Topic: The Ethereum Paradox - page 52. (Read 99876 times)

If all inputs to a script arrive via one transaction, you must be forced to conclude that there is no difference between said scripting blockchain and an asset transfer chain in terms of equilibrium breakdown?

edit: actually, I see what you're saying - if the script author's cause a cross partition dependency, that dependency can affect the resulting ordering on partition merger. The problem is when this dependency is not visible to the system.

further edit: however, I still struggle to see why this would result in anything more than the failure of the script in question?

More completely stated, the Nash equilibrium is that there is no other strategy other than the strategy of mining on the longest chain which is visible to all nodes, i.e. that there is no superior strategy other than the one that nodes are already doing and which is known to all nodes. Whereas, as I pointed out in my video, when a node (or colluding nodes) have > 33% of the hashrate, then for Satoshi' PoW design they can apply the selfish mining attack by withholding block solutions until the rest of the network catches up, thus the Nash equilibrium is destroyed by selfish mining in that case.

Also I have pointed out in my video and the follow up posts in this thread about a meta issue that destroys the Nash equilibrirum, that for the case where there are external failures (external to the block chain's perspective of itself) due to external actuation of cross-partition state (even if the block chain thinks it is enforcing a strict partitioning with no cross partition state), the Nash equilibrium fails because the entire coin fails, thus the validators of partitions can't trust the validators of other partitions (because although they get their block reward, the external market value of the reward fails). It remains under study whether this applies to asset transfers too (or just to partitioning of scripts) and whether it applies for asset transfers in the strict partitioning block chain (which I argued in my video is immune to the problem) and/or in the cross-partitioning block chain (which I did not address in my video and Fuserleer raised this point hence).

I hope readers don't get confused that I am making a distinction between when cross partitioned state is occurring by-design on the block chain and when it occurs externally because it can. For scripting it is impossible to enforce a strict partitioning because it is very clear that the external actuation can inject state from one partition into another partition (and even though the block chain can't determine this, the external users can and the external users can experience failure that the block chain is entirely unaware of due to this external Turing completeness, which is a very deep, meta, high IQ concept that apparently most people wouldn't think of ... note smooth indicated to me in a PM that he had thought of this issue of external Turing completeness before too). For asset transfers (no scripting), it is not yet 100% clear to me. I need to think about it more.


Yes as long as the state from the two partitions did not leak into each other by any means (including the external meta case mentioned again above).

Another point, I wanted to touch on this idea of 'Turing completeness'. I'm not sure it's entirely helpful to the discussion on Ethereum because, even the computer I am writing this message on cannot be classed as 'turing complete' in the true sense of the definition.

I believe your point is related with script inputs being external to the blockchain? If the inputs to a script must come to that script via a transaction, that internalises the inputs - so I must be missing something?

I've touched on this before, but you've reminded me again; partitions are the antithesis of consensus. Taking things to the extreme is helpful to illustrate the problem: with infinite partitions, in bitcoin, you are left with the DAG of the UTXO set, and no blocks or any agreement on what the order of transactions should be, in other words, no consensus. The LCR in bitcoin constantly forces miners to chose between candidate potential partitions (orphan chains). The nash equilibrium results in rational miners always choosing the longest branch to mine on to maximise their profits.

Talking about partition unification for a moment; if two partitions are totally separate, merging them doesn't have any consequences for ordering because the individual transactions in each partition have been separate from each other, you can order them however you like as long as you obey the  parent/child relationship in each partition.

Well that is what I meant by "such as Iota", and the other case that you are working on doesn't make any sense afaics (but we'll wait to review your white paper when it is published), so that is why I tend to dismiss that there is any other case worth associating with a DAG. Could I be wrong? Perhaps. Doubt it.

And yet you are happy to cross post the same thing in multiple threads at the same time repeatedly. I've seen three separate threads with quoted cross posts from you being updated at the same time. If that's not noise, I don't know what is. If you wanted control over your thread, you should have made it moderated.


A DAG does not inherently say anything about the model used upon it. There is a DAG with deterministic ordering, for example, however, Iota is not that, and that was my mistake.

Good luck with that.  One of the things that made Bitcoin great is consensus via economics that's advantageous for the individual and the group.

That multichain.com article is talking about the lack of parallelization, but as I (and CIYAM and even Vitalik) have pointed out (both upthread and in my video), parallelization can be fixed by simply restricting the inputs of scripts to data from already confirmed blocks (i.e. do not allow scripts in the current block to impact scripts in the current block).

Edit: by parallelization we mean the ability of a validator to validate more than one script simultaneously. Distinguish this from partitioning where we have validators who trust other validators (since not all validators validate all scripts).

That multichain.com article does not address my point about partition boundaries being impossible for scripting. That is my point and I am sure no one else has thought of it. I am the first to make that point. And this is a damning point because it means all validators (full nodes) must verify all scripts. And from that comes the other points I have made about why mining is forced economically to become centralized (and this applies to Bitcoin as well). Thus I have argued don't pretend that crypto currency has decentralized validation (verification) because it can't. And thus I argued just make the validation centralized and try to put the decentralized control else where in the design. Which is what I have contemplated for my design. Which I am working on.

Ethereum could hire me to solve their problem perhaps. I wouldn't charge $18 million.  


Not true in my contemplated design, because the mining is unprofitable.

My gosh I slept 13 hours. I didn't want to wake up except I was so hungry (which isn't good for my gall bladder to go such a long time between meals).

Btw, I spoke on Skype with my friend from Australia (who was in the Philippines with me back in the early 1990s) and he said yesterday that he has gallstones and the doctors want to remove his gall bladder. He is experiencing nausea every day so he is very close to being forced to have the surgery. I urged him to start taking curcumin immediately and hope the stones don't dislodge and hope it will dissolve them over the long-term. Most advice is to not take curcumin if one has gallstones, but some few have recommended small doses over the long-term to try to dissolve the stones without dislodging them (which is incredibly painful and if they are too large to pass through the duct then emergency surgery becomes necessary). I am perhaps dealing with a similar issue (bile duct blockage) or pancreatic cancer (obviously not the fast moving variety which killed Jobs). My condition has improved dramatically since the worst of September 2015, because I started taking specific supplements to replace the glutathione (and B vitamins) I wasn't getting due to lack of proper digestion of food (glutathione is very protective of my brain and neurological effects such as the peripheral neuropathy symptoms I used to have but now which are virtually non-existent). And then I started the high dose (3 - 6 grams per day) curcumin about 5 weeks ago and significantly increased the dose to 20 - 30 grams per day roughly 3 weeks ago, which seems to have radically good effects. My symptoms now have nearly entirely stopped involving my brain and neuropathy and now nearly entirely focused on my gut and sleep. So in spite of being tired (and wishing I could work with full energy always), I am very happy with the improvement and inspired that this might be leading to the cure to this shit I've had for the past years.


In case the video wasn't succinct (and some of you may have missed the end of the video as it was still uploading when I posted the link), if there exists external input/output ("I/O") (meaning the data on the block chain can be written to from a source external to the block chain independent of the data already on the block chain[1]), then the system is Turing complete (which was Nick Szabo and Gregory Maxwell's error when they didn't comprehend Wright's point[1]) thus it is not possible to bound the effects at partition boundaries. Although the validators will see a Nash equilibrium w.r.t. to their incentives (block rewards and in Casper consensus-by-betting penalties), the external effects will be chaotic and thus the value of the coin would collapse due to incorrect outcomes (from the perspective of the users of the system) and thus the Nash equilibrium collapses because validators need for their block rewards to have value in the external market. Apologies I didn't state this succinctly as I was already feeling somewhat tired yesterday evening when I started the video.

Whereas, the I/O for transactions (transferring assets such as crypto coins) is a directed acyclic graph and there is no external processing of coin outputs to input the values into another partition as a external input, thus the above does not apply to asset transfers in the context of strict partitioning (see below for caveat). For scripting on block chain, strict partitioning is impossible to obtain due to the point in the prior paragraph. That is why Ethereum's planned future Casper upgrade will not work.


I meant in the case where there are no transactions that cross partitions (i.e. where all transactions are fully contained within their partition). Once cross-partition transactions are considered, the design is much more complex as Fuserleer noted. In fact, I do believe that perhaps the same Nash equilibrium failure that applies to scripting (as stated above) may apply in the cross-partition design for asset transfers because there is a cascade of history. I need to think about this more. I will try to remember to comment on this point later.

[1]

please note the distinction that although asset transfer transactions (i.e. not scripting) can be written from external source, the data that can be written is dependent or restricted by the data that is already on the block chain, e.g. the UXTO. Wrights claim that Bitcoin's scripting is Turing complete hinges on the ability to store meta-data in the block chain from an external source, thus external state can be recorded in the block chain. However since Bitcoin does not support partitioning, I don't think (but I am not 100% certain yet) the smart contracting layers built on top of Bitcoin such as CounterParty can destroy the Nash equilibrium of Bitcoin.

---

It is humorous to watch people who lack technical understanding try to judge what they read based on how they misperceive the emotions/demeanor of the person writing.

Dude you have no fucking clue. I am not stumped. I have a very clear understanding as explained in this post.

---

As the above clarification shows after a 70 minute video explanation, these issues have many subtle points that are very difficult to explain such that no one has a misunderstanding. For example, you were upthread assuming that in my video I was talking about cross-partition spending, but I was not. I have now clarified that for you in this post (and my prior post just before I slept).

I'd prefer if you stop the erroneous generalizations ad hominem. Thanks.

This area is very difficult to discuss in a forum setting. Not only the technical aspects are easily misunderstood if one detail is not mentioned, but also there are all the egos to deal with such as stoat's above. Please stop putting all the BS of a forum on my shoulders. I do what I can.

I got upset with monsterer inciting you in my Decentralization thead to go spam the thread with 25 posts of nonsense because you were not ready to reveal your entire specification and monsterer was trying to reverse engineer your design by asking a zillion questions and I asked him to take it else where but he refused. Readers can't benefit from that noise at all.

Look I like organization. I like clarity. I like specifications. And yeah I get frustrated with those who don't prioritize those. Asking questions about a specification I understand. Asking questions about an enigma (because you Fuserleer stated you weren't ready to reveal a specification and monsterer was challenging you to release a summary) is far too noisy. If monsterer and you were interested to have that sort of public discussion, you could do it in your own thread. I created my Decentralization thread to talk about issues that I was prepared to completely specify. Then you somehow blamed all of that on me. Then later monsterer didn't understand that a DAG such as Iota is a mathematical model, not a structural model of the trees. I asked him to please discuss it with me outside of my thread until he understood, because his posts were getting noisy because he was pursuing a line of inquiry that was inapplicable. Any way, these are examples of what you refer to as "shouting". Yeah I have to shout over the noise and say "Stop".

As for changing designs again this where I have to deal with your ego. You have changed designs numerous times. Accusing me of changing designs as if that is some sort of sin, thus makes you a hypocrite. Do you think I like to write this? Do you think I love to put the mirror in your face? Not really. You force me to.

You're welcome.  If I had more free time maybe I could get more involved with discussions and cut through the chaff, but unfortunately I don't.

I think you are right to have concerns about POS, its drawbacks from a technical and economical point of view are well documented and I myself was surprised that they even considered it, let alone roll on with an implementation.  To me it feels a bit like a knee jerk reaction with a rushed "betting" solution built on top that won't have had extensive testing performed upon it to see if it actually provides any protection where POS is lacking.

As for the centralization, the road to that with POS is short.  At least with POW there are recurring costs, electricity, hardware maintenance, and at Bitcoin's current hash rate you need a lot of infrastructure just to keep up if you want to hold anything close to influence.  With POS you simply need more money to maintain or increase your stake with little additional or on going effort in comparison.  Sure pure POW is very wasteful, but that is in part what provides the bulk of the protection.

When it came to a consensus mechanism for eMunie I first looked at what was currently available,  POS was the first one struck off the list and swiftly, so make of that what you will.

Thank you for that answer and the further explanation! First: I don't care that much about personalities but sure, sometimes it can be a problem in discussions. In general I just try to filter out what's relevant.

I have to admit that I still don't understand it deeply but your explanation in the context of the discussion and my own researches about ETH seem to me like a confirmation that ETH is highly doubtable. I mean, it's a multi-million-dollar-project and there are obviously issues without solutions. Or with other words: It seems to me as if the ETH-team itself does not know yet how to solve these problems.

My own concerns were mainly about the eco-system if they switch to PoS. I know more about economy than about the technical side and I believe even if it would technically work, a PoS-System must lead into centralization and centralization would make the project irrelevant, impossible to trust it - even more when it comes to contracts. And if I'm right it has no chance in the real economy. Companies won't trust a system if it has no chance in longterm. Not totally sure if I'm right, maybe there are solutions to avoid centralization but if I understand this thread rudimentary correct, there are a lot more problems to solve.

And that in combination with the impression of a highly manipulated market... doesn't look good. As a speculator it's most likely possible to make some profit in the future, because of the hype, but if all those problems are real it's just a question of time that it's more widely discussed and more people would be careful with investing.

However, I'll continue to watch it and trying to understand if ETH has a chance or is doomed by design. But I won't invest any time soon.

Or he gets called out on his bullshit and then quits the thread claiming he's tired when he lost the argument.

He very well might be correct, but its a difficult task to figure it out as everything from him is a brain dump wall of text, with added shouting.

Not to mention that he changes his mind on things frequently, one day something isn't possible, then it is, then he has invented the best thing since slice bread relating to the issue at hand, then it isn't possible again.

Simply put, partitions for transactions are possible (we've done it, both with a block chain and with a ledger).  Block chain based partitions are inefficient, ledger based partitions are not.  ETH will be hard pressed to scale transactions with partitions on a block chain in excess of 2k+ tps, we've had public tests of partitioned ledgers nearing 3k tps (anyone is welcome to come and test it), and small WAN tests up to 10k+tps and I'm still not finished.

Partitions with scripts/contacts are a different kettle of fish, depending on the architecture, this may or may not be possible and the jury is still out on it.  

The main issue is that you have a lot more data, or states, and have to ensure that there are no cross-partition conflicts of those states.  

Further adding to the issue is that the states can be arbitrary, and a script/contract can have many of them.  In the case of a transaction it is defined in the core protocol what the state of a transaction is and how that transaction can behave.  With a script/contract you don't have that luxury as they can define their own states, sub-states and sub-protocols depending on the functionality.

You could simply say where he is wrong, bringing infos, arguments and facts. I'm somebody without much knowledge about the tech, that's why I read discussions about it. My impression in general is (and over months) that there is a lack of controversial discussions about ETH. It's nearly always about the price. My questions about Casper, just for example, were never replied - in no topic.

This is the first thread I know of that focusses on the technical side and the lonterm-potential and if ETH can be what it is aimed to be. And I don't believe that one side is completely right and the other side is totally wrong, but what I see is: All the concerns in this thread are not rebutted yet.

Being someone with first hand experience, and the technical knowledge to present challenges to TPTB's arguments when applicable, I can tell you its just too damn painful to do so.

No developer worth any salt is going to engage with him because it simply devolves into name calling if you don't align with his thoughts.  Before long the n00b, b-lister, you are not worthy comments start regardless of if he's right or not and its just a noisy waste of time.  I've seen the same towards people just trying to understand and not necessarily even disagreeing.

The discussion ends in a blaze of fire with one of the parties leaving, usually it's not TPTB that has departed (he can argue all damn day) so it "looks" from the outside that he won, even if in reality his "facts" are wrong.

If you DO have the patience and manage to back him into a corner, it's simply side stepped and a new argument is presented to justify the last, all twisted together in some obscure manner and the process repeats.

These days I'd rather go and pull my own teeth with a set of pliers than engage in any kind of deep prolonged discussion because its a less painful experience.

Can you explain anonymints technical complaint as you understand it?  Because not one time has he explained it clearly and unobscurely.  And also explain why EXACTLY it is a completely unsolvable problem?  My reading comprehension is second to none and I'm university educated, I have no problems understanding everything else I read but I can't seem for the life of me, able to reach that "aha moment" when anonymint explains this "completely unsolvable problem".

What was your "aha moment"? And can you explain how you arrived at it.

stoat. All you have to do is look at my post history on reddit (farage11) to see I was a fanboy for well over a year. You can also check my history and see that I went all in a few months ago after selling my IPO during the last run up in August. I could care less what you think. Imo, you are nothing more than a unkowledagble shill who badgers AnonyMint with sticks and stones, while he comes back with nukes. Your incessant shilling is juvenile. Name once where you have countered AnonyMints analysis with technical verbiage? Not once! You think people can't read your rebuttals? Show us your technical accumen and put AnonyMint in his place. Show us... that's what I thought. You are out of your depth, and that fact is a fact. Take care stoat, and know thyself. Something tells me that may be harder to attain than ETH's scaling problem.

what price did you sell at?

AnonyMint thank you so much for this video. This is coming from a guy who has been into ETH from day one. You may remember me from Reddit. I was the first one to bring up this issue on their thread. (Sorry again for the drama queen remark), I never sold my stake, but it did cause me concern that nobody took your analysis seriously. Instead you were called a scammer. Not cool for a community that prides itself on being different from this cesspool. As of now, I sold all my ETH. I would love to reinvest, but I am not satisfied that Ethereum is future proof.  I am done with cryptos for the time being, so my money is going back to Silver. I will keep an eye out for your project, but this is it for me. Take care trolls, shills, pumpers, dumpers, traders, devs, investors, bagholders, and Shelby.