Pages:
Author

Topic: GHash.IO has 51% of the network now (Read 4289 times)

sr. member
Activity: 448
Merit: 250
June 14, 2014, 01:58:26 PM
#48
means most if it is true, who holds a very large cex.io.
probably because many buyers GHS are there so that they invest more advanced
hero member
Activity: 770
Merit: 500
June 14, 2014, 01:36:24 PM
#47
Yes, they are having issues.
legendary
Activity: 1610
Merit: 1000
Well hello there!
June 14, 2014, 01:29:46 PM
#46
Anybody else having still having issue(s) accessing ghash.io site?

*Pool hashrate ~8.58PH the last time I checked.
hero member
Activity: 519
Merit: 502
June 14, 2014, 01:27:39 PM
#45
hey guys is cex.io opening or working. i am connected fine to the pool I just checked but i cant get to the website. Cex.io or ghash.io its been a few hours.
legendary
Activity: 4690
Merit: 1276
June 14, 2014, 01:02:05 PM
#44

There are working Commons in Germany, not sure what you want to say with the emphasis on Mongolia here.

I call attention to Mongolia because graze lands were the most widely thought of example of a 'commons' back in the day and even carried that label specifically.  This specific example continues to be observed in that country.

Again, you are failing to be very specific about 'the commons' that you are analyzing.  'In Germany' is still a fail.


Let me quote from the book I was also point at earlier. I found it now.

Quote
Hardin's [the one that published the paper "The Tragedy of the Commons" in 1968] argument was a hit because (1) it features an A+B=C simplicity that appears to be inarguably correct; and (2) it is useful in justifying seemingly heartless decisions by entrenched powers. [...] Hardin's articulation of the folly of communal ownership has provided cover repeatedly to those arguing for the privatization of government services and the conquest of native lands.
One other thing Hardin's elegant argument has in common with that of Mathus [does not matter here, thus left out earlier]: it colapses on contact with reality.
As Camdian author Ian Angus explains, "Hardin simply ignored what actually happens in a real commons: self regulation by the communites involved"
[...] any indivdual who tries to game the system is quickly found out and punished.

PhD C. Ryan and MD C. Jethá - Sex at Dawn - 2010

Thus my argument earlier that social punishment of GHash.io and cex.io is the way to go and centralisation/authority or privatization (not even sure how this would work with bitcoin). But lets get to the problem at hand:

Mining bitcoins is the pasture here. Everyone can herd their sheep (start mining) there. There are limited resources (Bitcoins) for everyone to have. If one has to many sheep (over 50% hashpower) in the pasture they risk ruining it (bitcoin) for everyone. A big scale double spending attack would have a big impact on the bitcoin price and thus hurt the attacker. Perfect example of a common that works because if you start gaming the system the angry mob will fuck you up.

You are framing the problem as it relates to Bitcoin incorrectly (IMO).  'bitcoins' are not the resource and don't play a part in things.  The bitcoin network is the operative construct.  It is supposed (by most people) to be available for common use by anyone who feels so inclined.

I would argue with you and Angus that as soon as control measures come into existence (e.g., getting beat up by an angry mob, or taking defensive measures such as running off your competition's sheep to make more resources available for your own (rather than moving a mile to virgin grass)) then we are no longer talking about pure commons.  All you are saying is that 'the commons' are not workable except in unusual circumstances and people adapt.  Big deal.  Pretty much all life forms face this reality at some point and they all adapt else they would not be here.

Hardin was quite unpopular within many social science communities for his various views and prescriptions.  Similar to many of those who've studied other other sensitive topics such as human intelligence.  One could achieve instant popularity which was not in alignment with the quality of one's work by 'dis'n' the guy.  I think this happened from time to time in his case.

I read Hardin's tragedy work as a kid and several times since.  What I took away was mostly the observation that people have a propensity to maximize their personal gain and it impacts corner-cases such as pure commons and exploitation of resources within them.  I personally never took from it that 'privatization' was his preferred default way of dealing with the problem.  I'm a strong proponent of pubic ownership of bulk resources.  I live on a thin strip of private lands surrounded by a state forest, and I am very much in favor of public ownership of resources where it is feasible, though I am opposed to that being the exclusive mode.  The state forest (which I am seeing out my window as I type these words) is distinctly not 'commons' and exploitation of it is fairly carefully controlled.

copper member
Activity: 1498
Merit: 1528
No I dont escrow anymore.
June 14, 2014, 06:41:43 AM
#43
...
Now I understand that these assumptions are wrong. Commons worked for several hundreds of years without any problems. Commons still exists and work fine.
...

That is a complete nonsense statement.  The whole point of 'the tragedy of the commons' is that indeed 'the commons' work find as an economic model (and in other ways) when there are abundant resources.  Saying that 'Commons still exist and work fine' makes zero sense unless you are specific about what instance you are pointing to.

Well I was pointing at the pastures where to term originated. I might have been more clear about that.

The principle applied to Bitcoin mining casts 'the commons' as the Bitcoin system generally with the underlying assumption (arguably incorrect) that centralization is bad.  Abundance would be when there is no threat of centralization.  In that instance maximizing personal gain can be done without threatening the system because it doesn't matter what pool he associates with.  In the more classic scenario it would be the same thing as the sheep herder growing his flock to the maximum number he can handle without worrying about ruining the graze on the commons since there is so much of it.

Yes, it is true that 'the Commons still exist and still work fine' in Mongolia.  They've ceased to exist and work fine in many other places where they once thrived.

There are working Commons in Germany, not sure what you want to say with the emphasis on Mongolia here.

Let me quote from the book I was also point at earlier. I found it now.

Quote
Hardin's [the one that published the paper "The Tragedy of the Commons" in 1968] argument was a hit because (1) it features an A+B=C simplicity that appears to be inarguably correct; and (2) it is useful in justifying seemingly heartless decisions by entrenched powers. [...] Hardin's articulation of the folly of communal ownership has provided cover repeatedly to those arguing for the privatization of government services and the conquest of native lands.
One other thing Hardin's elegant argument has in common with that of Mathus [does not matter here, thus left out earlier]: it colapses on contact with reality.
As Camdian author Ian Angus explains, "Hardin simply ignored what actually happens in a real commons: self regulation by the communites involved"
[...] any indivdual who tries to game the system is quickly found out and punished.

PhD C. Ryan and MD C. Jethá - Sex at Dawn - 2010

Thus my argument earlier that social punishment of GHash.io and cex.io is the way to go and centralisation/authority or privatization (not even sure how this would work with bitcoin). But lets get to the problem at hand:

Mining bitcoins is the pasture here. Everyone can herd their sheep (start mining) there. There are limited resources (Bitcoins) for everyone to have. If one has to many sheep (over 50% hashpower) in the pasture they risk ruining it (bitcoin) for everyone. A big scale double spending attack would have a big impact on the bitcoin price and thus hurt the attacker. Perfect example of a common that works because if you start gaming the system the angry mob will fuck you up.
legendary
Activity: 4690
Merit: 1276
June 14, 2014, 03:54:36 AM
#42
...
Now I understand that these assumptions are wrong. Commons worked for several hundreds of years without any problems. Commons still exists and work fine.
...

That is a complete nonsense statement.  The whole point of 'the tragedy of the commons' is that indeed 'the commons' work find as an economic model (and in other ways) when there are abundant resources.  Saying that 'Commons still exist and work fine' makes zero sense unless you are specific about what instance you are pointing to.

The principle applied to Bitcoin mining casts 'the commons' as the Bitcoin system generally with the underlying assumption (arguably incorrect) that centralization is bad.  Abundance would be when there is no threat of centralization.  In that instance maximizing personal gain can be done without threatening the system because it doesn't matter what pool he associates with.  In the more classic scenario it would be the same thing as the sheep herder growing his flock to the maximum number he can handle without worrying about ruining the graze on the commons since there is so much of it.

Yes, it is true that 'the Commons still exist and still work fine' in Mongolia.  They've ceased to exist and work fine in many other places where they once thrived.

---

It would be interesting to note when the first time Satoshi seemed to be aware of the concept of mining pools.  I suspect that it may have been one of the things that he (or they) really didn't anticipate very clearly in formulating the various constructs which were to influence the ecosystem.  One of the impressively few things I might add.

legendary
Activity: 1260
Merit: 1029
June 14, 2014, 02:53:57 AM
#41
Bitcoin protocol needs to be changed so noone never can get that much power again. It can be done. You can limit amount of blocks one entity can get in a row, you can change many, many things to prevent people from getting all into one basket.
Question is...why devs didint already change it?

Because you can't. It's anonymous. You can not tell who actually mined a block. It would be easy for one pool to make itself look like five pools with 10% each but still being the one 50% pool.

Of course it CAN be done. They can, if they wanted, force only p2p pools (which ar enothing like ghash and rest of em). They can do all sorts of changes to prevent massing of hash under control of 1 entity and they can do much muhc more. There are examples in other alt coins but hey...if it's easyer for you to think this way, who am i to break your dreams....
copper member
Activity: 1498
Merit: 1528
No I dont escrow anymore.
June 14, 2014, 02:51:33 AM
#40
Its a classic common resource problem: http://en.wikipedia.org/wiki/Tragedy_of_the_commons
"... individuals, acting independently and rationally according to each one's self-interest, behave contrary to the whole group's long-term best interests by depleting some common resource."

Historically, only 2 viable solutions exist:
1) An authority to regulate (% in this case)
2) Privatization

http://en.wikipedia.org/wiki/Tragedy_of_the_commons#Modern_solutions

Pick your poison..... but be quick about it.

When I first heard about the tragedy of the commons (one of my profs was talking about) I though the idea behind it was correct. But then I read more about it by accident in another book which is more about human nature than about economy. (sex before dawn maybe - not sure)

Now I understand that these assumptions are wrong. Commons worked for several hundreds of years without any problems. Commons still exists and work fine.

The basic asumption is correct if one person in the group behaves against the intrests of the group everyone suffers. But you have to keep in mind what times you are talking about and what situations. They all knew eachother. They knew which individual of the group was the black sheep. Thus the group had social means to punish which did not require an authority figure (e.g. a lord or cestallan) or privatization (wasnt a thing in medieval times anyway).
Same today, if you know a miner or two you can talk to them about Ghash.io and that you are concerned about the hashingpower the pool assembled. If you dont know a miner you can talk about it here (or reddit or twitter or facebook or...) and raise awairness.
Humans are creatures of cooperation not of selfishness. We are social animals. We need eachother. Thats the leverage here.
sr. member
Activity: 336
Merit: 260
June 14, 2014, 02:48:24 AM
#39
If more PH of power are added each 10 days than removed (hash rate graph) as the rising difficulty shows too, then it'll be back to square one sooner rather than later.
sr. member
Activity: 350
Merit: 250
Decentralized thinking
June 14, 2014, 02:28:52 AM
#38
According to Coindesk article...Bitfury to be moving 1.5PH of it's hashing power away from ghash.io with possibility of moving more in the future depending on where they are at later.

reference: http://www.coindesk.com/bitfury-pulls-power-ghash-community-uproar/

You beat me to it Smiley Lets hope more miners follow suit
copper member
Activity: 2898
Merit: 1465
Clueless!
June 14, 2014, 02:23:42 AM
#37
nothing is "solved". one pool having even 20% is a huge problem, and double spend attacks are not impossible under such conditions. this is a matter of probability. the problem is not "51%". the problem is centralization. and it is NOT solved.

heh don't know anything but too bad the bitcoin protocol could not show 'favortism' to go smaller rather then larger (decentralized rather then centralized) obvioualy
if that was possible/easy the devs of bitcoin would have done so already

Bitcoin always a new zombie behind every door it seems

Searing
hero member
Activity: 658
Merit: 500
June 14, 2014, 02:21:19 AM
#36
nothing is "solved". one pool having even 20% is a huge problem, and double spend attacks are not impossible under such conditions. this is a matter of probability. the problem is not "51%". the problem is centralization. and it is NOT solved.

i am amazed at how much blind trust bitcoiners can put into a hashing pool... why exactly are we here again?
copper member
Activity: 2898
Merit: 1465
Clueless!
June 14, 2014, 02:17:10 AM
#35
This sure didn't last long.  The link posted in the OP now shows Ghash down to 41%.  Thank you to all the miners who switched to a different pool.

yea! don't need the drama (even if to no ill effect) of bouncing around at 50 to 51% reminding everyone (and the press) of the dreaded 51% (hopefully)
boogie man!

it just goes to show the wisdom in such situation of Monty Python....(hey it always works if you run fast enough)

http://www.youtube.com/watch?v=7FPELc1wEvk

Run Away Run Away!

Searing

hero member
Activity: 672
Merit: 500
June 14, 2014, 12:15:52 AM
#34
It is a miracle how this situation got solved with no central authority? Did disciplined miners switch pools or ghash changed IP and masked some of their hash rate as unknown? I am sceptical on this being "solved"  Embarrassed
hero member
Activity: 490
Merit: 500
June 13, 2014, 11:55:54 PM
#33
This sure didn't last long.  The link posted in the OP now shows Ghash down to 41%.  Thank you to all the miners who switched to a different pool.
legendary
Activity: 2198
Merit: 1014
Franko is Freedom
June 13, 2014, 08:33:55 PM
#32
So, ghash.io acts as a benevolent dictator -- one who knows the penalty for abusing that privilege is instant decapitation.

They should be decapitated anyways...
hero member
Activity: 686
Merit: 500
June 13, 2014, 07:49:47 PM
#31
ghash gimme BTC bonus boss  Wink
sr. member
Activity: 313
Merit: 250
i ♥ coinichiwa
June 13, 2014, 07:46:32 PM
#30
Bitcoin protocol needs to be changed so noone never can get that much power again. It can be done. You can limit amount of blocks one entity can get in a row, you can change many, many things to prevent people from getting all into one basket.
Question is...why devs didint already change it?

Because you can't. It's anonymous. You can not tell who actually mined a block. It would be easy for one pool to make itself look like five pools with 10% each but still being the one 50% pool.
legendary
Activity: 1232
Merit: 1002
June 13, 2014, 05:44:17 PM
#29
Thanks for sharing so Gigahash did reach 51%
Well there goes the hashrate
(Went back to 49%) But that is still more or less a 50% network rate right there

Not really mate!
He explained it here clearly!

sssh! you know alot of people here don't like to go on facts!
but yeah, nicely said mate.

You DO know that chart isn't a representation of the total hashrate available, it's just the percentage of blocks found.

While it's CLOSE to 50%, it is NOT the actual hashrate.

https://blockchain.info/charts/hash-rate

This is the actual hashrate, currently at 93,507,419 GH/S, or 93.5 Petahash.

CEX.IO is reporting 40.13...

Lemme do the math for ya.

43.1%

Up there? Hell yeah... But is the sky falling? No. Not yet, but very close.

I've had enough FUD for one day. Please don't confuse your charts. While it's easy to scream "It's All CEX's Fault" and hit panic mode, it's consumers, not miners, on Ghash right now.

Consumers consume, Miners protect their resources.
Pages:
Jump to: