Gold collapsing. Bitcoin UP.

[AdrianX]

Well, in every system where you offer a service (may it be something like storj, or maybe something like amazon ec2, or an advertising network, or whatever) you need some kind of accounting.

Then you have two choices: charge some credit in the system, and trust it for good behaviour (that's fine, it's the actual model and it works well enough), but you can also imagine a system where you do not need to trust it.

Having this choice can be better in some scenarios, so I don't see why to dismiss it altogether.
Also, all other conditions met, if you can choose between them with no drawbacks, I don't see why anybody would choose to trust, if you don't have to.

I can think of lots of fringe financial products where you may want to lock money away in complex
schemes but I don't see any justification to make Bitcoin manage these fringe product's, their viability is not a necessary function of a good money (bitcoin).

To add to @Justus Ranvier point it's not credit when your pegging or exchanging one currency to another, its being held in trust.

Credit is when the company giving you credit takes the "money and uses it for something else and offers you credit in return. This existing system works well enough as you say. It has risks but they are well understood and managed.

The imaginary application for sidechains that you think of as "credit" or money held in trust is not practical for any of the user cases you present.

Having this choice doesn't come "with no drawbacks", the security trust/guarantee has to be provided some how. In bitcoin it's PoW and it's effective so long as the miners have an incentive secure the bitcoin system.

As I pointed out in my previous post one always needs to trust something adding another layer in between docent increase the trust.

 

[Erdogan]

The word trust should be limited to economic actors. You can of course say that you trust the weather to not sink the boat, but that is something else, it is nature, and it should be called risk.

 

[Justus Ranvier]

@Erdogan More precise than "trust" is "cost of attack."

Terms like "trust" and "trustless" are shorthand notation for saying "low cost of attack" and "high cost of attack" respectively, but frequently what happens in conversations is that meaning is lost and people start using them as if they meant "zero cost of attack" and "infinite cost of attack."

Any argument between alternative courses of action that takes as a premise that the cost of attack can either be zero or infinite is going to be unproductive, because both of those costs are impossible in reality.

It also makes comparison between different courses of action impossible, because everything is either one impossible value or the other.

If we drop the shorthand and work out a rational basis to estimate the costs of attacks (in non-binary terms), then we could actually make informed decisions.

 

[freetrader]

A new study by the Initiative for CryptoCurrencies and Contracts (IC3) at the Jacobs Technion-Cornell Institute authored by Christian Decker, Ittay Eyal, Andrew Miller and Emin Gün Sirer, among others, found that bitcoin’s blocksize could currently scale up to 4MB without affecting decentralization.

Source: https://www.cryptocoinsnews.com/cornell-study-recommends-4mb-blocksize-bitcoin/

PDF of study "On Scaling Decentralized Blockchains (A Position Paper)":
http://www.initc3.org/scalingblockchain/

 

[albin]

@freetrader

Now we just need to start placing bets on how long it takes for someone from Core to start yelling "spherical cow"!

 

[sickpig]

Just for the sake of procrastinating real work today I spent a bit of time posting about SegWit on BCT.

I can confirm, in case you have any doubts, that it is a total waste of time.

one last thing, @adamstgbit could please explain to me how is it possible that "blocks can be sent around without the individual sigs on each TX, and that block can still be validated without these sigs" ?

 

[adamstgbit]

@sigpig

you can validate the block as a whole without validating every TX within that block
once you know the block is valid, you know every TX within it is valid

 

[sickpig]

@adamstgbit

and how do you know is valid in the first place just because another full say so?

Isn't a valid block a group of valid txs?

 

[cypherdoc]

@sigpig

you can validate the block as a whole without validating every TX within that block
once you know the block is valid, you know every TX within it is valid

are you getting this concept from Lau's SW presentation where he says it's possible (not probable) that miners will start sending around blocks w/o the sigs of the tx's?

 

[adamstgbit]

@sickpig

idk, somthing about hashing the block with the previous block makes it impossible to fake a block.
i have expressed similar concerns and i was told it was nothing to worry about.

@cyherdoc

isn't that the thing? sigs will be dropped and forever lost after the block has XX confirms.
of course they will keep propagating blocks with sigs but onces the block "solidifies" there no use to keep the sigs.

 

[cypherdoc]

@adamstgbit

the sigs are only dropped by a full node after the block with sigs has been propagated to it. i sure as heck hope miners won't be pushing out blocks into the network w/o sigs in SW. altho i guess that is the concept behind XThins to a degree as they push out hashes instead.

 

[satoshis_sockpuppet]

So can someone explain to me why the miners are so eager to get rid of their income?
And who is going to run all those decentralizing nodes when you don't use the blockchain anymore?

 

[adamstgbit]

@adamstgbit

the sigs are only dropped by a full node after the block with sigs has been propagated to it. i sure as heck hope miners won't be pushing out blocks into the network w/o sigs in SW.

we agree.

but once dropped, a new node comes along and dwls all the old blocks. in that case nodes are sending blocks around without any sigs.

i think that is what @sickpig is referring to

 

[freetrader]

Please Slush, hit the switch:

 

[cypherdoc]

@adamstgbit

my understanding of how that works is that in SW, a node would first DL the entire blockchain with sigs from a bonafide full node to first verify it's integrity. then, it could pare down the blockchain data by discarding sigs. or even pare down the UTXO set if they want. thus, it becomes a "partially validating SW SPV node hybrid". seems like the sky is all pie as far as SW is concerned. i'm very skeptical of how this will all work out esp the ANYONECANSPEND tx's seen by old nodes. which if only the Satoshi 0.12 nodes upgrade to, leaves 75% of the network not upgrading to SW, potentially leaving unknown security concerns to be exploited.

 

[awemany]

Hey guys, after I actually accumulated some 10k points on reddit mostly from trying to wake people up against an endless stream of (paid) idiocy, I really felt that just about everything has been said in the blocksize debate and further talk doesn't help anyone. Don't expect me to be around a lot, I am quite busy with my daily life.

Let me just say I am still optimistic and holding my few coins, and that is last but not least due to @rocks proposal for a scheduled hardfork. I have by far not enough coins to retire - and that unfortunately means not enough coins to follow my passion of more fighting in and providing more help to win this war.

But in any case, @rocks' Bitcoin Original will allow some additional heat on the Miners of Moronia and an ejection seat to avoid being crushed in the quite possible train wreck as well.

Ledger vs. protocol and all that.

@jl777, I think I saw your nick somewhere else (BCT?) before, but you appear to be new here? In any case, your work on building a simple mmap()able structure to store Bitcoin data in is exactly what I had in mind would or should happen to scale Bitcoin up. So continue your great work and thanks for that!
I would love to work with you on that but I simply do not have the monetary (and thus time-wise) resources to do that right now.


@theZerg, @Peter Tschipper: Awesome that you can get from San Jose to Frankfurt in 250ms including all processing. Not long ago, the raw ping delay was about that.

I realized that I shouldn't characterize all devs and nerds that way when I wrote that. It's just that I made up that meme years ago that has played out imo:

"The geeks fail to understand that which they hath created."

Corrupt devs is more accurate for sure.

You give them too much credit here. Satoshi created Bitcoin.. 99.9% of what is Bitcoin was in place by 2009. The rest is optimization, edge cases and so forth. It is eerily impressive how well the self-election to 'wizards of Bitcoin' worked so far.

The excitement about the future, the limitless possibilities, it's all just been turning to ash lately. The gmax treehouse is just going to ride this first mover pony as long as they can, and miners' fear of uncertainty is running us right down the long road of irrelevance. All made possible with the centralized control of information, while using overblown fear of centralization as a tool of control... just like in good 'ol meatspace... it's pretty disgusting.

The Gmax treehouse is a great mental picture.

Fuck this. We won't have Bitcoin if people like him can't comprehend what is going on.
Gavin is the most experienced Bitcoin dev who has shown to create solutions instead of problems. Code written by Jeff is running on thousands of computers worldwide for years.
And still the current "core developers" are somehow the masterminds and bitcoin "wizards"?
Has he been living under a rock for the last months??

Multiply that by a couple million. Scale it up, if you will.

Linux is behind the scenes every Android phone, of which there are billions around. Each one containing quite some code from Jeff Garzik.

great series of charts showing the shift in economic policy over the last two years.
[...]

What anyone should be able to clearly see from this graph is that Bitcoin with an uncapped blocksize worked just fine, with the time preference of transaction confirmation building a nice distribution of coin fees. And on average well below 1MB for almost all history.

And... last but not least, on this:

https://www.reddit.com/r/Bitcoin/comments/4biob5/research_into_instantaneous_vote_behavior_in/

Has anyone ever requested his raw data? I do not post in North Korea...
Last time I used PRAW (the python reddit API wrapper), reddit had a 2/s API limit - so how did he acquire it with that resolution and number of users?

I want to see the raw data to believe anything here. Faking an analysis is just as likely as vote manipulation nowadays.

 

[sickpig]

@adamstgbit

You get it backward, it's not the validity of the block that determine the validity of txs it contains. It actually works in the exact opposite way. This is how the bitcoin protocol works.

Currently, and even if/when SegWit will be active, to validate a block a full node need the the block in the first place.

A block is composed as follows (https://en.bitcoin.it/wiki/Protocol_documentation#block):

A block to be valid has to have a valid header and all the txs it contains have to be valid. To validate a transaction you need tx signature. To a header is valid only if, among other things, from the group of txs it contains you're able to reproduce the merkle tree.

Hence it is not possible to fully validate a block without having at disposal all the txs (data + sign) included in the block.

If, after the validation step, a node decides to drop the signatures it can't provide blocks to any another node's initial sync (or IBD).

Taken it to the extreme, if all bitcoin nodes (even the one used by the miners) decide to prune witness data after validation the bitcoin p2p network won't grow any more (a part from new SPV clients), because new nodes won't be able to the blockchain from any nodes.

 

[awemany]

@Richy_T

(2:28:30)

Has it been edited out? I do not see Q&A about massive blocksize increase in that video?

 

[Peter R]

@awemany:

Nice to have you drop in and say hi!

You certainly did your part, given your dedicated debating of the big-block side on Reddit over the past year or so. Hopefully you'll have time to continue to stop by once and a while--I always enjoy reading your comments.

 

[AdrianX]

@awemany that's at awesome post, I've enjoyed your contributions to reddit and here too, I hope you hang around and drop in form time to time, your opinions and insights are obviously valued.