because the botnets are still fighting against the same amount of proof-of-work, whether mining is specialized or not. In these other "scrypt" coins, the decreased proof-of-work of the asics and gpus is just made-up-for by an increased amount of cpus. The bot-net still has the same amount of proof-of-work to fight against, which is mostly determined by the mining rewards and the value of the currency. Ultimately, the theory is that you have same proof-of-work, dictated by same factors of profitability, but distributed more widely.

I agree that this is a threat that has to be kept in mind. I'm not convinced that it's a more serious realistic risk than the threat of a 51% attack on a proof-of-work based system.

This is very bad. Bitcoin got this right. These other schemes are much weaker than Bitcoin's. To attack Bitcoin, you have to invest millions of dollars on mining ASICs that will become worthless if your attack succeeds. To attack these other coins, you just need a botnet or to rent some general-purpose computing power.

I think for the reasons stated above by me, a crypto-currency would be even more resilient if the proof-of-work is designed in such a way the encourages mining to be spread out among many nodes (ie. scrypt based proof-of-work). Many bitcoiners are very aware of this; hence the birth of litecoin, novacoin, yacoin... Ripple is going in the opposite direction...

So to summarize, I think having miners who are rewarded for verifying and recording transactions is a much more resilient system than having voluntary servers consenting with one another. Yes, in Bitcoin running a client is voluntary, and as the # of transactions increase, likely there will be fewer and fewer people tracking them all. Likewise with Ripple, you will likely end up with most people running light-weight clients or using web interfaces. So the fact that your servers are voluntarily run is a weakness. It means the entire Ripple network will be weaker than the Bitcoin network... easier to attack... and entities running these servers will likely be organizations that profit from running public Ripple services, so they will also be sitting ducks; contrary to bitcoin mining which is incentive in and of itself. An entity could have incentive to mine without having any other interests in bitcoin.

I think this is a weak point in Ripple, for the very reason that the Ripple server is analogous with a Bitcoin client. Problem being that the Bitcoin client also has a real risk of decreasing to only a few clients (that actually track all transactions), but that will not affect the security of Bitcoin, because there will still be competition among miners; although there is the risk of mining power being concentrated.... Ripple relies on the server in order to secure the network through consensus. If there ends up being only a few servers, your Ripple ledger is under a much higher security risk than the network of Bitcoin miners is. If Bitcoin is reduced to only a few Bitcoin clients that track the entire block-chain, and if those few bitcoin clients are hacked, it is no big deal because the ledger can not be tampered with, people only use Bitcoin clients to view transactions... clients are read-only. Ultimately, I see Ripple as having most people using light-weight Ripple clients, and major players/investors running Ripple servers. Those few Ripple servers will be the heart of the Ripple network. Will you rely on a trust model? Where Ripple clients must connect only to trusted servers? If you do not do this, then what is the risk that someone might connect to a rogue server? This problem doesn't exist in Bitcoin. I see Ripple having a network topology where ripple clients are forced to rely on trusting certain servers. This is a big drawback. Lightweight bitcoin clients would only need to trust certain full-chain clients for actually reading the blockchain, they would not need to trust such servers to ensure their transactions are correctly written. So I see the weakness in Ripple being that the server (analogous to the Bitcoin client as you described), is also responsible for writing transactions to the ledger. Bitcoin doesn't have this problem of its clients being responsible for verifying the ledger, we leave that to the miners, who compete vigorously for the profits, and watch each other like hawks.

Yeah

There are 3 things at play here:

1) Ripple the concept
2) Ripple the opencoin implementation, "rippled"
3) XRP the currency of (2).

(1) It's a good system which some people will love, some people will hate, it compliments other currencies at it's core, but it competes with many financial systems and financial services offered by companies.
(2) Who knows, it's not generally opensource yet, available on request / to gateways, I know people with the source and they think it does the job but the code could be better quality (poor as in coding standards, not poor as in vulnerable)
(3) XRP is just a currency which is a utility of (2), it can be valued in different ways, either at face value, at what people are willing to pay, and as a utility of the network where 1 XRP is the value of 10k-100k transactions on the public ledger.

Ultimately, (1) and (3) are just different approaches.

(2) is the only real talking point, and time will tell, but it's worth noting the route they are taking and that they are dealing with businesses and investors, as a company they are transparent, even if the software currently isn't.

BITCOIN!

That's a safeguard mearure for mental sanity.


There is a saying:

"Be careful when you argue with a zealot, because he might be right"

Well if you aren't sure an accusation is justified why make it?

good points both of you.

And in addition, I want to point out that open source implies a mindset.


Thus, to summarise.

• the people of Opencoin do not want to share their ingineering work with outsiders
• the people of Opencoin do not want to get feedback and corrections from other people scrutinising their work
• the people of Opencoin do not want their creation to be judged based on facts, but prefer to sell it through a marketing deparment
• the people of Opencoin think it is better to create an elitist group who knows how things are done right, and the users are just to consume their final product, while a marketing department creates the necessary Spin.

These are clearly telltale signs which indicate we should be cautions with anything coming from Opencoin.
Please note, this might be an unjustified accusition.

Which all could have been avoided by just publishing a f**cking Git.

Ok what is it then? Closed source/proprietary is when there is a binary but no source code right?
 
Then I am glad it isn't up to you, what a strange, strange world that would be.

You clearly have no idea what "Open Source" means.

"Open source in future" is NOT Open Source. Period.

If it were for me to decide, you would be marked as scammer together with Ripple for doing PR for them.

They shouldn't claim that the source is "open" until they've actually opened it. They could say "we plan on releasing this software under an open-source licence", but to call it "open" in their marketing material is clearly false and misleading, since it is not yet open.

Closed source does imply the binary is otherwise available, which it is not. If you are just using the client you are not a user. Users are those who run a node, and those get access to their github.

I don't know what their reasons are to make it not all public immediately, but I can't see anything that would lead me to the conclusion that it's fraud. It might be, but claiming that it is is so far stretched that it requires serious mental gymnastics to get there.

What a strange logic.

Please just publish the Git url to pull all code form. All of it.

Until this does happen, it's closed source.
And the creators or ripple get the smell of fraudsters for using the positive connotation of "open source".

Being an open source developer myself, I can tell you: there is absolutely not the slightest practical problem to publish every line of code right away. It does not "help" you not to publish. The contrary is true. Publishing helps spotting your own "group think".