Bitcoin Forum
May 09, 2024, 01:38:26 PM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 [14] 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 »
  Print  
Author Topic: Bitcoin halving to be canceled?  (Read 33687 times)
USB-S
Sr. Member
****
Offline Offline

Activity: 574
Merit: 250

In XEM we trust


View Profile
November 25, 2015, 06:44:12 PM
 #261

if it was done (the halving) every year, right now miners would be fucked basically

So how does it change the things now, before the next halving? If the halving weren't as rare as it is, would the miners have already been fucked by now? If they would (which seems to be your point), I see no reason why they won't be ruined by this halving...

What goes around comes about

it change the thing because they had time to improve the efficiency, now they have a better efficiency and can sustain the halving even with a <<<$100 price, if the halving was every years they would still be with the efficiency of 2011....

A better efficiency is fully accommodated over time by the rising difficulty, so we are effectively time-invariant in the long run (efficiency vs difficulty), i.e. the halving interval is irrelevant if we still hit the efficiency wall before the halving. In other words, your assumption would hold true only if (and for as long as) the efficiency outpaces the difficulty. Previously, this condition had been satisfied by the new more efficient equipment hitting the market on a pretty regular basis...

But now we seem to be well past that point
What are you talking about?
New chips hit the market just this summer. Give them some time if you already want new chips. However current chips are pretty effective at around
0.25 J/GH


If we were to cancel the halving, can we remove the 21mil hard cap as well?


````````````````████████
_`````````██████████████████████
_`````█████████████████████████████
_```█████████████████████████████████
_``████████████████████████████████████
_█████████```````████████```````████████
_███████````████````██`````███````███████
_██████````████████`````████████``███████
_██████````██████````██``██████```███████
_███████```````````████``````````████████
_██████████████████████████████████████
_``████████████████████████████████████
_```_████████████████████████████████
_``````████████████████████████████
_`````````3█████████████████████
play.infinity
        Eжeднeвный ДЖEКПOT
TELEGRAM CHAT   SITE   TELEGRAM
                   Get free eth
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
1715261906
Hero Member
*
Offline Offline

Posts: 1715261906

View Profile Personal Message (Offline)

Ignore
1715261906
Reply with quote  #2

1715261906
Report to moderator
Amph
Legendary
*
Offline Offline

Activity: 3206
Merit: 1069



View Profile
November 25, 2015, 06:58:10 PM
 #262

if it was done (the halving) every year, right now miners would be fucked basically

So how does it change the things now, before the next halving? If the halving weren't as rare as it is, would the miners have already been fucked by now? If they would (which seems to be your point), I see no reason why they won't be ruined by this halving...

What goes around comes about

it change the thing because they had time to improve the efficiency, now they have a better efficiency and can sustain the halving even with a <<<$100 price, if the halving was every years they would still be with the efficiency of 2011....

A better efficiency is fully accommodated over time by the rising difficulty, so we are effectively time-invariant in the long run (efficiency vs difficulty), i.e. the halving interval is irrelevant if we still hit the efficiency wall before the halving. In other words, your assumption would hold true only if (and for as long as) the efficiency outpaces the difficulty. Previously, this condition had been satisfied by the new more efficient equipment hitting the market on a pretty regular basis...

But now we seem to be well past that point

difficulty will not increase if the price do not increase, so no one can not rely on difficulty and dismiss the impact of the halving, for the efficiency

satoshi predicted that the price would be increased slowly not in few years, this was another reason for the halving
deisik (OP)
Legendary
*
Offline Offline

Activity: 3444
Merit: 1280


English ⬄ Russian Translation Services


View Profile WWW
November 25, 2015, 07:00:47 PM
 #263

A better efficiency is fully accommodated over time by the rising difficulty, so we are effectively time-invariant in the long run (efficiency vs difficulty), i.e. the halving interval is irrelevant if we still hit the efficiency wall before the halving. In other words, your assumption would hold true only if (and for as long as) the efficiency outpaces the difficulty. Previously, this condition had been satisfied by the new more efficient equipment hitting the market on a pretty regular basis...

But now we seem to be well past that point
What are you talking about?
New chips hit the market just this summer. Give them some time if you already want new chips. However current chips are pretty effective at around
0.25 J/GH

Will these new chips be cost effective at 12.5 BTC reward per block (given the price and difficulty we have today)?

deisik (OP)
Legendary
*
Offline Offline

Activity: 3444
Merit: 1280


English ⬄ Russian Translation Services


View Profile WWW
November 25, 2015, 07:03:14 PM
 #264

difficulty will not increase if the price do not increase, so no one can not rely on difficulty and dismiss the impact of the halving, for the efficiency

This essentially means that the current level of efficiency should be doubled at halving, to keep things as they were/would be before the halving...

Are the new chips twice as efficient?

Amph
Legendary
*
Offline Offline

Activity: 3206
Merit: 1069



View Profile
November 25, 2015, 07:08:46 PM
 #265

aside from this i believe that the primarily reason for the halving is to delay as long as possible the fee era, and to permit to have a very valuable fees for the miners too mine, otherwise if we were to mine all the supply very quickly and enter the fee era prematurely

miners would not have profit anymore and they would leave the game, leaving the network unsecure
deisik (OP)
Legendary
*
Offline Offline

Activity: 3444
Merit: 1280


English ⬄ Russian Translation Services


View Profile WWW
November 25, 2015, 07:09:55 PM
 #266

If we were to cancel the halving, can we remove the 21mil hard cap as well?

The latter follows from the former

n691309
Legendary
*
Offline Offline

Activity: 1526
Merit: 1001



View Profile
November 25, 2015, 07:20:52 PM
 #267

If we were to cancel the halving, can we remove the 21mil hard cap as well?

The latter follows from the former

If the supply is unlimited then how much would cost a bitcoin? it is better to have restriction and a limit on supply as the price should be increased year by year until there are a small amount in market.
lissandra
Sr. Member
****
Offline Offline

Activity: 434
Merit: 250


View Profile
November 25, 2015, 07:22:16 PM
 #268

If we were to cancel the halving, can we remove the 21mil hard cap as well?

The latter follows from the former

i thought the hard 21 mill cap is to save bitcoin in its value.

Like how we aint a printing press, like the feds do no?

Amph
Legendary
*
Offline Offline

Activity: 3206
Merit: 1069



View Profile
November 25, 2015, 07:27:46 PM
 #269

i was thinking that the halving would have been useless if bitcoin had risen enormously in value in the past and within the first 3-5 years

but even satoshi didn't believe in it to skyrocket so fast

Are the new chips twice as efficient?

yes they are, s7 is x2 more efficient than the s5 antminer
GannickusX
Hero Member
*****
Offline Offline

Activity: 490
Merit: 500


View Profile
November 25, 2015, 08:26:01 PM
 #270

if it was done (the halving) every year, right now miners would be fucked basically

So how does it change the things now, before the next halving? If the halving weren't as rare as it is, would the miners have already been fucked by now? If they would (which seems to be your point), I see no reason why they won't be ruined by this halving...

What goes around comes about

it change the thing because they had time to improve the efficiency, now they have a better efficiency and can sustain the halving even with a <<<$100 price, if the halving was every years they would still be with the efficiency of 2011....

A better efficiency is fully accommodated over time by the rising difficulty, so we are effectively time-invariant in the long run (efficiency vs difficulty), i.e. the halving interval is irrelevant if we still hit the efficiency wall before the halving. In other words, your assumption would hold true only if (and for as long as) the efficiency outpaces the difficulty. Previously, this condition had been satisfied by the new more efficient equipment hitting the market on a pretty regular basis...

But now we seem to be well past that point
What are you talking about?
New chips hit the market just this summer. Give them some time if you already want new chips. However current chips are pretty effective at around
0.25 J/GH


If we were to cancel the halving, can we remove the 21mil hard cap as well?

How would we cancel the halving exactly? I thought that was something built in the system itself and couldn't be stopped, pretty much in all coins. Last one that got halved was litecoin and the price wasn't really affected at all by it. I Hardly doubt there is a way to remove the 21mil cap.
Amph
Legendary
*
Offline Offline

Activity: 3206
Merit: 1069



View Profile
November 25, 2015, 08:35:44 PM
 #271

if it was done (the halving) every year, right now miners would be fucked basically

So how does it change the things now, before the next halving? If the halving weren't as rare as it is, would the miners have already been fucked by now? If they would (which seems to be your point), I see no reason why they won't be ruined by this halving...

What goes around comes about

it change the thing because they had time to improve the efficiency, now they have a better efficiency and can sustain the halving even with a <<<$100 price, if the halving was every years they would still be with the efficiency of 2011....

A better efficiency is fully accommodated over time by the rising difficulty, so we are effectively time-invariant in the long run (efficiency vs difficulty), i.e. the halving interval is irrelevant if we still hit the efficiency wall before the halving. In other words, your assumption would hold true only if (and for as long as) the efficiency outpaces the difficulty. Previously, this condition had been satisfied by the new more efficient equipment hitting the market on a pretty regular basis...

But now we seem to be well past that point
What are you talking about?
New chips hit the market just this summer. Give them some time if you already want new chips. However current chips are pretty effective at around
0.25 J/GH


If we were to cancel the halving, can we remove the 21mil hard cap as well?

How would we cancel the halving exactly? I thought that was something built in the system itself and couldn't be stopped, pretty much in all coins. Last one that got halved was litecoin and the price wasn't really affected at all by it. I Hardly doubt there is a way to remove the 21mil cap.

miners consensus above a certain % with the help of other people can potentially hard fork bitcoin and cancel the halving

it's all about consensus + hard fork, after all many alt coin are hard fork of bitcoin...
deisik (OP)
Legendary
*
Offline Offline

Activity: 3444
Merit: 1280


English ⬄ Russian Translation Services


View Profile WWW
November 25, 2015, 09:06:07 PM
Last edit: November 25, 2015, 09:37:21 PM by deisik
 #272

i was thinking that the halving would have been useless if bitcoin had risen enormously in value in the past and within the first 3-5 years

but even satoshi didn't believe in it to skyrocket so fast

Are the new chips twice as efficient?

yes they are, s7 is x2 more efficient than the s5 antminer

The quick search on Google reveals that "the S5+ antminer offers nearly 3,000 more GH/s than the announced S7". How's that? Besides, the S5 uses 60 chips per miner while the S7 uses 162, so is this new miner in fact more efficient? Efficiency is not only about power consumption (operating costs), you should also take into account capital expenditures. In this way, if the new miner consumes half but costs twice as much (or even the same if it has the same hashrate), then you would still be less profitable at 12.5 BTC with it than with the old miner at 25 BTC per block per same price...

That's why Bitcoin halving may have devastating effects on its future (too many factors need to be "upgraded")

deisik (OP)
Legendary
*
Offline Offline

Activity: 3444
Merit: 1280


English ⬄ Russian Translation Services


View Profile WWW
November 25, 2015, 09:12:56 PM
 #273

If we were to cancel the halving, can we remove the 21mil hard cap as well?

The latter follows from the former

If the supply is unlimited then how much would cost a bitcoin? it is better to have restriction and a limit on supply as the price should be increased year by year until there are a small amount in market.

I have created another topic which discusses this issue in particular. Short version, the higher price (due to limited supply) turns Bitcoin into a Ponzi scheme, which would kill it eventually...

anyzhugo789
Newbie
*
Offline Offline

Activity: 1
Merit: 0


View Profile WWW
November 26, 2015, 02:11:10 AM
 #274

What?

Who has the most power over Bitcoin? Right, these are mining pools. Who is most interested in preserving the current block reward? The same mining pools...



Who gives a fuck what the pools do, it's what the miners do.

Mining pools act on behalf of the miners. Besides, do you understand that the miners will be the first to welcome the cancellation of halving?
btckold24
Sr. Member
****
Offline Offline

Activity: 434
Merit: 250


★Bitvest.io★ Play Plinko or Invest!


View Profile
November 26, 2015, 06:18:24 AM
 #275

isnt it built into the system to halve? also who is this "they" that would be able to make a decision like that



YOU CAN ▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀
PLAY or INVEST


▄████████████████████▄
██████████████████████
██████████▀▀██████████
█████████░░░░█████████
██████████▄▄██████████
███████▀▀████▀▀███████
██████░░░░██░░░░██████
███████▄▄████▄▄███████
████▀▀████▀▀████▀▀████
███░░░░██░░░░██░░░░███
████▄▄████▄▄████▄▄████
██████████████████████
▀████████████████████▀
▄████████████████████▄
██████████████████████
█████▀▀█▀▀▀▀▀▀██▀▀████
█████░░░░░░░░░░░░░▄███
█████░░░░░░░░░░░░▄████
█████░░▄███▄░░░░██████
█████▄▄███▀░░░░▄██████
█████████░░░░░░███████
████████░░░░░░░███████
███████░░░░░░░░███████
███████▄▄▄▄▄▄▄▄███████
██████████████████████
▀████████████████████▀
▄████████████████████▄
███████████████▀▀▀▀▀▀▀
███████████▀▀▄▄█░░░░░█
█████████▀░░█████░░░░█
███████▀░░░░░████▀░░░▀
██████░░░░░░░░▀▄▄█████
█████░▄░░░░░▄██████▀▀█
████░████▄░███████░░░░
███░█████░█████████░░█
███░░░▀█░██████████░░█
███░░░░░░████▀▀██▀░░░░
███░░░░░░███░░░░░░░░░░
▀██░▄▄▄▄░████▄▄██▄░░░░
▄████████████▀▀▀▀▀▀▀██▄
█████████████░█▀▀▀█░███
██████████▀▀░█▀░░░▀█░▀▀
███████▀░▄▄█░█░░░░░█░█▄
████▀░▄▄████░▀█░░░█▀░██
███░▄████▀▀░▄░▀█░█▀░▄░▀
█▀░███▀▀▀░░███░▀█▀░███░
▀░███▀░░░░░████▄░▄████░
░███▀░░░░░░░█████████░░
░███░░░░░░░░░███████░░░
███▀░██░░░░░░▀░▄▄▄░▀░░░
███░██████▄▄░▄█████▄░▄▄
▀██░████████░███████░█▀




Rainbot
Happy Hour
Faucet
Amph
Legendary
*
Offline Offline

Activity: 3206
Merit: 1069



View Profile
November 26, 2015, 07:38:16 AM
 #276

i was thinking that the halving would have been useless if bitcoin had risen enormously in value in the past and within the first 3-5 years

but even satoshi didn't believe in it to skyrocket so fast

Are the new chips twice as efficient?

yes they are, s7 is x2 more efficient than the s5 antminer

The quick search on Google reveals that "the S5+ antminer offers nearly 3,000 more GH/s than the announced S7". How's that? Besides, the S5 uses 60 chips per miner while the S7 uses 162, so is this new miner in fact more efficient? Efficiency is not only about power consumption (operating costs), you should also take into account capital expenditures. In this way, if the new miner consumes half but costs twice as much (or even the same if it has the same hashrate), then you would still be less profitable at 12.5 BTC with it than with the old miner at 25 BTC per block per same price...

That's why Bitcoin halving may have devastating effects on its future (too many factors need to be "upgraded")

i was talking about the old s5, the s5+ was released with the s7 basically, but it's still not efficient liek the s7, it's a bulky version of the s5 with a slightly more better efficiency
deisik (OP)
Legendary
*
Offline Offline

Activity: 3444
Merit: 1280


English ⬄ Russian Translation Services


View Profile WWW
November 26, 2015, 08:10:21 PM
 #277

The quick search on Google reveals that "the S5+ antminer offers nearly 3,000 more GH/s than the announced S7". How's that? Besides, the S5 uses 60 chips per miner while the S7 uses 162, so is this new miner in fact more efficient? Efficiency is not only about power consumption (operating costs), you should also take into account capital expenditures. In this way, if the new miner consumes half but costs twice as much (or even the same if it has the same hashrate), then you would still be less profitable at 12.5 BTC with it than with the old miner at 25 BTC per block per same price...

That's why Bitcoin halving may have devastating effects on its future (too many factors need to be "upgraded")

i was talking about the old s5, the s5+ was released with the s7 basically, but it's still not efficient liek the s7, it's a bulky version of the s5 with a slightly more better efficiency

It seems that your definition of efficiency is heavily lopsided. My understanding of efficiency is entirely economic, i.e. I don't care about power consumption and hashrates as such. All I care is the balance of costs and revenues. In other words, the miner that incurs less cost and brings in more revenue is more efficient (i.e. does more with less) than the one that does the opposite...

That's what all technical parameters ultimately boil down to

Amph
Legendary
*
Offline Offline

Activity: 3206
Merit: 1069



View Profile
November 27, 2015, 07:53:52 AM
 #278

The quick search on Google reveals that "the S5+ antminer offers nearly 3,000 more GH/s than the announced S7". How's that? Besides, the S5 uses 60 chips per miner while the S7 uses 162, so is this new miner in fact more efficient? Efficiency is not only about power consumption (operating costs), you should also take into account capital expenditures. In this way, if the new miner consumes half but costs twice as much (or even the same if it has the same hashrate), then you would still be less profitable at 12.5 BTC with it than with the old miner at 25 BTC per block per same price...

That's why Bitcoin halving may have devastating effects on its future (too many factors need to be "upgraded")

i was talking about the old s5, the s5+ was released with the s7 basically, but it's still not efficient liek the s7, it's a bulky version of the s5 with a slightly more better efficiency

It seems that your definition of efficiency is heavily lopsided. My understanding of efficiency is entirely economic, i.e. I don't care about power consumption and hashrates as such. All I care is the balance of costs and revenues. In other words, the miner that incurs less cost and brings in more revenue is more efficient (i.e. does more with less) than the one that does the opposite...

That's what all technical parameters ultimately boil down to

my definition is different, talking about the initial investment as efficiency is a no-sense, the efficiency is only related to consumption, and the s7 consume half of the s5, so x2 efficiency

the fact that it cost more, does not mean anything, you can always sell it for the same value even after many months, guaranteed
deisik (OP)
Legendary
*
Offline Offline

Activity: 3444
Merit: 1280


English ⬄ Russian Translation Services


View Profile WWW
November 27, 2015, 08:27:57 AM
Last edit: November 27, 2015, 08:42:59 AM by deisik
 #279

It seems that your definition of efficiency is heavily lopsided. My understanding of efficiency is entirely economic, i.e. I don't care about power consumption and hashrates as such. All I care is the balance of costs and revenues. In other words, the miner that incurs less cost and brings in more revenue is more efficient (i.e. does more with less) than the one that does the opposite...

That's what all technical parameters ultimately boil down to

my definition is different, talking about the initial investment as efficiency is a no-sense, the efficiency is only related to consumption, and the s7 consume half of the s5, so x2 efficiency

The problem is it doesn't in the least guarantee that with an s7 you will profit twice as much as with an s5 given only its half as much power consumption. Initial investment (aka capital expenditure) as a part of fixed costs defines your break-even point. You still have to shell out on the equipment first, whether you love it or not. Your personal opinion (lol) that you can sell your miner is not related to this, and I highly doubt that you will get for it the same price which you bought it at after a few months of operation (and more so before the halving)...

Even if you personally will be lucky to sell your device at its original price tag

deisik (OP)
Legendary
*
Offline Offline

Activity: 3444
Merit: 1280


English ⬄ Russian Translation Services


View Profile WWW
November 27, 2015, 08:47:05 AM
Last edit: November 27, 2015, 09:21:03 AM by deisik
 #280

the fact that it cost more, does not mean anything, you can always sell it for the same value even after many months, guaranteed

If the equipment still brings in cash, it makes no sense economically to sell it (prior to break-even at least). Otherwise, you would be better off by not buying it in the first place (all other things being equal). But if it stops bringing in enough revenue at some point (negative cash flow), you won't be able to profitably sell it...

If only out of pure luck to some goof, lol

Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 [14] 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 »
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!