Pages:
Author

Topic: Stop fuckin' around, fork the son-of-a-bitch already. - page 10. (Read 9352 times)

legendary
Activity: 2674
Merit: 2965
Terminated.
-snip-
Tl;dr; I would sacrifice decentralization for money.  Roll Eyes

lots of people have done benchmarks and deemed 4mb raspberry pi safe.
but here is lauda being ignorant and now avoiding other bench tests just to provoke an argument that i didnt do as he said.
No. You're the one that mentioned validation time on a Pi in this thread. I asked YOU to run YOUR own benchmarks, not pull stuff from the internet to back it up.

ill repeat what he and his immature friends say when i ask them to research something
"dont tell me what to do unless your paying me"
I'm not going to waste my time researching for someone who thinks that libsecp256k1 mitigates the quadratic validation time problem.

and even if i did spoonfeed him. i predict he would just say i made it up (yes he is that predictable)
Of course you'd likely make it up since you can't even properly document your testing methodology.

anyway using todays settings 2mb spam is 4 seconds.. not a 11 minute bottleneck
You clearly have no idea what you're talking about.
hero member
Activity: 3052
Merit: 685
That is the nature of the game therefore we have no choice but play with it, so far in my opinion ETH is the best investment as I can easily read the movement of the market and I make consistent money trading it.
legendary
Activity: 4410
Merit: 4766
I only ask that people try to understand that leveraging miners against users to try to force a network migration is unethical, and I ask that people reflect on whether it is right for miners to provoke users into removing consensus rules. Hard forks are a matter of user consent, not miner consent.

Miner consent only establishes the longest valid chain (so, it enables soft forks). It has nothing to do with the consensus rules that every node on the network enforces.

cores softfork does not need user consent at actual code requirement level. (as part of the real consensus mechanism).
absolutely no nodes needs to upgrade to make a softfork possible.
but a mining pool DOES need to upgrade to be able to validate signatures of a segwit transaction to include them in blocks.
yes mining pools actually validate data to ensure it doesnt orphan.. so miners are not there just to decide blockheight..

seriously im starting to think im talking to a noob right now

core adding in an activation mechanism. is not due to consensus rules requiring it, but because they have bypassed the consensus mechanism to be able to change the rules without using the consensus mechanism and as an afterthought putting in a social activation mechanism to pretend its the same thing as consensus mechanism
but fundamentally not required.

old nodes are not forward compatible in the sense of being able to validate signatures of a segwit transaction. they are tricked into just passing on the data unchecked. this is not ensuring an old node remains a full validation node.. but just a blind participant in a game of pass the parcel

you say that segwit has been around for months.. on other networks, not bitcoin.
there are many BITCOIN implementations that have code for a hardfork that have been around since last year.. core have only achieved having the code IN BITCOIN a few weeks ago.

your rebuttal is like saying bitcoin has had ~2.5 minute blocks for 5 years because there is code being tested on a different network.
(segnet does not have 7 years of bitcoin data and is not talking to 6000 bitcoin nodes.. much like litecoin isnt)

but i will just laugh at your other false assumptions because its obvious no matter what core does, they are your king.. you have lost full understanding of decentralization and have been fooled into blindly following the centralization of false choice

by the way. if you have not already been indoctrinated to start favouring monero (the usual path of those wanting to keep bitcoin down) you soon will be. much like all the other core fanboys.

i have seen it happen many times, you probably are not going to be immune to it, because you are already repeating the standard spoonfed false rhetoric and hypocrisy

so the day you start thinking monero is the future, at this point realise that someone had warned you that your 'friends' have been tweaking at your mind to move you to favour monero with all the spoon fed false scripts. slap yourself and wake up

...
but here is the mind blowing part.
because core (didnt have to) but are putting in an activation mechanism where 95% of nodes need to use a piece of software..
guess what.
put the real capacity block buffer increase (meaning 2mb baseblock and 4mb segwit blockweight) so that when it activates the network CONTINUES to add new blocks onto the chain. and the minority get orphaned (via consensus mechanism)
after all there is no point running a full node if your not going to be validating transactions and just blindly passing on data you cant check.

killing 2 birds within one stone(mechanism/oppertunity)


and dont reply with the fake rhetoric of split coins, which can only occur by blacklisting nodes to by pass the orphan mechanism
legendary
Activity: 1806
Merit: 1521
Franky, you're spreading a lot of misinformation. I won't go so far as to call it disinformation, but I'm not sure that your intentions here are honorable. I can't keep responding to you because every time I address your points, you repeat the same falsities that I've already shown to be false.

No, I am not. This has nothing to do with controversy. A hard fork, by definition, entails creating a new network that is incompatible with the old one. The intention of a hard fork is for users of the old network to migrate to the forked network. There is no propaganda in that, whatsoever.

Supporting Core's conservative method of rigorous analysis and testing, or their overall roadmap, or simply supporting an implementation that retains Bitcoin's consensus rules, is not anything you can blame Core for.
1. hardfork code has been publicly available since last year, meaning core is not conservative. because they ignore code available for over a year to quickly push something only finalised a few weeks ago..

Segwit has been worked on since late last year. It's been rigorously tested for many months. Simply stating that "hard fork code has been publicly available since last year" is meaningless. That doesn't suggest that the hard fork code is sensible, optimal nor rigorously tested. Rather, peer review and testing has been minimal.

2. core is not retaining consensus rules.. old nodes cant validate segwit, so old nodes are not part of the consensus. they are just blindly confused into not caring what the data is by using the flag that tells them to ignore it.

Please point to the consensus rules being broken by Segwit. You can't, because there are none. This is why Segwit is backward compatible.

Old nodes certainly can validate Segwit transactions against the consensus rules. And they will. You seem disturbed by the idea of forward compatibility, but it's really quite elegant and appropriate. See Satoshi's thoughts on forward compatible soft forks:

The nature of Bitcoin is such that once version 0.1 was released, the core design was set in stone for the rest of its lifetime.  Because of that, I wanted to design it to support every possible transaction type I could think of.  The problem was, each thing required special support code and data fields whether it was used or not, and only covered one special case at a time.  It would have been an explosion of special cases.  The solution was script, which generalizes the problem so transacting parties can describe their transaction as a predicate that the node network evaluates.  The nodes only need to understand the transaction to the extent of evaluating whether the sender's conditions are met.

The script is actually a predicate.  It's just an equation that evaluates to true or false.  Predicate is a long and unfamiliar word so I called it script.

The receiver of a payment does a template match on the script.  Currently, receivers only accept two templates: direct payment and bitcoin address.  Future versions can add templates for more transaction types and nodes running that version or higher will be able to receive them.  All versions of nodes in the network can verify and process any new transactions into blocks, even though they may not know how to read them.

The design supports a tremendous variety of possible transaction types that I designed years ago.  Escrow transactions, bonded contracts, third party arbitration, multi-party signature, etc.  If Bitcoin catches on in a big way, these are things we'll want to explore in the future, but they all had to be designed at the beginning to make sure they would be possible later.

I don't believe a second, compatible implementation of Bitcoin will ever be a good idea.  So much of the design depends on all nodes getting exactly identical results in lockstep that a second implementation would be a menace to the network.  The MIT license is compatible with all other licenses and commercial uses, so there is no need to rewrite it from a licensing standpoint.

3. core is not retaining consensus rules.. 4mb blockweight is a total different rule than maxblocksize.. but again by disabling consensus(consent) they are changing rules without needing consent.

Again: Please point to the consensus rules being broken by Segwit. You can't, because there are none. Maxblocksize is a consensus rule. It is not broken by Segwit. Segregating the witness data doesn't affect validity. It's really quite a clever way to optimize transactions size while permitting features which require non-malleability.

4. the road map belongs to core, the segwit code is a core feature and core are avoiding accepting hard fork code so much they are throwing devs out of the core group

Anyone is free to contribute to Core. Anyone is also free to develop other clients. Being a small, vocal minority should not enable someone to steamroll the rest of the project contributors into merging code they disagree with. You have no authority to say what Core should or should not do. You are free to contribute to other projects. Go ahead.

You can blame the lack of convincing arguments by fork proponents for the lack of support to fork. You could also probably blame Ethereum, which has likely scared much of the community away from the risks of hard forks.
the REKT and fake doomsdays stories to sway people started way before ethereum

Indeed, anti-fork sentiment existed long before the Ethereum network split. But the failed hard fork scared a lot of people into recognizing that Core's position has been reasonable and thoughtful, while the Ethereum Foundation has clearly been reckless and disdainful towards their userbase and custodians.

"Consensual [hard] fork" is a strange term. Consent flows from users opting into a network. Anyone is free to opt out of the Bitcoin network and consent to a new hard fork. But that doesn't imply at all that all users of the Bitcoin network consent to a hard fork. Knowing that every user consents is not possible in any way, shape or form -- and even if it were, we could only know after the fork had occurred, after which users actually migrate and affirmatively consent to the hard fork's new rules.
see there you go again "opt out of bitcoin network".. more subtly propaganda that if its not the core way then its not bitcoin..
yet core are changing the rules without needing users consent for the new direction core want to go.

No, it's not propaganda at all. You either fundamentally misunderstand what a hard fork entails, or you yourself are attempting to spread lies to further your agenda.

Are you suggesting that the Bitcoin network doesn't exist? Because if the Bitcoin network exists, a hard fork of Bitcoin's consensus protocol will create an additional network that is incompatible with Bitcoin. It's curious that you don't deny this, yet you continue to try to deceive people into thinking that a hard fork is merely a software change that is compatible with Bitcoin. It is not. It splits the network into 2 or more incompatible networks.

Core has no obligation to attempt to break the consensus rules. I don't know where you get the impression that they do. Anyone is free to fork Core's code and try to convince the community to join them. However, I urge everyone to oppose any fork that attempts to leverage miners against users. That raises serious ethical concerns about whether migrating users actually consent, or whether they felt forced by colluding miners (who have an apparent monopoly on hash rate and thus, confirmations for transactions).

"anyone is free to fork cores code"
hmm there you go again. subtle propaganda saying bitcoin is core and core is bitcoin, meaning anything not core, is not bitcoin.. seems you actually want core to be king and own bitcoin and make decisions without users consent, and you want anyone revolting away from your ideology to be classed as a alternative network/coin.

I didn't say "Bitcoin is Core and Core is Bitcoin." But it's the reference implementation. Why do you think XT, Classic and Unlimited are all forks of Core? What else would you expect people to fork, if not Core's repository? Feel free to code a new implementation from scratch, if you'd prefer.

but here is the thing that will trip you up..
orphans are the mechanism.. and orphans happen daily. meaning bitcoin is suppose to change but only in a direction the majority agree on.
so a true consensual fork is the same as the daily orphans.. the chain CONTINUES in the direction of the majority..

"Orphans" refer to valid blocks that don't have a known parent in the longest block chain. They don't refer to invalid blocks that aren't part of the Bitcoin network.

Also, you're only addressing processing power (miners). You're not addressing users at all. What gives miners power over the software that users decide to run? Nothing.

if someone makes a rule change by submitting a block with odd data using the real consensual mechanism, but doesnt have majority, it gets orphaned
every day there are decisions being made for which direction the majority should move.
so by your failed definition of the consensus mechanism.. the majority of users are moving to a different network every time there is a fork(orphan)..

No, an orphan doesn't imply a new network at all. Orphans are valid blocks, so they are compatible with the original network. Miners are simply agreeing on which valid chain has the most cumulative work.

It seems you need a reminder of how this works:
http://cointimes.tech/2016/08/hard-forks-and-consensus-networks-meta-questions-and-limitations/
http://www.gsd.inesc-id.pt/~ler/docencia/rcs1314/papers/P2P2013_041.pdf

a controversial fork still gets orphaned off..
but.. only when intentionally blacklisting certain nodes to prevent the orphaning mechanism from taking care of the situation is where a realistic and continual split occur.

A fork that breaks the consensus rules doesn't get orphaned. It gets ignored. Saying that it is "orphaned" suggests it is valid according to the network. It isn't.

the community on the whole dont want a intentional split by adding in code to bypass the natural orphaning mechanism, all that is being said is a consensual fork for true capacity growth

which wont happen if core prevent a healthy majority by doing all this propaganda crap to keep their sheep inline to veto a change. and blindly allow another change by not telling their sheep its not requiring their consent for that route either

think long and hard about what i have just said.. and dont reply just with waffle to defend cores control

Core doesn't have any control. Please come to terms with the fact that most of the community supports Core's maintenance of the Bitcoin code. You may disagree with their views -- it's only natural in such a diverse ecosystem that not every participant sees eye to eye. But I'm becoming more and more confused as to what you are even arguing for.

Anyone is free to opt out of the network and run an incompatible node. Anyone is free to contribute to projects other than Bitcoin Core. Please feel free to support such efforts.

I only ask that people try to understand that leveraging miners against users to try to force a network migration (as XT and Classic attempted to do) is unethical, and I ask that people reflect on whether it is right for miners to provoke users into removing consensus rules. Hard forks are a matter of user consent, not miner consent. Miner consent only establishes the longest valid chain (so, it enables soft forks). It has nothing to do with the consensus rules that every node on the network enforces.

Many of us view the block size limit as a moral issue -- key to allowing Bitcoin to remain peer-to-peer and sufficiently decentralized to prevent attacks -- and further, that consensus rules must be safe from tyranny of the majority. For miners to leverage "fear of being on the weaker chain" against users to remove the block size limit, for me, is not very different from miners doing so to remove the 21M supply limit. This idea that consensus rules are subject to democratic vote is extremely dangerous, and I oppose it unequivocally.

And frankly, I don't think that you are convincing anyone here of anything -- your understanding of the protocol is clearly lacking and you are clearly very angry, which obviously gets in the way of your message.
legendary
Activity: 4410
Merit: 4766
No, I am not. This has nothing to do with controversy. A hard fork, by definition, entails creating a new network that is incompatible with the old one. The intention of a hard fork is for users of the old network to migrate to the forked network. There is no propaganda in that, whatsoever.

Supporting Core's conservative method of rigorous analysis and testing, or their overall roadmap, or simply supporting an implementation that retains Bitcoin's consensus rules, is not anything you can blame Core for.
1. hardfork code has been publicly available since last year, meaning core is not conservative. because they ignore code available for over a year to quickly push something only finalised a few weeks ago..
2. core is not retaining consensus rules.. old nodes cant validate segwit, so old nodes are not part of the consensus. they are just blindly confused into not caring what the data is by using the flag that tells them to ignore it.
3. core is not retaining consensus rules.. 4mb blockweight is a total different rule than maxblocksize.. but again by disabling consensus(consent) they are changing rules without needing consent.
4. the road map belongs to core, the segwit code is a core feature and core are avoiding accepting hard fork code so much they are throwing devs out of the core group

You can blame the lack of convincing arguments by fork proponents for the lack of support to fork. You could also probably blame Ethereum, which has likely scared much of the community away from the risks of hard forks.
the REKT and fake doomsdays stories to sway people started way before ethereum

"Consensual [hard] fork" is a strange term. Consent flows from users opting into a network. Anyone is free to opt out of the Bitcoin network and consent to a new hard fork. But that doesn't imply at all that all users of the Bitcoin network consent to a hard fork. Knowing that every user consents is not possible in any way, shape or form -- and even if it were, we could only know after the fork had occurred, after which users actually migrate and affirmatively consent to the hard fork's new rules.
see there you go again "opt out of bitcoin network".. more subtly propaganda that if its not the core way then its not bitcoin..
yet core are changing the rules without needing users consent for the new direction core want to go.


Core has no obligation to attempt to break the consensus rules. I don't know where you get the impression that they do. Anyone is free to fork Core's code and try to convince the community to join them. However, I urge everyone to oppose any fork that attempts to leverage miners against users. That raises serious ethical concerns about whether migrating users actually consent, or whether they felt forced by colluding miners (who have an apparent monopoly on hash rate and thus, confirmations for transactions).

"anyone is free to fork cores code"
hmm there you go again. subtle propaganda saying bitcoin is core and core is bitcoin, meaning anything not core, is not bitcoin.. seems you actually want core to be king and own bitcoin and make decisions without users consent, and you want anyone revolting away from your ideology to be classed as a alternative network/coin.

but here is the thing that will trip you up..
orphans are the mechanism.. and orphans happen daily. meaning bitcoin is suppose to change but only in a direction the majority agree on.
so a true consensual fork is the same as the daily orphans.. the chain CONTINUES in the direction of the majority..

if someone makes a rule change by submitting a block with odd data using the real consensual mechanism, but doesnt have majority, it gets orphaned
every day there are decisions being made for which direction the majority should move.
so by your failed definition of the consensus mechanism.. the majority of users are moving to a different network every time there is a fork(orphan)..

a controversial fork still gets orphaned off..
but.. only when intentionally blacklisting certain nodes to prevent the orphaning mechanism from taking care of the situation is where a realistic and continual split occur.

the community on the whole dont want a intentional split by adding in code to bypass the natural orphaning mechanism, all that is being said is a consensual fork for true capacity growth

which wont happen if core prevent a healthy majority by doing all this propaganda crap to keep their sheep inline to veto a change. and blindly allow another change by not telling their sheep its not requiring their consent for that route either

think long and hard about what i have just said.. and dont reply just with waffle to defend cores control
legendary
Activity: 1806
Merit: 1521
Propaganda? Huh

By definition a hard fork creates a new network. This a matter of fact, not opinion.

What I said didn't contemplate whether a hard fork is controversial. What I said was, why don't you fork the code? Why do you need miners to pressure users to switch clients?

It seems the answer you buried in your long-winded posts was "nobody would support my fork." Okay then. There's your answer.

again your only seeing a hard fork through the eyes of controversial.

No, I am not. This has nothing to do with controversy. A hard fork, by definition, entails creating a new network that is incompatible with the old one. The intention of a hard fork is for users of the old network to migrate to the forked network. There is no propaganda in that, whatsoever.

but by core not even trying to make it (lets hope luke releases it before getting thrown out) then there will never be a healthy majority, due to
the loyal flock wanting core to dominate(even if the flock dont understand the route core prefers) and make decisions for them (centrally)

You're blaming the community for supporting the reference implementation. Take it up with the community. Supporting Core's conservative method of rigorous analysis and testing, or their overall roadmap, or simply supporting an implementation that retains Bitcoin's consensus rules, is not anything you can blame Core for. You can blame the lack of convincing arguments by fork proponents for the lack of support to fork. You could also probably blame Ethereum, which has likely scared much of the community away from the risks of hard forks.

so yea bitcoin knots, airbitz and others can have perfect code.. but core fanboys will doomsday call anything not core.. purely because core doesnt want a consensual fork(real onchain capacity).. they want to lead and control decisions. not allow an open decentralized free choice.

"Consensual [hard] fork" is a strange term. Consent flows from users opting into a network. Anyone is free to opt out of the Bitcoin network and consent to a new hard fork. But that doesn't imply at all that all users of the Bitcoin network consent to a hard fork. Knowing that every user consents is not possible in any way, shape or form -- and even if it were, we could only know after the fork had occurred, after which users actually migrate and affirmatively consent to the hard fork's new rules.

the funny thing is. core can easily release 0.14B(segwit+ hardfork) and then just see if people download A or B.
and if 95% have not downloaded version B of their favourite. then it will not activate and nothing is lost.
but by core vetoing any chance of a core version B. they are not even allowing users a free choice, even amungst their own flock

Core has no obligation to attempt to break the consensus rules. I don't know where you get the impression that they do. Anyone is free to fork Core's code and try to convince the community to join them. However, I urge everyone to oppose any fork that attempts to leverage miners against users. That raises serious ethical concerns about whether migrating users actually consent, or whether they felt forced by colluding miners (who have an apparent monopoly on hash rate and thus, confirmations for transactions).
legendary
Activity: 4410
Merit: 4766
I really don't think a hard fork is a good move.
Maybe a side-chain or something like that.

And well, we have a lot of good alts over there, bitcoin is not unanimous anymore.

^ this here is the mindset of why bitcoin could lose utility.
yes thats right im talking about function, im talking about using actual bitcoin to buy actual products

in short they dont want to expand bitcoin, instead use an altcoin that has the desired capacity and make bitcoin no longer a functional currency and just an investment

we all know gmaxwell and lauda+chums want monero to rule supreme

and its funnier that they say its us wanting capacity growth purely for the value increase
while they are trying to get monero value to increase by baiting people over to it and turn bitcoin into just an pump and dump reserve.

(every doomsday and argument they can shout out about others is usually exactly what they want themselves)
EG
"big blockers" has been proven to now be the term for the core team, 4mb bigger then 2mb afterall
sr. member
Activity: 364
Merit: 250
I really don't think a hard fork is a good move.
Maybe a side-chain or something like that.

And well, we have a lot of good alts over there, bitcoin is not unanimous anymore.
legendary
Activity: 4410
Merit: 4766
lots of people have done benchmarks and deemed 4mb raspberry pi safe.
but here is lauda being ignorant and now avoiding other bench tests just to provoke an argument that i didnt do as he said.

ill repeat what he and his immature friends say when i ask them to research something
"dont tell me what to do unless your paying me"

so if he wants to ignore the many many benchmarks that google can supply him. then its obvious he doesnt care about actual statistics.
and even if i did spoonfeed him. i predict he would just say i made it up (yes he is that predictable)

anyway using todays settings 2mb spam is 4 seconds.. not a 11 minute bottleneck

have a nice day
hero member
Activity: 2506
Merit: 645
Eloncoin.org - Mars, here we come!
Planned, well executed, not rushed hard fork for the sake of achieving  better performance, stability, speed -  pushing evolution in general is welcomed and needed.
The problem is that some users are scared/don't know what forking is about and I'm afraid that forking will always cause problems and panic - at least initially.


It is fine and makes sense as long as everyone knows the deal.  There are a lot of people that have a wallet and use it on occasion, but do not follow the Bitcoin news and what not. 
hero member
Activity: 546
Merit: 500
I am going to walk on the safe path, and go with the direction these Core guys are going. Why do we need to change anything if it is working? The big blockers were creating Apocalyptic scenarios since this debate has started many moons ago, and guess what, nothing happened.

We have seen strange stress tests being done, to make people believe that bigger blocks are necessary, but the network came through with flying colors and these people lost money.
Because it is not working as intended, a congested network leads to longer and more unreliable confirmation times and higher fees. This is not good for Bitcoin adoption, much has happened over the last year in this debate including Bitcoin continuously losing more of its market share to its competitors. Keeping this limit in place restricts the amount of people that can use Bitcoin regardless of the fee that is payed. This one simple parameter is hurting the adoption and growth of Bitcoin which is critical over the long term for its decentralization and expansion. Here is proof that this is occurring:




legendary
Activity: 1862
Merit: 1004
Planned, well executed, not rushed hard fork for the sake of achieving  better performance, stability, speed -  pushing evolution in general is welcomed and needed.
The problem is that some users are scared/don't know what forking is about and I'm afraid that forking will always cause problems and panic - at least initially.
legendary
Activity: 1596
Merit: 1026
Blah, blah, blah

Hey Lauda, do you take drugs?
legendary
Activity: 2674
Merit: 2965
Terminated.
i even bothered to check a few websites to see several benchmarks to to make sure its not a fluke..
but heres one.. which also covers the bits about 8mb too
https://rusty.ozlabs.org/?p=515
Initially I've asked you to provide your own benchmarks on a raspberry PI. These are not your results.

seems rusty can get better than 2 seconds, but im guessing from reading it he was using a i3 laptop.
there were other sites and benchmarks saying increasing the dbcache also improved things too, but yea that digresses off the point
Increasing the dbcache does speed up overall time, but that's not what we're trying to accomplish. I'm talking about default settings, and raspberry PI (since you've mentioned it).

basically 2mb spam tx is not a bottleneck of doom 11minute scare story.. but a 4 second validation time
That's not correct.

by the way google and a thing called research is your friend.
You've made claims, ergo you should back them up with sources. I'm not going to do the homework for you.
legendary
Activity: 4410
Merit: 4766
lol so transactions are no longer median 226 bytes? or as i said average ~450bytes
Median is still 226 bytes, as shown on this website.

remember the sigops issue is not a problem for median/average transactions.(you know transactions that dont have much signatures) and only a problem for bloated transactions (you know such as what will be noticeable when lightning hubs start dealing with many signatures) where the average and even your magical median increase.
No, neither one holds true. Nobody ever claimed that the quadratic scaling is problematic with normal transactions. It is just an attack vector that could be abused by a malicious miner. "Could be" doesn't mean that it "will be" though.

block: 364292 validated in 2 seconds..
as oppose to pre libsecp256k1 which took far far far longer.
Where is the source of this result ("2 seconds")?

other standard blocks 0-400,000 validated in under 6 hours (18 blocks per second)
as oppose to pre libsecp256k1 which took far far far longer.
That doesn't matter. As previously stated, while libsecp256k1 improves validation time significantly it does not mitigate the attack vector at 2 MB (Gavin even confirmed this somewhere).

i even bothered to check a few websites to see several benchmarks to to make sure its not a fluke..
but heres one.. which also covers the bits about 8mb too
https://rusty.ozlabs.org/?p=515

seems rusty can get better than 2 seconds, but im guessing from reading it he was using a i3 laptop.
there were other sites and benchmarks saying increasing the dbcache also improved things too, but yea that digresses off the point

basically 2mb spam tx is not a bottleneck of doom 11minute scare story.. but a 4 second validation time
by the way google and a thing called research is your friend.
legendary
Activity: 2674
Merit: 2965
Terminated.
lol so transactions are no longer median 226 bytes? or as i said average ~450bytes
Median is still 226 bytes, as shown on this website.

remember the sigops issue is not a problem for median/average transactions.(you know transactions that dont have much signatures) and only a problem for bloated transactions (you know such as what will be noticeable when lightning hubs start dealing with many signatures) where the average and even your magical median increase.
No, neither one holds true. Nobody ever claimed that the quadratic scaling is problematic with normal transactions. It is just an attack vector that could be abused by a malicious miner. "Could be" doesn't mean that it "will be" though.

block: 364292 validated in 2 seconds..
as oppose to pre libsecp256k1 which took far far far longer.
Where is the source of this result ("2 seconds")?

other standard blocks 0-400,000 validated in under 6 hours (18 blocks per second)
as oppose to pre libsecp256k1 which took far far far longer.
That doesn't matter. As previously stated, while libsecp256k1 improves validation time significantly it does not mitigate the attack vector at 2 MB (Gavin even confirmed this somewhere).
legendary
Activity: 4410
Merit: 4766
sorry to tell you this but 2mb is actually healthy right now..
libsecp256k1 alone (released ages ago) has actually made enough efficiency boost to make 2mb blocks healthily able to run on a raspberry pi.
No, that's not right (if it was right, there would be no need for sigops limitations in Bitcoin Classic). I've addressed this issue about 6 months ago already:
Quote from: Lauda
Quote
in order to sign or verify the tx, each input has to construct a special version of the tx and hash it. so if there are n inputs there are nxn hashes to be done. hence quadratic scaling.
the TLDR I believe is: ecdsa operations are the most computationally expensive part of verifying transactions, for normal, small size transactions, but they scale linearly with the size (number of inputs).whereas if a transaction in current bitcoin has tons of inputs, the bottleneck moves over to the hashing/preparing data to to be signed, because that time depends on the *square* of the number of inputs.
so usually it's ultra small, but it blows up for large N inputs.
Why doesn't libsecp256k1 have an effect on this?
Quote
because libsecp256k1 is an ECC library so it's only the "ecdsa" part in the above.
I don't know why you keep persisting that it doesn't. You should run a benchmark on that raspberry PI and validate only the block full of spam that was mined by F2Pool back in 2015 or 2014 (it was mentioned in the conference).

i was gonna waffle but..

block: 364292 validated in 2 seconds..
as oppose to pre libsecp256k1 optimisation, which took far far far longer.

other standard blocks 0-400,000 validated in under 6 hours (18 blocks per second)
which took far far longer pre libsecp256k1 optimisation

yes pre-libsecp256k1 estimated 11 minutes for an 8mb block(without the antispam sigops/hash limits)..
but due to the antispam limits, the only way to fill an 8mb block with spam is as 8x 1mb transactions..
which libsecp256k1 optimisations would bring about ~16 seconds validation time for 8mb (1mb processing repeated 8 times)
but we all know that for the last year, 8mb-20mb was not rational and 2mb was rational and overall approved. your buddy carlton approved it and even today your saying 2mb is ok...

so based on 2mb blocks of spam would require tweaking out some of the antispam limits to perform the test based on old metrics. but lets say with the anti-spam limits inplace it was treated as block364292 repeated twice (the only way to get 2mb of spam is to do 2 transactions to stay under the sigop/hash limits).

which would result in a 4second validation time.

so its not exactly killing a raspberry pi.. compared to the non antispam, non optimised doomsday pitch you and your cohorts of last year were crying by saying 11minutes+..

i do agree that linear would make a 4second spam block into a 10milisecond validation time just like any non spam transaction block.

but 4 seconds for the rare occassion is not killing bitcoin
and is definitely not anywear near a doomsday pitch "more than 10 minutes of DDoS via spam" Armageddon war-cry
legendary
Activity: 2674
Merit: 2965
Terminated.
sorry to tell you this but 2mb is actually healthy right now..
libsecp256k1 alone (released ages ago) has actually made enough efficiency boost to make 2mb blocks healthily able to run on a raspberry pi.
No, that's not right (if it was right, there would be no need for sigops limitations in Bitcoin Classic). I've addressed this issue about 6 months ago already:
Quote from: Lauda
Quote
in order to sign or verify the tx, each input has to construct a special version of the tx and hash it. so if there are n inputs there are nxn hashes to be done. hence quadratic scaling.
the TLDR I believe is: ecdsa operations are the most computationally expensive part of verifying transactions, for normal, small size transactions, but they scale linearly with the size (number of inputs).whereas if a transaction in current bitcoin has tons of inputs, the bottleneck moves over to the hashing/preparing data to to be signed, because that time depends on the *square* of the number of inputs.
so usually it's ultra small, but it blows up for large N inputs.
Why doesn't libsecp256k1 have an effect on this?
Quote
because libsecp256k1 is an ECC library so it's only the "ecdsa" part in the above.
I don't know why you keep persisting that it doesn't. You should run a benchmark on that raspberry PI and validate only the block full of spam that was mined by F2Pool back in 2015 or 2014 (it was mentioned in the conference).
legendary
Activity: 4410
Merit: 4766
I'm also not aware of any "false doomsdays or fake stats" that were used as arguments against any of those BIPs.

Yes, I'm talking about ridiculous block sizes. Examples of those being suggested by irrational people: 8 MB, 20 MB (remember 2015 20 MB now or doomsday?), unlimited. 2 MB should be fine after validation time is linear and not quadratic.

sorry to tell you this but 2mb is actually healthy right now..
libsecp256k1 alone (released ages ago) has actually made enough efficiency boost to make 2mb blocks healthily able to run on a raspberry pi.

the debate about linear validation. is absolutely nothing to do with bitcoin right now. absolutely nothing about a 2mb block of traditional bitcoin transaction, but only a worry in regards to when lightning network is active, which is not any time soon.

ill emphasise it again, linear signatures only really see a benefit in regards to multisigs which will gain dominance when lightning network is activated, but offers little to no noticeable difference to.... wait.. you may pick up your own rhetoric here... "median" size transactions(226byte), or as i say average transactions(~450byte)

(and im being subtle, but hopefully previous conversations give you enough of a hint)
so try not to back track on average/median transaction sizes to now argue about large 1kbyte multisigs being a big deal right now
legendary
Activity: 2674
Merit: 2965
Terminated.
ridiculous blocksizes??

sorry but segwit is 4mb.. i proven that in a different topic and even supplied you a link.
so how is 2mb ridiculous??
Yes, I'm talking about ridiculous block sizes. Examples of those being suggested by irrational people: 8 MB, 20 MB (remember 2015 20 MB now or doomsday?), unlimited. 2 MB should be fine after validation time is linear and not quadratic.

i actually agree with lauda's sentiment..
Additionally, I strongly believe that any controversial fork is an altcoin in a sense where a very small minority split away from the majority. We define pretty much any coin that does this as an altcoin; the primary difference is that they: 1) Change some specifications; 2) Don't retain the current blockchain information but rather start 'fresh'.

well there already has been logical bips submitted, and the decision against implementing it was not vetoed out due to code, or logic or real reason. but due to false doomsdays and fake stats.. mainly illogical scare stories
You can't get something merged into Core without consensus. What you could do is: Write a BIP -> provide the code for it -> if it gets rejected, then provide your own implementation for it. I'm also not aware of any "false doomsdays or fake stats" that were used as arguments against any of those BIPs.
Pages:
Jump to: