Ripple SOUNDS nice but there are some MAJOR problems

[misterbigg]

Looking over the documentation (https://ripple.com/wiki/) there is a lot of information about the protocol that is not explained. There is certainly nowhere nearly enough information to produce a competing implementation. A lot of important details are swept under the rug. For example, in "How it works", it is claimed that "Ledgers are really hash trees". No further explanation is given.

This nonsense is repeated all over the documentation, with promises that "mathematical proofs are coming soon." It is said that the ledger achieves consensus between decentralized nodes without proof of work (a bold claim) and yet, no step by step algorithm is provided. The closest it comes to anything resembling detail is that a transaction is either "passed", "soft failed", or "hard failed." With no other information provided. Answers to important questions, like how Ripple peers are discovered, the messages passed between nodes, who manages the central list of "unique nodes" (for every client's UNL) are totally missing.

The original Bitcoin paper from Satoshi was amazing. It provided enough algorithmic information to implement a baseline protocol yourself, and addressed almost every important issue that one might come up with when first exposed to Bitcoin. On the other hand, Ripple is almost completely opaque and comes to us not in the form of a rigorous paper but a deployed implementation that we are supposed to "just trust." Looking over the Ripple forum and reading some parts of the wiki, it seems that Ripple credits ("XRPs") are distributed by hand by the administrators of the system? In what way is this decentralized?

In the other Ripple thread, people are fawning over this rubbish like it's the second coming of Jesus. I suspect these are non technical individuals who have fallen in love with the idea (which is decent) but don't have any inkling of whether or not it can work at the technical level.

[HostFat]

It's in beta and so it isn't complete.
You should wait the release here: https://github.com/rippleFoundation/ripple

[misterbigg]

It's in beta and so it isn't complete.
Wait for this: https://github.com/rippleFoundation/ripple

Yeah nice job stating the obvious. Bitcoin is also in beta and yet we know about every intricate detail.

[cypherdoc]

i have to agree with OP altho i base this off the info gleaned from the Ripple website and haven't bothered to hunt for more detailed info.

i don't get the concept and the lack of granular detail MBigg is complaining about seems valid.

[Herodes]

Agree with OP. I also think there lacked some clear cut info about how it actually works on a technical level.

[D.H.]

It's in beta and so it isn't complete.
Wait for this: https://github.com/rippleFoundation/ripple

Yeah nice job stating the obvious. Bitcoin is also in beta and yet we know about every intricate detail.

If you wanna make a fair comparison you should compare it to how complete the documentation of Bitcoin was in 2009.

[Intueri]

Anytime i feel smart around normal people, i come on here and get brought back down to earth by the sheer amount of brilliance that occupies this forum.

Never ceases to amaze me. Always so far above my head lol

[misterbigg]

If you wanna make a fair comparison you should compare it to how complete the documentation of Bitcoin was in 2009.

Agreed. Here's the original paper, published in 2008:

Bitcoin: A Peer-to-Peer Electronic Cash System

Every revolutionary idea is explained, along with an analysis. This is completely absent for Ripple.

[D.H.]

If you wanna make a fair comparison you should compare it to how complete the documentation of Bitcoin was in 2009.

Agreed. Here's the original paper, published in 2008:

Bitcoin: A Peer-to-Peer Electronic Cash System

Every revolutionary idea is explained, along with an analysis. This is completely absent for Ripple.

Yeah, well, it doesn't contain "every intricate detail" and it isn't "enough information to produce a competing implementation".

[cypherdoc]

in the ECB paper on Bitcoin, which alt currency platform was the one that allowed selective "inflation" of the money supply by its owners?

[misterbigg]

Yeah, well, it doesn't contain "every intricate detail" and it isn't "enough information to produce a competing implementation".

I disagree. There's enough there to where you can create your own client that functions similarly to Bitcoin. It might not have binary compatibility (i.e. the messages passed between peers may not be compatible, and the format of the blocks would probably be different). But all of the "difficult" problems are explained in the paper and this is the guts of Bitcoin.

Compare this with Ripple, where every "difficult" problem is just swept under the rug with a non-explanation in the wiki.

[HostFat]

I think that the simpler answer is that they are probably trying to avoid that someone come up with a competing implementation/clone before of their own product. (if you want to speculate on the "bad" view)
It's probably a good idea, and they want to be sure to take this part of the market before anyone else.
When they will be sure that the product is secure from bug/hacks and they have a good marketplace, they will release all sourcecode for servers/notes and it will become completely decentralized.
This can already happen in the next few weeks.

Anyway, this is just speculation now, but if someone is already jealous it is his own problem.

[D.H.]

Yeah, well, it doesn't contain "every intricate detail" and it isn't "enough information to produce a competing implementation".

I disagree. There's enough there to where you can create your own client that functions similarly to Bitcoin. It might not have binary compatibility (i.e. the messages passed between peers may not be compatible, and the format of the blocks would probably be different). But all of the "difficult" problems are explained in the paper and this is the guts of Bitcoin.

Compare this with Ripple, where every "difficult" problem is just swept under the rug with a non-explanation in the wiki.

OK, I get your point. I still think that you overreact a lot when saying that is has "MAJOR problems" (or as in the original subject "flawed implementation"). So yes, don't put all your money on Ripple right now. Wait for a bit, make sure that they release the code as promised, ask questions about the protocol and have them update the wiki.

[misterbigg]

I still think that you overreact a lot when saying that is has "MAJOR problems"

Did you know that there is a hard limit of only 500 million accounts? And that once this limit is reached it is no longer possible to create transactions? Even with the little documentation that is there, this fact can be factored out:

- Limit of 100 billion Ripple credits
- Required to have 200 Ripple credit balance in an account
- Transactions require Ripple credits

Hardly an insignificant issue.

[markm]

I got the impression that it conveyed to me a decent gist of how it does consensus, except that it did not seem really clear on how to deal with someone trying to instantly doublespend such as by connecting to two nodes very distant from each other in connectivity telling each one a different place to spend the same coins to.

Thinking about it it seems to me it could be useful to flag the account doing the sends as somehow corrupt or problematic rather than to penalise other nodes for happening to prefer the spend they heard about first; otherwise both nodes might end up thinking they have the right one and the other has a wrong one.

So yeah I can see that some detail seems to be missing. I don't remember seeing anything about flagging accounts for such things, just about whether to trust nodes about items nodes have heard of.

I also agree that the desire to have their personal 20 billion ripples aquire value and their company's 80 billion ripples (minus however many it gives away) aquire value does seem decent motive for trying to establish "network effect" before all the altripples pop up.

As to the limit of 500 million accounts, that would only be if reserve requirement per account stayed at 200 ripples per account forever. Aren't reserve requirements, like transaction fees, modifiable?

-MarkM-

[misterbigg]

Aren't reserve requirements, like transaction fees, modifiable?

Exactly, these important details are not explained. I sure hope reserve requirements aren't modifiable or else they could just be lowered to zero and then they would serve no purpose to prevent spam.

[cypherdoc]

i would think that the proof of work principle would need to be applied to Ripple to encourage correct behavior.

[HostFat]

XRP aren't an asset as Bitcoin, their main function is to avoid spam on the network.
I won't be surprised if there will be the possibility to have more than 100 billion XRP in the future if they are needed.

[D.H.]

I still think that you overreact a lot when saying that is has "MAJOR problems"

Did you know that there is a hard limit of only 500 million accounts?

No, I didn't. And I don't think that you know that either since, as you pointed out, not all details are explained. It's just that my initial reaction to this is not to start alarming topics but to ask them to clarify things. They are Bitcoiners you know, they are around for questions.

[D.H.]

XRP aren't an asset as Bitcoin, their main function is to avoid spam on the network.

Regardless of their original purpose it seems to me that they have the all necessary properties to be an asset like Bitcoin.

I won't be surprised if there will be the possibility to have more than 100 billion XRP in the future if they are needed.

Is this pure speculation or is there some reasoning behind it? I thought that the number of existing XRP would be in the "genesis ledger" and not possible to change just like that.