[XMR] Monero Speculation

[wachtwoord]

...

Thanks. That means it's not solved at all (which is ashame of course, but good for Monero).

I do like they dont have a tail emission (I hate that in Monero, smells like Keynsianism, just secure the chain with fees).

The tail emission of Monero places the inflation rate of Monero below the historical inflation rate of gold which after all is the "gold standard" of hard money of Austrian school economists. If anyone has a solution to both secure the chain only with fees and at the same time allow for an adaptive blocksize limit there are many in the Bitcoin community that would love to see this solution. This is after all the real issue behind the blocksize debate in Bitcoin.

 

Securing the chain only with fees is fixed. Just don't increase the block size. People don't seem to listen to this because they want their damn coffee on the chain for some reason.

With Bitcoin you can make secure payments. You have to pay for that. Only make transactions that need that high level of security.

[owm123]

The GUI project is fully funded now in less then a day !!

https://forum.getmonero.org/8/funding-required/2476/the-official-qt-gui-project

What is the expected release date for this?

[dnaleor]

...

Thanks. That means it's not solved at all (which is ashame of course, but good for Monero).

I do like they dont have a tail emission (I hate that in Monero, smells like Keynsianism, just secure the chain with fees).

The tail emission of Monero places the inflation rate of Monero below the historical inflation rate of gold which after all is the "gold standard" of hard money of Austrian school economists. If anyone has a solution to both secure the chain only with fees and at the same time allow for an adaptive blocksize limit there are many in the Bitcoin community that would love to see this solution. This is after all the real issue behind the blocksize debate in Bitcoin.

 

Securing the chain only with fees is fixed. Just don't increase the block size. People don't seem to listen to this because they want their damn coffee on the chain for some reason.

With Bitcoin you can make secure payments. You have to pay for that. Only make transactions that need that high level of security.

small bocks will lead to bitcoin banks who will process transactions for you.
Not really desirable imho.

[wachtwoord]

...

Thanks. That means it's not solved at all (which is ashame of course, but good for Monero).

I do like they dont have a tail emission (I hate that in Monero, smells like Keynsianism, just secure the chain with fees).

The tail emission of Monero places the inflation rate of Monero below the historical inflation rate of gold which after all is the "gold standard" of hard money of Austrian school economists. If anyone has a solution to both secure the chain only with fees and at the same time allow for an adaptive blocksize limit there are many in the Bitcoin community that would love to see this solution. This is after all the real issue behind the blocksize debate in Bitcoin.

 

Securing the chain only with fees is fixed. Just don't increase the block size. People don't seem to listen to this because they want their damn coffee on the chain for some reason.

With Bitcoin you can make secure payments. You have to pay for that. Only make transactions that need that high level of security.

small bocks will lead to bitcoin banks who will process transactions for you.
Not really desirable imho.

For small transactions extremely desirable. For big transactions not. It works out perfectly

[dEBRUYNE]

The GUI project is fully funded now in less then a day !!

https://forum.getmonero.org/8/funding-required/2476/the-official-qt-gui-project

What is the expected release date for this?

Not a specific one. He will start with 10 hours per week, until his other project is finished. After that, he will work 20 hours (perhaps even more) per week. The announcement states this as well:

Ilya will work 10 hours per week at the beginning, expanding into more hours per week later when some other project is finished.

The announcement also states:

The hours that remain after having completed the GUI will be used to work on other features. There should be a better estimate of hours needed to complete the GUI once the work really has started.

My personal guess is +- 2-3 months. Take this with a bit of a grain of salt though.

[GingerAle]

The GUI project is fully funded now in less then a day !!

https://forum.getmonero.org/8/funding-required/2476/the-official-qt-gui-project

What is the expected release date for this?

Not a specific one. He will start with 10 hours per week, until his other project is finished. After that, he will work 20 hours (perhaps even more) per week. The announcement states this as well:

Ilya will work 10 hours per week at the beginning, expanding into more hours per week later when some other project is finished.

The announcement also states:

The hours that remain after having completed the GUI will be used to work on other features. There should be a better estimate of hours needed to complete the GUI once the work really has started.

My personal guess is +- 2-3 months. Take this with a bit of a grain of salt though.

So another 2-3 months of low hashrate and low prices. I wonder at which pricepoint the bonnets and cloud-cowboys will kick back on. And 2-3 months brings us to the hardfork with 2 minute blocks (with double blockrewards compared to now), so in the realm of 12-14 xmr per block again. Price actions gonna get bumpy.... if the GUI is really all that big of a deal.

[dEBRUYNE]

The GUI project is fully funded now in less then a day !!

https://forum.getmonero.org/8/funding-required/2476/the-official-qt-gui-project

What is the expected release date for this?

Not a specific one. He will start with 10 hours per week, until his other project is finished. After that, he will work 20 hours (perhaps even more) per week. The announcement states this as well:

Ilya will work 10 hours per week at the beginning, expanding into more hours per week later when some other project is finished.

The announcement also states:

The hours that remain after having completed the GUI will be used to work on other features. There should be a better estimate of hours needed to complete the GUI once the work really has started.

My personal guess is +- 2-3 months. Take this with a bit of a grain of salt though.

So another 2-3 months of low hashrate and low prices. I wonder at which pricepoint the bonnets and cloud-cowboys will kick back on. And 2-3 months brings us to the hardfork with 2 minute blocks (with double blockrewards compared to now), so in the realm of 12-14 xmr per block again. Price actions gonna get bumpy.... if the GUI is really all that big of a deal.

That remains to be seen. Fact is though that a lot of people have been either put off by the CLI or would wait before a more usable wallet would be out (a GUI).

[iCEBREAKER]

Some sort of anonymous coin WILL succeed, if cryptocurrencies are here to stay. With people like Mike Hearn declaring Bitcoin has failed, and really traction slowing down, it's not an absolute guarantee that cryptocurrencies won't become a passing fad yet.

Hearn's whiny ragequit is a contrarian indicator (ie good news for Bitcoin).

No idea wtf you are talking about wrt "traction slowing down."  Did you miss Blockstream's $55 million Series-A announcement and GBTC's huge premium?

The cryptocurrency genie isn't going back in the bottle.  The e-cash toothpaste can't be put back into the tube.

The blockchain horse is out of the barn, but by all means consider shutting the gate if you enjoy that sort of thing.   

[iCEBREAKER]

The GUI project is fully funded now in less then a day !!

https://forum.getmonero.org/8/funding-required/2476/the-official-qt-gui-project

Well that escalated quickly!  Even though we've seen that movie before, the magnanimous Monero Mustangs' monumental munificence continues to astound me.

What is the expected release date for this?

[iCEBREAKER]

small bocks will lead to bitcoin banks who will process transactions for you.
Not really desirable imho.

For small transactions extremely desirable. For big transactions not. It works out perfectly

Exactly.  Nice catch, noticing dnaleor failed to distinguish between (the economics of) small and large tx.

The mistaken notion that trustless payment channels like Lightning are "bitcoin banks" is equally invalid.

Even if it was true, the "bitcoin banks" would be kept sound and honest by the transparent/auditable nature of the blockchain(s).

Finally, let's reiterate that large blocks alone cannot even begin to get us anywhere even remotely near Visa scale, no matter how we wish they could.

[ArticMine]

...
Finally, let's reiterate that large blocks alone cannot even begin to get us anywhere even remotely near Visa scale, no matter how we wish they could.

I respectfully disagree with the above statement and Monero is extremely well positioned to prove it false. All I will say is that when one has written code for a mainframe computer with a total memory capacity of 2 MB using punched cards as the data input method one has a very different perspective on the subject.  

Edit: By comparison a mini computer of the day would have a total memory capacity of about 2 KB and use a teletype and paper tape reader / punch as the data input output methods. I also programmed on those.

[iCEBREAKER]

large blocks alone cannot even begin to get us anywhere even remotely near Visa scale

I respectfully disagree with the above statement and Monero is extremely well positioned to prove it false. All I will say is that when one has written code for a mainframe computer with a total memory capacity of 2 MB using punched cards as the data input method one has a very different perspective on the subject.  

Edit: By comparison a mini computer of the day would have a total memory capacity of about 2 KB and use a teletype and paper tape reader / punch as the data input output methods. I also programmed on those.

That's nice.  Reminds me of my childhood, making what I would later find out are called "Demos" using BASIC on a 256k Tandy XT with no hard drive.  And there was no mouse, so using the paint program required copious amonts of arrow keys and space bar.  

Alas, our ancient history and nostalgic reveries have no bearing on the current fact that sig_op verification scales as a quadratic of the tx size.

You are entitled to your opinion, but not your own facts.  EG:

The Toomininstas are confronting the same problem the Gavinistas did, which is that multiplying a tiny number such as 3tps by another tiny (ie sane) number such as 2 or 4 or even 8 still only produces another tiny number such as 6tps or 12tps or 24tps.

You can't get to Visa tps from here.  Our only realistic path to Visa is orthogonal scaling, where each tx does the maximum economic work possible.

Right, card payments are currently at around 5000 TPS on a _year round_ average basis, with highest day peaks probably at over 100k TPS. And that is now, these figures have been rapidly growing.

These are numbers high enough that just signature processing would completely crush even high end commercially available single systems.  Even if you took some really great drugs and believed plain-old-bitcoin could get anywhere near matching that in the near future while having any decentralization remaining.... so what? it wouldn't be close again after just a couple more years growth.

A trip to the moon requires a rocket with multiple stages or otherwise the rocket equation will eat your lunch... packing everyone in clown-car style into a trebuchet and hoping for success is right out.

Combined the major card networks are now doing something on the other of 5000 transactions per second on a year round average; and likely something on the order of 120,000 transactions per second on peak days.

The decentralized Bitcoin blockchain is globally shared broadcast medium-- probably the most insanely inefficient mode of communication ever devised by man. Yet, considering that, it has some impressive capacity. But relative to highly efficient non-decentralized networks, not so much. The issue is that in the basic Bitcoin system every node takes on the whole load of the system, that is how it achieves its monetary sovereignty, censorship resistance, trust cost minimization, etc. Adding nodes increases costs, but not capacity. Even the most reckless hopeful blocksize growth numbers don't come anywhere close to matching those TPS figures. And even if they did, card processing rates are rapidly increasing, especially as the developing world is brought into them-- a few more years of growth would have their traffic levels vastly beyond the Bitcoin figures again.

No amount of spin, inaccurately comparing a global broadcast consensus system to loading a webpage changes any of this.

So-- Does that mean that Bitcoin can't be a big winner as a payments technology? No. But to reach the kind of capacity required to serve the payments needs of the world we must work more intelligently.

IIRC, Monero's sig_op/verification is vastly improved over Bitcoin's first generation technology, but there is also an elastic (ie scaling) reward penalty for large blocks.

So the idea that Monero, using some kind of Giant Block Superweapon, is going to dethrone Bitcoin, is not supported by the facts.

Even if Monero gradually and without penalty accommodates larger blocks, they will never be able to shovel 5k-120k tps onto Layer One.

[wpalczynski]

Staying out of that discussion but I have to say I love the picture!!

[smooth]

Even if Monero gradually and without penalty accommodates larger blocks, they will never be able to shovel 5k-120k tps onto Layer One.[/b]

That's a pretty big range. 5K is somewhat more plausible than 120k, for longer term.

5K can fit in a 100 megabit connection and require something like a large hard drive a month to store the blockchain. Not going to be home-based nodes at that point but it could still fit easily within satoshi's vision of many independent nodes in data centers, even small data centers (10 years is only 100 hard drives -- not really much a storage array). Ignoring future improvements in storage technology which I think is ArticMine's point.

Considering a peak 60x higher means you would need 6 gigabit connections, also plausible for a small-ish data center. As long as the peak isn't sustained too long, storage wouldn't become a problem, and verification can be almost arbitrarily parallelized.

I agree with your main point about there being better ways to scale than jamming everything into one chain though.

Small point: The block size can't grow at all without some penalty. Despite what the white paper says the penalty kicks in at the median size, not 10% above the median. The implementation has always worked that way.

[iCEBREAKER]

Even if Monero gradually and without penalty accommodates larger blocks, they will never be able to shovel 5k-120k tps onto Layer One.[/b]

That's a pretty big range. 5K is somewhat more plausible than 120k, for longer term.

5K can fit in a 100 megabit connection and require something like a large hard drive a month to store the blockchain. Not going to be home-based nodes at that point but it could still fit easily within satoshi's vision of many independent nodes in data centers, even small data centers (10 years is only 100 hard drives -- not really much a storage array). Ignoring future improvements in storage technology which I think is ArticMine's point.

Considering a peak 60x higher means you would need 6 gigabit connections, also plausible for a small-ish data center. As long as the peak isn't sustained too long, storage wouldn't become a problem, and verification can be almost arbitrarily parallelized.

I agree with your main point about there being better ways to scale than jamming everything into one chain though.

Small point: The block size can't grow at all without some penalty. Despite what the white paper says the penalty kicks in at the median size, not 10% above the median. The implementation has always worked that way.

I'll take that as confirmation Monero can verify 5K tps on something like a reasonably priced server.  What about 120k?  ~25 Xeon cluster?

Once crypto has defeated fiat and we all live in Libertopia, I'll have much less of a problem with full nodes residing mostly in data centers.

Of course there is every chance Libertopian homes will have full duplex 6 gigabit connections.

Thanks for nifty insight into the spec vs reality of block growth penalties.  It's nice to know Monero is protected from block bloating attacks, even if well-intentioned!   

[luigi1111]

...

Thanks. That means it's not solved at all (which is ashame of course, but good for Monero).

I do like they dont have a tail emission (I hate that in Monero, smells like Keynsianism, just secure the chain with fees).

The tail emission of Monero places the inflation rate of Monero below the historical inflation rate of gold which after all is the "gold standard" of hard money of Austrian school economists. If anyone has a solution to both secure the chain only with fees and at the same time allow for an adaptive blocksize limit there are many in the Bitcoin community that would love to see this solution. This is after all the real issue behind the blocksize debate in Bitcoin.

 

Securing the chain only with fees is fixed. Just don't increase the block size. People don't seem to listen to this because they want their damn coffee on the chain for some reason.

With Bitcoin you can make secure payments. You have to pay for that. Only make transactions that need that high level of security.

small bocks will lead to bitcoin banks who will process transactions for you.
Not really desirable imho.

For small transactions extremely desirable. For big transactions not. It works out perfectly

This is of course true for some definition of "small" and "big" (and I guess "perfectly" too).

If Bitcoin continued growing while keeping the same limit indefinitely, eventually ~nobody would like those definitions. 1 MB Bitcoin has approximately enough space for 1% of the world's population to make 1 transaction annually.

Whether Bitcoin will continue growing (in terms of value transferred), and for how long, is an open and interesting question.

[smooth]

I'll take that as confirmation Monero can verify 5K tps on something like a reasonably priced server.  What about 120k?  ~25 Xeon cluster?

In NoodleDoodle's performance commit he noted a benchmark of 2.5ms/tx on i7-2600. That's 400 tx/sec on a 2011 desktop. A reasonably priced current-gen server (say dual-Xeon 10-core CPUs) is probably several times faster so close to 5K/sec, but I don't know the exact numbers. There is more optimization available still (we aren't using the most optimized elliptic curve asm library available from Bernstein for example, just his sort-of-optimized C library).

With the move to ringCT, it will probably be different (though some of the differences will offset, such as having fewer outputs/tx), and we will have to reevaluate.

[luigi1111]

Even if Monero gradually and without penalty accommodates larger blocks, they will never be able to shovel 5k-120k tps onto Layer One.[/b]

That's a pretty big range. 5K is somewhat more plausible than 120k, for longer term.

5K can fit in a 100 megabit connection and require something like a large hard drive a month to store the blockchain. Not going to be home-based nodes at that point but it could still fit easily within satoshi's vision of many independent nodes in data centers, even small data centers (10 years is only 100 hard drives -- not really much a storage array). Ignoring future improvements in storage technology which I think is ArticMine's point.

Considering a peak 60x higher means you would need 6 gigabit connections, also plausible for a small-ish data center. As long as the peak isn't sustained too long, storage wouldn't become a problem, and verification can be almost arbitrarily parallelized.

I agree with your main point about there being better ways to scale than jamming everything into one chain though.

Small point: The block size can't grow at all without some penalty. Despite what the white paper says the penalty kicks in at the median size, not 10% above the median. The implementation has always worked that way.

I'll take that as confirmation Monero can verify 5K tps on something like a reasonably priced server.  What about 120k?  ~25 Xeon cluster?

Once crypto has defeated fiat and we all live in Libertopia, I'll have much less of a problem with full nodes residing mostly in data centers.

Of course there is every chance Libertopian homes will have full duplex 6 gigabit connections.

Thanks for nifty insight into the spec vs reality of block growth penalties.  It's nice to know Monero is protected from block bloating attacks, even if well-intentioned!  

I did some thumb-sucking math a number of pages ago: I "built" a single system server that exists today (8-way) that could process 40-something-k TPS IIRC.

Edit: found post: https://bitcointalk.org/index.php?topic=753252.msg12768096#msg12768096

This would suggest to me that a 120k TPS node would cost ~$200,000 for just the system(s).

Assuming a transaction size of 2 KB (made up), we'd need a ~2gbps link.

[dEBRUYNE]

I disagree with you on this one (and I think I owe you a response on Reddit, now that I think of it). In short, I think this is at best an open question, but if it turns out there is any orphan risk or marginal-cost-per-tx, and/or if miners don't actually compete, longrun, on mostly the commodity block-production anyway, then this just amounts to unnecessary economic meddling, and I hate it that people are so quick to justify this sort of thing.

I just fail to see how a fee market is going to develop. In either case, limited or unlimited blocksize, you need a huge amount of transactions per block to keep miners incentivized and the network secure. Let's take the example of 1 Exahash the network hashrate is around today. With an average price of 400 over the last month, miners earn ~$10k per block (taking this number for convenience). Coincidentally, Bitcoin's daily transaction fees hit almost $10K on a 200 day moving average. For convenience, I will take that number as well. From another chart, we also see that number of transactions just hit an all time high of 242,129 tx/day. That's ~ 2.8 tx p/s with blocks becoming kind of full on average and a limited blocksize. In sum, currently ~ 2.8 tx p/s is making up for ~$10k in transaction fees daily. When the block reward runs out, assuming miners are currently mining at equilibrium, we would need (6*24) * ~2.8 tx p/s = 403 tx p/s on the main chain to yield the same proceeds for miners as of today. That is conditional on most current variables remaining the same and blocksize being limited, since I think it would likely be a race to the bottom with an unlimited blocksize and no block rewards. I personally fail to see how to accomplish this, certainly with the push of lower value transactions to side-chains. Perhaps the fees of higher-value transactions will reduce the amount of tx per seconds needed, but it would still be a large number to accomplish. Also, bear in mind that if the network is doing such a high amount per transactions, we likely need more hashrate to make the network more secure.

Anyway, it will be an interesting experiment, that is for sure.

Links:
https://blockchain.info/charts/n-transactions
https://blockchain.info/charts/transaction-fees-usd?timespan=360days&showDataPoints=false&daysAverageString=200&show_header=true&scale=0&address=

That said, it's money supply transparency that matters most, so if a coins comes with a tail-emission out of the gate, that's fine (except for the above caveat about it suggesting the devs/community may be prone to over-engineering in the economic realm).

But in Monero's case, you tail-emission supporters broke *the* single biggest implicit contract: don't change the supply dynamics!! I loathe the fact that I bought-in to one economic model in 2014 and then a bunch of economic-meddlers decided to change the model out from under me. Yes, it's small relative to my holdings, but quite the slippery slope. That might as well be fiat, and the move *drastically* reduced my respect for the coin, the developers, and the community.

They did not do such a thing? Smooth already stated this earlier and on the earliest web.archive I could find it is also stated clearly -> http://web.archive.org/web/20140520215957/https://bitcointalk.org/index.php?topic=583449.0. See:

[2] Actual number of atomic units is M = 264 - 1. A minimum subsidy may be implemented in the future with <1% inflation to preserve mining incentives.

The actual number was determined later, but it was always made clear that there would be a tail emission with <1% inflation.

Seems minor to me, but hard for me to truly judge without being a cryptographer.

Well, if the cryptography breaks, the whole system breaks and it will be rendered worthless. Therefore, it is in my opinion extremely important. Furthermore, Bitcoin is also build on solid cryptography. Cryptography usually strengthens over the years, when it has been thoroughly tested, vetted and peer-reviewed. Relying on mature cryptography, opposed to an infancy one, is definitely an advantage in my opinion.  

This post of gmaxwell emphasizes the importance of cryptography:

Cryptography has never been a significant part of cryptocurrency - even though it may share the first few letters. It works on a system of digital signatures.

It would seem that you actually do not understand what cryptography is in the modern sense.

A fundamental nature of information is that it wants to be freely copied everywhere to everyone. That any bit is equal and indistinguishable from any other bit of the same value and that any bit is eventually known to all who care.  Cryptography is all that technology by which we hope to confine and constrain the nature of information, to put up fences and direct it to our exclusive purposes, against all attacks and in defiance of the seemingly (and perhaps actually) impossible.  Digital signatures are cryptography by any modern definition and utilize the same tools and techniques (for example, a DSA signature is a linear equation encrypted with an additively homorphic encryption), and suffer from most of the same challenges as the message encryption systems to which you seem to be incorrectly defining cryptography as equivalent.  Moreover, the use of digital signatures isn't the only (or even most relevant) aspect of cryptography in cryptocurrencies-- e.g. the prevention of double spending of otherwise perfectly copyable and indistinguishable information in a decentralized system is a cryptographic problem which we address using cryptographic tools, and-- like all other practical cryptography-- achieve far less than perfect confidence in our solution. As are more modest ends like interacting with strangers but not being subject to resource exhaustion from them.

Far more so than other sub-fields of engineering, cryptographic systems are doing something which is fundamentally at odds with nature and share an incredible fragility and subtly as a result (and perhaps all are failures, we have no proof otherwise).

A failure to understand and respect these considerations has resulted in a lot of harmful garbage and dysfunctional software.

In addition, it might take 10-20 years for Zcash's cryptography to become mature/vetted and considered "safe". Bear in mind that Zcash is also prone to (thus not safe from) quantum computing, which might by the time their cryptography is considered "safe", break their "system".


I agree with your other comments.

[iCEBREAKER]

In NoodleDoodle's performance commit he noted a benchmark of 2.5ms/tx on i7-2600. That's 400 tx/sec on a 2011 desktop. A reasonably priced current-gen server (say dual-Xeon 10-core CPUs) is probably several times faster so close to 5K/sec, but I don't know the exact numbers. There is more optimization available still (we aren't using the most optimized elliptic curve asm library available from Bernstein for example, just his sort-of-optimized C library).

With the move to ringCT, it will probably be different (though some of the differences will offset, such as having fewer outputs/tx), and we will have to reevaluate.

Sweet!  

When I tell people about this good news during my Monero evangelizing, how do I explain why our sig_ops are so much faster than Old Grandpa Bitcoin's?

Or should I even bother, since Bitcoin's verification is being fixed (and supercharged) with segwit?

I did some thumb-sucking math a number of pages ago: I "built" a single system server that exists today (8-way) that could process 40-something-k TPS IIRC.

Edit: found post: https://bitcointalk.org/index.php?topic=753252.msg12768096#msg12768096

This would suggest to me that 120k TPS would cost ~$200,000 for just the system(s).

Assuming a transaction size of 2 KB (made up), we'd need a ~1.9gbps link.

Super sweet!!!  

Watch your back, Visa. 

_Soon...