Gold collapsing. Bitcoin UP.

[satoshis_sockpuppet]

this can't be good:

https://news.microsoft.com/2018/06/04/microsoft-to-acquire-github-for-7-5-billion/

Why not? Github wasn't a very trustworthy company, why should MS make that worse?

 

[cypherdoc]

@satoshis_sockpuppet

every worthwhile open source project i know of is on there. ownership influence by MS wouldn't be a good thing.

 

[bitsko]

I'm sure everyone remembers sourceforge, they were rather important at one time too.

 

[lunar]

@Peter R Interesting difference of opinion? Care to expand a little on your logic?

https://coingeek.com/coingeek-nchain-announce-miners-choice-initiative-bitcoin-cash-transaction-fees-values/

CoinGeek and nChain’s mining firms plan to adjust settings in the software running on their BCH mining units to make two changes:

(1) Remove the current minimum “dust limit” of 546 Satoshi. This will makes it is possible for users to send as little as 1 Satoshi (one hundred millionth of a single Bitcoin Cash coin) via a BCH transaction. Any miner can accept a 1 Satoshi transaction amount, and it should be included in a valid block that all other miners will accept.

(2) Accept some free (zero mining fee) transactions in each block they mine. CoinGeek and nChain believe that miners should be able to accept free transactions, rather than requiring users to pay a minimum of 1 Satoshi per byte. Their mining firms will each designate a number of free transactions they will accept in each mined block.

How would you go about better understanding the costs (what costs and to whom?) without real market conditions attached? The burden here, as with all other supply and demand issues for blockspace, falls clearly with the miners. Accepting single satoshi transactions would have many benefits for coloured coins, tokens and UTXO consolidation.

Your argument seems to be predicated with Core reasoning. 'if we remove the blocksize limit, the sudden rush would bloat the chain, or won't somebody care for those poor home nodes ' Well after a year that didn't happen, and home nodes are redundant. This then, is same argument over again; as if miners won't act economically rational.

Let's say worst fears happen and suddenly millions of 1sat TX are broadcast, any rational miner will just filter them out, rather than waste time (orphan risk) verifying them all. How about combining this initiative with the original coin age parameter to prevent people churning out chains of DoS Txs? Well for now we still have the 32MB limit in place, would this not act as a good test buffer?

I see a future where miners are compensated for including transactions that are not economical at face value. Imagine a stock exchange or streaming service using coloured coins could pay a dedicated miner to include the days trades in one batch?

Anyway, this initiative would be great for adoption and new economic use cases. We've all spent the last 3 years debunking what ifs. How about more, let's try and see?

 

[Peter R]

I'm suggesting that we do some stress testing with very large UTXO sets before we implement a policy change that could result in a big spike in the size of the UTXO database. Before we increased the block size limit (first to 8 MB and then to 32 MB) we did lots of research and testing until we understood the implications on block propagation, validation times, etc.

"Let's say worst fears happen and suddenly millions of 1sat TX are broadcast, any rational miner will just filter them out, rather than waste time (orphan risk) verifying them all."

A miner is human on a long time scale, but a machine on a short time scale.

That is, over a long time scale, you can view a miner as a rational actor because the human controlling the mining operation can react to unforeseen events or fix things that break. But over a short time scale, the mining equipment simply does whatever it is programmed to do, even if that is to segmentation fault and crash.

For example, it was not rational for the chain to split a few years ago when a few miners were SPV mining. And it was not rational for Bitcoin.com to mine the 1.00003 MB block that got orphaned. But these things happened because that's what the mining equipment was programmed to do.

So I think it is unwise to think that bad things can't happen because miners are incentivized not to let bad things happen. It is prudent then to make changes slowly.

 

[SPAZTEEQ]

Is the 2nd change, re min tx fees also controversial. Interesting that these are being put forward together.

 

[Norway]

Can someone put this in a (collapsing) gold frame?

 

[Norway]

I'm suggesting that we do some stress testing with very large UTXO sets before we implement a policy change that could result in a big spike in the size of the UTXO database. Before we increased the block size limit (first to 8 MB and then to 32 MB) we did lots of research and testing until we understood the implications on block propagation, validation times, etc.

"Let's say worst fears happen and suddenly millions of 1sat TX are broadcast, any rational miner will just filter them out, rather than waste time (orphan risk) verifying them all."

A miner is human on a long time scale, but a machine on a short time scale.

That is, over a long time scale, you can view a miner as a rational actor because the human controlling the mining operation can react to unforeseen events or fix things that break. But over a short time scale, the mining equipment simply does whatever it is programmed to do, even if that is to segmentation fault and crash.

For example, it was not rational for the chain to split a few years ago when a few miners were SPV mining. And it was not rational for Bitcoin.com to mine the 1.00003 MB block that got orphaned. But these things happened because that's what the mining equipment was programmed to do.

So I think it is unwise to think that bad things can't happen because miners are incentivized not to let bad things happen. It is prudent then to make changes slowly.

Regarding dust limits: They should be removed, because the cost of doing the 1 satoshi transactions is regulated by tx fees. T'sup with magic numbers and minimum amounts? This is very weired coming from you, Peter.
[doublepost=1528237850][/doublepost]We need to get BCH investors as new members in BU.

No more code wizards or cheerleaders.

 

[Peter R]

@Norway: the dust limits are not consensus level. Individual miners can already override them if they choose to.

I'd like to see massive on-chain scaling and I'd like to see the dust limits fully removed in the future. But to the extent that miners might listen to me, I'd suggest they get here by taking small incremental steps rather than losing balance with a few giant leaps.

 

[lunar]

Peter has valid concerns, new equipment should never tested with all the dials set to maximum. However i'm a firm believer in 'necessity is the mother of all invention'. The dust limit is an individual miners choice, with the majority of wallets currently unable to send a single satoshi anyway. As more miners adopt this strategy, it will be an incentive for the rest to improve their processing technology. Keep up or loose out.

Gavin wrote a little on this here, https://gist.github.com/gavinandresen/f209a02ee559905aa69bf56e3b41040c#file-utxo_bitvector-md

"Back-of-the-envelope calculations for gigabyte-sized blocks: A gigabyte block full of 250-byte 2-input, 2-output transactions would have 4 million outputs, which would be a 0.5 megabyte bit-vector. 10 years of gigabyte-sized blocks would fit in 263 gigabytes of RAM, which is reasonable for a powerful server (servers with terabytes of RAM are common these days)."

So long as only a small % of the blockspace contains completely free transactions, the resistance created by small Tx fees, and the free market should take care of the rest. This looks more like a group of miners leading by example. They've seen something that will create value for the BCH network and they've done something about it.. Permissionless innovation at it finest.

Will the rest of the miners immediately adopt this strategy? Undoubtedly no. After a few months without complications, and the innovation that will bring, more are sure to follow.

​

​

 

[AdrianX]

Is the 2nd change, re min tx fees also controversial. Interesting that these are being put forward together.

these two things are related they got lumped together when Core moved from a cost per transaction model to a cost per byte model.
[doublepost=1528244772][/doublepost]

Regarding dust limits: They should be removed, because the cost of doing the 1 satoshi transactions is regulated by tx fees. T'sup with magic numbers and minimum amounts? This is very weired coming from you, Peter.

The issue is you can spam the 32MB block with 1 sat transactions. I agree with Peter. You don't just want to make that possible, a DOSS attack with 1 say transactions could go on indefinitely and there are dozens of people who would love to do it just to make the argument that 1MB prevents spam.

 

[79b79aa8]

those 1 sat txs will still pay 1 sat/byte, or whatever the going rate, to ensure inclusion in a block. if you pay nothing, you are guaranteed nothing. whoever pays is by definition not spamming, but paying-to-use.

It’s based on open market competition, and there will probably always be nodes willing to process transactions for free.

miners are the best and only assessors of their own risk/reward ratio.

 

[lunar]

@Norway

@DanielKrawisz has been known to frequent these parts.

I tend to agree, we need more investors and miners. I wouldn't object to some sort of proof of reserves for new members, but only if it could be done in a private manner. I hear several of the Chinese chat groups have a 100BCH as a pre-requisite.

 

[Norway]

@Norway: the dust limits are not consensus level. Individual miners can already override them if they choose to.

I'd like to see massive on-chain scaling and I'd like to see the dust limits fully removed in the future. But to the extent that miners might listen to me, I'd suggest they get here by taking small incremental steps rather than losing balance with a few giant leaps.

I don't get this dust limit. It's not related to either blocksize or UTXO size. What am I missing?

@lunar Yeah, I've been thinking about this "new members" issue lately. My initial stance was elect "good guys". But now, I think investors should be the prefered new members.

Some devs are really crazy. They just don't get the full picture. (Not talking about current BU devs now.)
[doublepost=1528250427,1528249701][/doublepost]I do regret that I voted YES to get @sken as a member. Turns out he thinks abortion is murder. Even a few seconds after the sperm hits the egg.

I don't want BU to be a bunch of young, extreme guys. I like the fact that we are older and wiser than the rest of the space.

But more important: Members should have a stake and care. That's why I will promote investors as new members in the future, and not just people who say they agree. Time has changed.

 

[cypherdoc]

The dust limit ought to be removed

 

[Christoph Bergmann]

I don't get this dust limit. It's not related to either blocksize or UTXO size. What am I missing?

you can imagine the dust limit as a minimal weight of a physical coin, and the UTXO-set a list of all coins circulating. If you remove the minimal weight, it becomes possible to flood the market with extremely tiny coins and thus increase the lengths of the list of coins.

Also, the dust limit is some kind of minimum price an attacker has to pay for polluting the UTXO-set. Blowing up UTXO seems to be one of the worst attacks on Bitcoin, as nodes can't delete the UTXO set and need to access it fastly.

I think if UTXO-commitments become a thing, it eats more CPU power to calculate them when the UTXO-set is bigger, but this is just an uninformed guess.

Anybody has an idea if it is possible to prune the UTXO-set? @Tom Zander had some interesting insights into loading it in other databases instead of leveldb, which might be searchable (?) or can be filtered (?). For example, if I run a purely financial node, I don't care about UTXO which are smaller than dust or have just the purpose to hammer messages / pictures on the blockchain. So it would be nice to have the option to delete them. I guess this would need some kind of proof that these UTXO are legit. Maybe this could be a subset / another kind of UTXO commitment ...

 

[Zarathustra]

The Trend Is Your Friend ...

https://www.tradingview.com/chart/BCHBTC/k1oQu8js-Bitcoin-Cash-The-Trend-Is-Your-Friend/

 

[lunar]

Thanks for the tip off about this guy, @Zangelbert Bingledack he's great.


Debunking Vitalik Buterin on Decentralization

(perhaps skip the little rant at the end of vid)

 

[Tom Zander]

Free transactions have existed since the beginning, and we should eventually re-enable them on the mainnet. It is wrong to conclude that free transactions will fill up blocks or flood the network because for years that has been protected against by rate-limiting them.

That code is still operational in all (satoshi originated) codebases today. (assuming they didn't remove it manually)

 

[Tom Zander]

Why not? Github wasn't a very trustworthy company, why should MS make that worse?

I've seen Microsoft purchase various companies or services which head downhill quite fast afterwards. They purchased Nokia & Skype as two good examples.

I have no doubt github will get major changes to its services over the next year(s).

I've started playing with gitlab, looks pretty good. Much improved since I last tried it.

Is BU interested in trying gitlab?

ps. Whohoo! post 24000