Bitcoin Forum
April 23, 2024, 02:21:44 PM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: [1]
  Print  
Author Topic: Incentivizing Decentralization: Tick Tock, turn Bitcoin into a clock?  (Read 557 times)
Cielo1 (OP)
Newbie
*
Offline Offline

Activity: 4
Merit: 0


View Profile
April 06, 2016, 02:01:16 PM
 #1

https://www.reddit.com/r/Bitcoin/comments/4dkkku/incentivizing_decentralization_tick_tock_maybe_we/



Just got back from a BitDevs Meetup. The topic was Decentralization.
 
http://www.meetup.com/BitDevsNYC/events/229692528/
 
At one point the discussion turned toward the perverse incentives for miners to centralize and/or withhold information and how this may get worse as technology gets better and the speed of light stays the same.
 
Soooooo … I started playing around in my head on the way back home …
 
Tick-Tock Proposal:
 
*Tick Block: Traditional Block
*Tock Block: Includes the traditional Block plus a Tock nonce and Reward transaction
 
*Tick Block: 1/16th difficulty and 60% of the Reward
*Tock Block: 100% difficulty and 40% of the Reward
 
If a miner were lucky enough to successfully mine a Tick Block, they would be incentivized to propagate that block immediately and as widely as possible. There would be no reason to hold back as the miner would risk losing out on the majority Reward to another miner.
 
The difference in difficulty would lead to quickly generated Tick Blocks which then have the whole length of the Tock Block mine to propagate.
 
Thus the network would have a higher tolerance for larger blocks and network topography gradients.
 
If a miner were to receive a Tick Block they could choose to:
*Mine the Tock: to get the Tock Reward
*Idle: If the Tock difficulty isn’t worth the reward
*Mine their own Tick: A miner would not be incentivized to do this as they would put themselves behind the pack
 
The successfully mined Tock Block would be small and propagate quickly as the hefty Tick Block on which it was based would already be propagated.
 
The split would not have to be 60/40. It could be more or less but the goal would be to give the Tick Miner an incentive to rapidly and properly propagate data through the network rather than withhold it for the purpose of generating an advantage.
 
It might make sense for the network auto rebalance the amount of reward based on the hash power dedicated to the Ticks relative to the Tocks.
 
It might make sense to allow the Tick miner to formally post an additional bounty for the Tock miner to collect.
 
We could also consider having auto adjusting (or equal) difficulties for the Tick vs. Tock. Regardless of how it is achieved The Tick should somehow be relatively cheaper in order to reward miners for propagating data.
 
I’m sure I glossed over some details but I think the gist is here. Am I missing something?




=============================

Q:
The commercialised and highly effecient miner still has the highest chance of winning? How does that incentivize decentralization?

A:
Yes, they still have the highest chance of winning. But additionally, if they mine a block, they then have a head start in mining the next block in the time that it takes for the original block to propagate.
This could be because of network topography (great firewall) or possibly intentionally.
The idea is to incentivize the miner to publish information as fast as possible to lock down the big easy reward.
Secondarily, smaller miners would benefit in two ways:
*The big reward is an order of magnitude easier to mine
*They would be more likely to be mining on a valid block
The network tries to produce one block per 10 minutes. It does this by automatically adjusting how difficult it is to produce blocks.
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
1713882104
Hero Member
*
Offline Offline

Posts: 1713882104

View Profile Personal Message (Offline)

Ignore
1713882104
Reply with quote  #2

1713882104
Report to moderator
1713882104
Hero Member
*
Offline Offline

Posts: 1713882104

View Profile Personal Message (Offline)

Ignore
1713882104
Reply with quote  #2

1713882104
Report to moderator
1713882104
Hero Member
*
Offline Offline

Posts: 1713882104

View Profile Personal Message (Offline)

Ignore
1713882104
Reply with quote  #2

1713882104
Report to moderator
CIYAM
Legendary
*
Offline Offline

Activity: 1890
Merit: 1075


Ian Knowles - CIYAM Lead Developer


View Profile WWW
April 06, 2016, 02:04:22 PM
 #2

I don't see how this is really any different to the weak blocks proposal (which actually has several such "tick" blocks per "tock").

Perhaps you are not aware of that?

With CIYAM anyone can create 100% generated C++ web applications in literally minutes.

GPG Public Key | 1ciyam3htJit1feGa26p2wQ4aw6KFTejU
Cielo1 (OP)
Newbie
*
Offline Offline

Activity: 4
Merit: 0


View Profile
April 06, 2016, 02:30:57 PM
 #3

Agreed.

I was not aware of that.

Thanks!

I think there might a slight difference in that I'm not sure a weak block approach incentivizes propagation?

Also weak blocks idea was more to target to empty blocks rather than propagation?
CIYAM
Legendary
*
Offline Offline

Activity: 1890
Merit: 1075


Ian Knowles - CIYAM Lead Developer


View Profile WWW
April 06, 2016, 02:34:54 PM
 #4

I think once the Lightning Network is live this will actually perhaps not really be so important anyway (I would think that shrinking the bandwidth being wasted by getting rid of repeated transaction content from relaying txs and blocks would be the next priority).

Perhaps the main thing that weak blocks helps with is zero confirmation transactions (as you can effectively have 1/4 confirmations rather than 0) but if most txs end up being LN ones then I'm not sure that it is going to really matter so much.

With CIYAM anyone can create 100% generated C++ web applications in literally minutes.

GPG Public Key | 1ciyam3htJit1feGa26p2wQ4aw6KFTejU
Cielo1 (OP)
Newbie
*
Offline Offline

Activity: 4
Merit: 0


View Profile
April 06, 2016, 02:51:42 PM
 #5

There's two aspects we are looking at ... scalability and propagation.

Kind of related. Kind of distinct.

LN addresses scalability only no?

Incentivizing propagation feels like may still become important.

But, I agree that this is far from clear.
CIYAM
Legendary
*
Offline Offline

Activity: 1890
Merit: 1075


Ian Knowles - CIYAM Lead Developer


View Profile WWW
April 06, 2016, 02:58:43 PM
 #6

LN addresses scalability only no?

LN addresses the scalability of transactions (from the current 3-7 TPS to potentially millions per second) although the massive scaling that it can bring will likely require increasing block sizes as well (if not wanting payment channels to have to last a long time which is a why it is expected that the block size will be increased in 2017).

For the LN propagation is separate from the Bitcoin network itself and so would be expected to actually be much quicker (as generally most people are going to do most of their transactions within a small geographical sub-network and confirmations are not relevant until the channel is closed).

With CIYAM anyone can create 100% generated C++ web applications in literally minutes.

GPG Public Key | 1ciyam3htJit1feGa26p2wQ4aw6KFTejU
Cielo1 (OP)
Newbie
*
Offline Offline

Activity: 4
Merit: 0


View Profile
April 06, 2016, 03:04:09 PM
 #7

Fully agree on the Sub-Networks being localized and that it is probably ok.

Just worried that the Prime-Network might also become localized and that it is probably not ok.
CIYAM
Legendary
*
Offline Offline

Activity: 1890
Merit: 1075


Ian Knowles - CIYAM Lead Developer


View Profile WWW
April 06, 2016, 03:06:19 PM
 #8

Just worried that the Prime-Network might also become localized and that it is probably not ok.

I'm not sure why you'd think that - LN is using Bitcoin txs which are just not broadcast to the normal Bitcoin network until the channel is closed (so all settlement is still being done using the normal Bitcoin system).

The localised effect that speeds up your LN txs (prior to settlement) is simply because the nodes you are transacting with would be expected to be nearby (such as your local food delivery service for example).

If you open up a channel with your Pizza Delivery Service and the channel is given a duration of say 1 month then your txs for buying pizzas will be almost instant and at the end of the month the settlement will occur as a normal Bitcoin tx.

With CIYAM anyone can create 100% generated C++ web applications in literally minutes.

GPG Public Key | 1ciyam3htJit1feGa26p2wQ4aw6KFTejU
franky1
Legendary
*
Online Online

Activity: 4200
Merit: 4435



View Profile
April 06, 2016, 03:57:29 PM
Last edit: April 06, 2016, 04:10:54 PM by franky1
 #9

Just worried that the Prime-Network might also become localized and that it is probably not ok.

I'm not sure why you'd think that - LN is using Bitcoin txs which are just not broadcast to the normal Bitcoin network until the channel is closed (so all settlement is still being done using the normal Bitcoin system).

The localised effect that speeds up your LN txs (prior to settlement) is simply because the nodes you are transacting with would be expected to be nearby (such as your local food delivery service for example).

If you open up a channel with your Pizza Delivery Service and the channel is given a duration of say 1 month then your txs for buying pizzas will be almost instant and at the end of the month the settlement will occur as a normal Bitcoin tx.


 the large channel closing tx still gets relayed across the network. so its not exactly a cure for bandwith usage. but more so a efficient usage of the relay network. by communicating offchain on a different port, to then merging some of the transactions before throwing a large tx onchain.

LN will ultimately be useful for gambling, faucets and other things(users doing many tx an hour/day).. but if someone is only making one transaction a day/week/month.. then LN is not helpful because the tx to lock in, tx to release is more hassle and data compared to just sending onchain.

what we will see is that channels will stream line 5000-20000 tx's into a hopefully single 4200 input/outputs large tx.. instead of 5000-20000 different and separate transactions where upto 75% may be the same people receiving and sending the same funds back and forth..

the issue becomes when txs themselves become as big as the blocks because the channel is lumping it all together, rather than people sending transactions adhok at slightly different times.

eg. instead of
1 onchain tx every 0.14seconds (the 7tx/s theory of 4200 capacity)
its 1 onchain tx (with the 4200 input outputs) every 10 minutes

(note my numbers are not science but random for demonstration purposes only)

I DO NOT TRADE OR ACT AS ESCROW ON THIS FORUM EVER.
Please do your own research & respect what is written here as both opinion & information gleaned from experience. many people replying with insults but no on-topic content substance, automatically are 'facepalmed' and yawned at
CIYAM
Legendary
*
Offline Offline

Activity: 1890
Merit: 1075


Ian Knowles - CIYAM Lead Developer


View Profile WWW
April 06, 2016, 04:06:35 PM
 #10

Unfortunately this topic is now being derailed by a poster of nonsense (who doesn't even understand how these things work).

@OP - I would advise you to ignore his posts and lock the topic (if you don't know how then scroll down to the bottom and look to the left).

(he is a well known shill/troll who constantly posts inaccurate information to try and attack the Bitcoin Core devs)

With CIYAM anyone can create 100% generated C++ web applications in literally minutes.

GPG Public Key | 1ciyam3htJit1feGa26p2wQ4aw6KFTejU
franky1
Legendary
*
Online Online

Activity: 4200
Merit: 4435



View Profile
April 06, 2016, 04:14:44 PM
Last edit: April 07, 2016, 10:07:17 AM by franky1
 #11

Unfortunately this topic is now being derailed by a poster of nonsense (who doesn't even understand how these things work).

@OP - I would advise you to ignore his posts and lock the topic (if you don't know how then scroll down to the bottom and look to the left).

(he is a well known shill/troll who constantly posts inaccurate information to try and attack the Bitcoin Core devs)


care to prove me wrong or just use insults without proof..

1. closing the channel, does it involve lots of individual transactions.. spread out evenly every few milliseconds over the 10 minutes.. or 1 LARGE tx with all the offchain tx's lumped together as a multi- input/output single tx?(answer: 1 LARGE tx with all the offchain tx's lumped together as a multi- input/output single tx)

2. is the closing channel tx going to be just 500bytes OR closer to 1mb if there are 4000 input/outputs to settle?(answer:closer to 1mb if there are 4000 input/outputs to settle)

3. is the closing channel transaction only pushed directly to 20 pool and not the whole relay network(6000 nodes) or will the whole relay network(6000 nodes) see the LARGE closing channel tx before it gets added to a block..?(answer:  the whole relay network(6000 nodes) see the LARGE closing channel tx before it gets added to a block)

i numbered them so be brutally honest. its 3 easy answers
(in short. bandwidth due to larger single transactions is still going to cause bandwidth issues. which weakblocks and other such things wont solve)

one last question

4. prove LN actually works right now within bitcoins ledger. show me a single transaction that is an LN channel closing transaction on bitcoins blockchain, so that you can prove that you have seen 'how things work', or realise your theories are not yet a reality because LN does not work, yet.


I DO NOT TRADE OR ACT AS ESCROW ON THIS FORUM EVER.
Please do your own research & respect what is written here as both opinion & information gleaned from experience. many people replying with insults but no on-topic content substance, automatically are 'facepalmed' and yawned at
CIYAM
Legendary
*
Offline Offline

Activity: 1890
Merit: 1075


Ian Knowles - CIYAM Lead Developer


View Profile WWW
April 06, 2016, 04:33:04 PM
 #12

@OP - as I warned you this is the crap that you are going to see.

I do not bother with the above troll as he has been told again (and again and again) how things work and refuses to accept it.

(it is clear that he is being paid to make such posts as he does so every day on this forum in many different topics)

With CIYAM anyone can create 100% generated C++ web applications in literally minutes.

GPG Public Key | 1ciyam3htJit1feGa26p2wQ4aw6KFTejU
franky1
Legendary
*
Online Online

Activity: 4200
Merit: 4435



View Profile
April 07, 2016, 04:23:55 AM
 #13

i hear only insults and nothing more.

if only he stopped calling people trolls and insulting people. and instead stuck to technicals like a few years ago..
i just checked his post history and grabbed 6 examples (there are soo many more though)

https://bitcointalk.org/index.php?topic=1402980.msg14342928#msg14342928
https://bitcointalk.org/index.php?topic=1402980.msg14343042#msg14343042
https://bitcointalk.org/index.php?topic=1402980.msg14343105#msg14343105
https://bitcointalk.org/index.php?topic=1402980.msg14343262#msg14343262
https://bitcointalk.org/index.php?topic=1330553.msg14273368#msg14273368
https://bitcointalk.org/index.php?topic=1330553.msg14272256#msg14272256

it seems someone else is also coming to the conclusion that CIYAM is just an insulting person
https://bitcointalk.org/index.php?topic=1400837.msg14223573#msg14223573

the real funny thing is that he knows that blockstreams features are not made for bandwidth reduction

but if you ever question blockstreams motives or question the chance that there are bugs in the code (nothing is perfect, after all)
he will reign down fire on you. without showing technicals or reasons why. but pure insults and flame wars.

its best to research everything and make your own opinion. instead of blindly following one person who has a very negative ego purely trying to gain some reputation to make his poor little company worth some VC investment.

yep thats right he is shilling himself out to the altcoiners hoping one day for a VC payday.

so dont be afraid of insulting people like ciyam. instead keep asking the questions you have and get as much information from as many sources as possible. dont be fooled into being spoonfed one source of info

I DO NOT TRADE OR ACT AS ESCROW ON THIS FORUM EVER.
Please do your own research & respect what is written here as both opinion & information gleaned from experience. many people replying with insults but no on-topic content substance, automatically are 'facepalmed' and yawned at
CIYAM
Legendary
*
Offline Offline

Activity: 1890
Merit: 1075


Ian Knowles - CIYAM Lead Developer


View Profile WWW
April 07, 2016, 09:20:44 AM
 #14

https://bitcointalk.org/index.php?topic=1396525.msg14444220;topicseen#new

(just one example of @franky1 making a fool of himself through his lack of knowledge - if you check other topics he posts in you'll see the same sorts of basic incorrect understanding about Bitcoin being repeated)

I'd rather be known as rude and correct rather than polite and completely wrong.

With CIYAM anyone can create 100% generated C++ web applications in literally minutes.

GPG Public Key | 1ciyam3htJit1feGa26p2wQ4aw6KFTejU
Pages: [1]
  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!