Topic: btc (Read 5688 times)

Quote from: QuestionAuthority on June 03, 2015, 10:34:57 PM

Quote from: Newar on June 03, 2015, 10:20:16 PM

Quote from: bitcoinjesus2.0 on June 02, 2015, 02:21:02 AM

Historically, the bitcoin community loves to eat whatever Gavin is serving up, [...]

Huh? Stopped reading after that.

It is kind of true though. Not because he's a cult leader. Anyone in the lead dev position would have the same support unless he's a a loser that steers us in the wrong direction constantly. The surprising thing is that four members of his own team don't support him.

Which is why this may be the first time in human history that the bitcoin community stands up to tyranny and earns its freedom.

The bitcoin community has historically eaten every fork full that Gavin has fed them, and they lapped it up. But now the community has woken up to see that if Gavin successfully pulls off this centralization fork to make the big miners stronger (more expensive for the little guy to compete), then that means that Gavin basically has the big centralized mining cartels in his back pocket meaning that each consecutive centralization fork will become more and more easier to implement until he is able to finally raise the total coin limit to compensate the mining cartels handsomely for following his forks year after year.

But before you cast the first stone, tell me honestly that you would not seize absolute power if it was within your grasp.

In other words, would you turn down the opportunity to have the hottest chick on the planet (and she was madly in love with you and wanted your baby) for an opportunity to save bitcoin anonymously?

I did not think so. Gavin has already won.

Tinfoil hat much? I don't know what it is about crypto that seems to attract conspiracy theorists, but that's some definite crazy right there. This isn't about individual developers "winning" or "losing". Grow up. The choice we're being given is whether bitcoin can scale to support the masses and remain affordable for all, which it simply can't do at the moment unless the fork goes ahead, or whether it becomes a niche, elitist network that most people won't be able to afford to use. I support a network that's open and accessible to all, not just the privileged few like the anti-fork crowd want.

Fuserleer

legendary

Activity: 1050

Merit: 1016

Quote from: troglodactyl on June 03, 2015, 09:40:08 PM

Quote from: Fuserleer on June 03, 2015, 11:27:55 AM

I know that 100,000 TPS sounds impossible, but so does an "automatic blockchain that pays for its own development with its profits"

100,000 TPS isn't impossible, its just impossible to do on a block chain in a true P2P decentralized manner without some form of centralization, or "magic trick". It seems to me BitShares may be doing both as per these 2 quotes in the link you provided.

"...assuming that all the witness nodes have internet connections capable..." so it is indeed confined/centralized to a set of nodes

and

"...with an average transaction size of about 100 bytes."

The latter concerns me, the absolute minimum core basic requirements of a secure verifiable transaction are a sender, a receiver, a value and a signature. Just the signature alone with a 256bit key is in the order of 70 bytes, sender pubkey is 32, receiver RIPEMD160 address is 20, and a value is 8 which is a total of 130 bytes with the bare minimum. If the key space is reduced to 160bit, then it will just fit in 100 bytes, but with a huge loss of security.

I'm assuming that to achieve this 100,000 TPS something similar to this happens:

Transactions are filtered through these "witness nodes" and I send 100 BTS to A. If within a certain time frame, A moves it to B, and B to C, etc etc that TX isn't recorded on the block chain until such a time that it lives at X for longer than a specified period of time. The 100 BTS movements between A -> B-> C....X are not recorded in full on the block chain (if at all), only the transaction A -> X

Taking that approach could indeed give you very high TX throughput, but if that is the method used (or something similar) it's a total hack in my opinion and may well lead to issues later.

Of course I'm speculating as I don't have the time to research this properly, so if you could provide some links/docs/something that details how this works rather than me hunting, I'd appreciate it. I stand by the fact that it is not possible to do, on a block chain, while recording all transactions to said chain, and allow any node to be a full node without special requirements.

Anyway, I'm getting off topic. 180,000 TPS might sound great to some, but Monero destroys BitShares in this arena with the best TPS ceiling of "infinity," so there's that option too of course. In other words, the bitcoin community has infinitely more options than they are currently looking at, and I am just trying to show them that the state of crypto circa 2015 is "not your dad's crypto"

If Monero really has stated "infinity" as their TPS limit, then someone there really needs a reality check! Regardless of what is possible on a block chain or not, the laws of thermodynamics will step in and dispatch a tough and thorough spanking waaaaay before "infinity" is even close

IMO the only way to achieve anything near a sustainable VISA level transaction throughput, stay in keeping with real decentralization (no special node sub-sets that are selected or voted), not perform any "tricks" which may compromise security, AND have all transactions on a public ledger is to scale horizontally, and NOT vertically!

Chain based ledgers can't scale vertically past a certain point, no matter how big your bag of tricks, nor your processing setup, horizontal is the only way and by that I mean a distributed and partitioned ledger of the ilk that we are doing over here. No one has even attempted to do this, because its assumed impossible or too difficult, and if it is so be it, at least it was attempted. However it's not impossible, we've ran it in many betas now and its is very close to being ready for use.

I like what you've done with e-munie. You should come work for the BitShares blockchain. Just submit a proposal, and the community would certainly vote you into a paid position (that's how BitShares members fund development).

https://bitsharestalk.org/index.php/board,61.0.html?PHPSESSID=2170a8f0b09b8fa2bdc7d35908ab4517

Heh thanks but no thanks, I've ploughed my life and everything I have into eMunie and I'm not jumping ship, ever

EDIT
----

So I did some more digging and came across this:

"...the idea being that if transactions have their signatures validated as they propagate across the network, a witness can have any number of computers surrounding him that validates all of these signatures, and then he gets a list of transactions and puts them in a block, and he doesn’t have to check those signatures himself, because he has got all these other nodes surrounding him that are dividing up the task."

Can someone clarify this? Witness nodes, which build the blocks DO NOT check transaction signatures themselves, but rely on 3rd parties (which may be dishonest) to inform them that the signature for said transactions validate? How does a witness node know if a 3rd party is providing false information regarding a transaction, claiming that it contains a valid signature when it may not? If this happens, how then does the network resolve it, someone, somewhere must be doing a full validation of those 100,000 TPS to ensure that all transactions really are legitimate.

The issue is that sequencing and finalizing a block production is done single threaded, because validity is dependent on sequencing. That makes that step the bottleneck, since it can't be scaled to more threads. The signature checking doesn't have to be done by third parties, just separate threads. Those separate threads could be spread across a cluster, or possibly handled by GPUs, or anything else, they just have to be running somewhere trusted by the Witness in order to feed the final Witness thread valid transactions to sequence and include in the block.

It's always a trade-off between decentralization and therefore redundant block validation and transaction fees. Once you get to a globally useful scale, processing transactions has a real cost, and it has to be paid either by inflation of the currency, transaction fees, or subsidies from somewhere else. The BitShares approach is to let the shareholders decide how much redundancy and decentralization they're willing to pay for, and where to set the transaction fees based on that.

Yes block production is single threaded, and multi-threaded signature verification is common sense, but even with the latest CPU's, performing 100,000 signature validations per second is an extremely tall order, commodity CPU's wont get anywhere close to that. GPU's could be an option, but has anyone actually attempted to perform signature verification on a GPU? I'm not aware of anyone doing it, or looking into it at the moment.

The statement made by BitShares was that witness nodes can rely on 3rd parties to verify the signatures, and that doesn't sit comfortably with me, because you have to trust these nodes. Any of those nodes could become dishonest at any moment and then you have to resolve it in the network. A proofing scheme could be created that proves the 3rd party is honest with regard to the transaction, but that will likely be just as expensive as simply doing the verification. Seems like a lot of risk for a TPS value that will likely never be required, or at least not for many many years.

luigi1111

legendary

Activity: 1105

Merit: 1000

Quote from: Fuserleer on June 03, 2015, 02:20:41 PM

Quote from: Fuserleer on June 03, 2015, 02:14:06 PM

Please show me how you are more or less, practically, almost, nearly, close to, verging on, just about, as good as, essentially, to all intents and purposes, roughly, approximately, extracting all the energy from the universe to get close to infinity processing capability?

Now this is just getting retarded, you know very well what I meant with virtually infinite and I have shown by definition of both words its the correct term for Monero, now go play with your emunie.

Indeed I do, I'm just pointing out that by using the definition you presented yourself, is not an accurate representation of Monero's capabilities, and so, it is wrong.

You are only convincing yourself here Cheesy

He's actually quite right (on the definition).

As for Monero scaling to infinity: block size may have no upper bound, but you can be sure current processing efficiency will cause issues before a "high" TPS is achieved (define it as whatever). Lots of work to do...

QuestionAuthority

legendary

Activity: 2156

Merit: 1393

You lead and I'll watch you walk away.

Quote from: Newar on June 03, 2015, 10:20:16 PM

Quote from: bitcoinjesus2.0 on June 02, 2015, 02:21:02 AM

Historically, the bitcoin community loves to eat whatever Gavin is serving up, [...]

Huh? Stopped reading after that.

It is kind of true though. Not because he's a cult leader. Anyone in the lead dev position would have the same support unless he's a a loser that steers us in the wrong direction constantly. The surprising thing is that four members of his own team don't support him.

Newar

legendary

Activity: 1358

Merit: 1001

https://gliph.me/hUF

Quote from: bitcoinjesus2.0 on June 02, 2015, 02:21:02 AM

Historically, the bitcoin community loves to eat whatever Gavin is serving up, [...]

Huh? Stopped reading after that.

troglodactyl

newbie

Activity: 31

Merit: 0

Quote from: Fuserleer on June 03, 2015, 11:27:55 AM

I know that 100,000 TPS sounds impossible, but so does an "automatic blockchain that pays for its own development with its profits"

100,000 TPS isn't impossible, its just impossible to do on a block chain in a true P2P decentralized manner without some form of centralization, or "magic trick". It seems to me BitShares may be doing both as per these 2 quotes in the link you provided.

"...assuming that all the witness nodes have internet connections capable..." so it is indeed confined/centralized to a set of nodes

and

"...with an average transaction size of about 100 bytes."

The latter concerns me, the absolute minimum core basic requirements of a secure verifiable transaction are a sender, a receiver, a value and a signature. Just the signature alone with a 256bit key is in the order of 70 bytes, sender pubkey is 32, receiver RIPEMD160 address is 20, and a value is 8 which is a total of 130 bytes with the bare minimum. If the key space is reduced to 160bit, then it will just fit in 100 bytes, but with a huge loss of security.

I'm assuming that to achieve this 100,000 TPS something similar to this happens:

Transactions are filtered through these "witness nodes" and I send 100 BTS to A. If within a certain time frame, A moves it to B, and B to C, etc etc that TX isn't recorded on the block chain until such a time that it lives at X for longer than a specified period of time. The 100 BTS movements between A -> B-> C....X are not recorded in full on the block chain (if at all), only the transaction A -> X

Taking that approach could indeed give you very high TX throughput, but if that is the method used (or something similar) it's a total hack in my opinion and may well lead to issues later.

Of course I'm speculating as I don't have the time to research this properly, so if you could provide some links/docs/something that details how this works rather than me hunting, I'd appreciate it. I stand by the fact that it is not possible to do, on a block chain, while recording all transactions to said chain, and allow any node to be a full node without special requirements.

Anyway, I'm getting off topic. 180,000 TPS might sound great to some, but Monero destroys BitShares in this arena with the best TPS ceiling of "infinity," so there's that option too of course. In other words, the bitcoin community has infinitely more options than they are currently looking at, and I am just trying to show them that the state of crypto circa 2015 is "not your dad's crypto"

If Monero really has stated "infinity" as their TPS limit, then someone there really needs a reality check! Regardless of what is possible on a block chain or not, the laws of thermodynamics will step in and dispatch a tough and thorough spanking waaaaay before "infinity" is even close

IMO the only way to achieve anything near a sustainable VISA level transaction throughput, stay in keeping with real decentralization (no special node sub-sets that are selected or voted), not perform any "tricks" which may compromise security, AND have all transactions on a public ledger is to scale horizontally, and NOT vertically!

Chain based ledgers can't scale vertically past a certain point, no matter how big your bag of tricks, nor your processing setup, horizontal is the only way and by that I mean a distributed and partitioned ledger of the ilk that we are doing over here. No one has even attempted to do this, because its assumed impossible or too difficult, and if it is so be it, at least it was attempted. However it's not impossible, we've ran it in many betas now and its is very close to being ready for use.

Quote from: ArticMine on June 03, 2015, 06:31:31 PM

I like what you've done with e-munie. You should come work for the BitShares blockchain. Just submit a proposal, and the community would certainly vote you into a paid position (that's how BitShares members fund development).

https://bitsharestalk.org/index.php/board,61.0.html?PHPSESSID=2170a8f0b09b8fa2bdc7d35908ab4517

Heh thanks but no thanks, I've ploughed my life and everything I have into eMunie and I'm not jumping ship, ever

EDIT
----

So I did some more digging and came across this:

"...the idea being that if transactions have their signatures validated as they propagate across the network, a witness can have any number of computers surrounding him that validates all of these signatures, and then he gets a list of transactions and puts them in a block, and he doesn’t have to check those signatures himself, because he has got all these other nodes surrounding him that are dividing up the task."

Can someone clarify this? Witness nodes, which build the blocks DO NOT check transaction signatures themselves, but rely on 3rd parties (which may be dishonest) to inform them that the signature for said transactions validate? How does a witness node know if a 3rd party is providing false information regarding a transaction, claiming that it contains a valid signature when it may not? If this happens, how then does the network resolve it, someone, somewhere must be doing a full validation of those 100,000 TPS to ensure that all transactions really are legitimate.

The issue is that sequencing and finalizing a block production is done single threaded, because validity is dependent on sequencing. That makes that step the bottleneck, since it can't be scaled to more threads. The signature checking doesn't have to be done by third parties, just separate threads. Those separate threads could be spread across a cluster, or possibly handled by GPUs, or anything else, they just have to be running somewhere trusted by the Witness in order to feed the final Witness thread valid transactions to sequence and include in the block.

It's always a trade-off between decentralization and therefore redundant block validation and transaction fees. Once you get to a globally useful scale, processing transactions has a real cost, and it has to be paid either by inflation of the currency, transaction fees, or subsidies from somewhere else. The BitShares approach is to let the shareholders decide how much redundancy and decentralization they're willing to pay for, and where to set the transaction fees based on that.

kazuki49

sr. member

Activity: 350

Merit: 250

The idea that one can push an infinite number of transactions through the Monero network is utter nonsense. Monero uses adaptive limits that limit the blocksize dynamically. This is explained in section 6.2 of the Cryptonote Whitepaper https://cryptonote.org/whitepaper.pdf. This means that there is no fixed maximum number of TPS that cannot be exceeded regardless of the market conditions. This is the critical difference with not just Bitcoin, but with Litecoin, Dodgecoin, Dash and many other alt-coins.

yup, thats why I'm trying to make OP change to at least virtually infinite, its pretty clear that respecting laws of physics is a major hole in the XMR ark Cheesy

celestio

sr. member

Activity: 770

Merit: 250

The only coins in that poll that are worth anything is Monero and Bitcoin of course.

Fuserleer

legendary

Activity: 1050

Merit: 1016

Quote from: ArticMine on June 03, 2015, 06:31:31 PM

The idea that one can push an infinite number of transactions through the Monero network is utter nonsense. Monero uses adaptive limits that limit the blocksize dynamically. This is explained in section 6.2 of the Cryptonote Whitepaper https://cryptonote.org/whitepaper.pdf. This means that there is no fixed maximum number of TPS that cannot be exceeded regardless of the market conditions. This is the critical difference with not just Bitcoin, but with Litecoin, Dodgecoin, Dash and many other alt-coins.

Thats what I was looking for, thank you!

ArticMine

legendary

Activity: 2282

Merit: 1050

Monero Core Team

The idea that one can push an infinite number of transactions through the Monero network is utter nonsense. Monero uses adaptive limits that limit the blocksize dynamically. This is explained in section 6.2 of the Cryptonote Whitepaper https://cryptonote.org/whitepaper.pdf. This means that there is no fixed maximum number of TPS that cannot be exceeded regardless of the market conditions. This is the critical difference with not just Bitcoin, but with Litecoin, Dodgecoin, Dash and many other alt-coins.

Anima

member

Activity: 63

Merit: 10

Quote from: kazuki49 on June 03, 2015, 02:26:01 PM

haha ok...

just confirming you wanted to FUD Monero btw Kiss

https://lab.getmonero.org/

Monero is one of the few coins besides Bitcoin with real mathematicians and cryptographers behind.

emunie is just a joke like XEM Undecided

I'm a little biased by nature since i am a part of the emunie project, but i cannot see why asking critical questions regarding claims made by different devs/projects is speading FUD. There are several problems highlighted in this thread surrounding these claims, which i do not see you or anyone else have been able to answer to a satisfactory degree.

Then you have the audacity to go so low and proclaim that emunie is a joke. Its the same thing every time here on BTT. You go in and try and have a civil debate and all it turns out to, when people have no counter-arguments, is schoolyard bullying.

Either acknowledge that Monero's performance claims are not true or don't. Let that be your opinion and lets close it there. Do not downplay other peoples projects just because they found a hole in your Noahs ark.

Fuserleer

legendary

Activity: 1050

Merit: 1016

Quote from: kazuki49 on June 03, 2015, 02:26:01 PM

Quote from: Fuserleer on June 03, 2015, 02:24:40 PM

Quote from: Fuserleer on June 03, 2015, 02:20:41 PM

Quote from: Fuserleer on June 03, 2015, 02:14:06 PM

Please show me how you are more or less, practically, almost, nearly, close to, verging on, just about, as good as, essentially, to all intents and purposes, roughly, approximately, extracting all the energy from the universe to get close to infinity processing capability?

Now this is just getting retarded, you know very well what I meant with virtually infinite and I have shown by definition of both words its the correct term for Monero, now go play with your emunie.

Indeed I do, I'm just pointing out that by using the definition you presented yourself, is not an accurate representation of Monero's capabilities, and so, it is wrong.

You are only convincing yourself here Cheesy

Not really, this is one of the many reasons that the mass market is confused, because people involved with projects make statements such as that which are fundamentally wrong, can not be achieved or proven, and only serve the purpose of "bragging rights" against other technologies.

Additionally statements such as that result in good projects such as Monero not being taken seriously by academics and other educated individuals in the real world, because statements are presented as fact, that can not be fact.

haha ok...

just confirming you wanted to FUD Monero btw Kiss

https://lab.getmonero.org/

Monero is one of the few coins besides Bitcoin with real mathematicians and cryptographers behind.

emunie is just a joke like XEM Undecided

I didn't want to FUD anything, just pointing these claims and the possibility of them being true or not.

Yeah yeah, eMunie is a joke, XEM is a joke, XYZ is a joke, awesome, lets descend to childish playground statements.

kazuki49

sr. member

Activity: 350

Merit: 250

Quote from: Fuserleer on June 03, 2015, 02:24:40 PM

Quote from: Fuserleer on June 03, 2015, 02:20:41 PM

Quote from: Fuserleer on June 03, 2015, 02:14:06 PM

Please show me how you are more or less, practically, almost, nearly, close to, verging on, just about, as good as, essentially, to all intents and purposes, roughly, approximately, extracting all the energy from the universe to get close to infinity processing capability?

Now this is just getting retarded, you know very well what I meant with virtually infinite and I have shown by definition of both words its the correct term for Monero, now go play with your emunie.

Indeed I do, I'm just pointing out that by using the definition you presented yourself, is not an accurate representation of Monero's capabilities, and so, it is wrong.

You are only convincing yourself here Cheesy

Not really, this is one of the many reasons that the mass market is confused, because people involved with projects make statements such as that which are fundamentally wrong, can not be achieved or proven, and only serve the purpose of "bragging rights" against other technologies.

Additionally statements such as that result in good projects such as Monero not being taken seriously by academics and other educated individuals in the real world, because statements are presented as fact, that can not be fact.

haha ok...

just confirming you wanted to FUD Monero btw Kiss

https://lab.getmonero.org/

Monero is one of the few coins besides Bitcoin with real mathematicians and cryptographers behind.

emunie is just a joke like XEM Undecided

Fuserleer

legendary

Activity: 1050

Merit: 1016

Quote from: Fuserleer on June 03, 2015, 02:20:41 PM