Bitcoin Forum
May 27, 2024, 02:38:05 AM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
Poll
Question: Is it time to shift the sacred cow?
Yes. Shift the sucker! - 7 (38.9%)
No. It was touched by the Holy Hand of Satoshi - 2 (11.1%)
FFS! Wake me up when this is over. - 9 (50%)
Total Voters: 18

Pages: [1]
  Print  
Author Topic: Time to SHIFT the Sacred Cow!  (Read 794 times)
solex (OP)
Legendary
*
Offline Offline

Activity: 1078
Merit: 1002


100 satoshis -> ISO code


View Profile
May 28, 2015, 03:14:43 AM
Last edit: May 29, 2015, 09:28:30 PM by solex
 #1



Q; What is a conservative course of action?
A: Conservative action (for a successful enterprise) is always to maintain the status quo as much as possible in the face of necessary change. Because Bitcoin has prospered for 6 years the status quo is to maintain the network as it is now.
Unfortunately, neither keeping the 1MB or increasing it can maintain the network how it is now!
This is a dilemma, but all enterprises face dilemmas at some time: necessity for change from competition, customers, technology or regulation. So, a conservative position is to adapt to an external change while maintaining the status quo as much as possible.

Q; Why isn’t “doing nothing” the best approach?
A: Because “doing nothing” has no real fall-back plan.
The fall-back plan for an increase to, say 20MB, is a later soft-fork down to 2 or 3MB. This may or may not become necessary, but at least it can be achieved smoothly. Think: duck paddling hard in a millpond.
If the 1MB limit is kept and causes mayhem to confirmation times, bad publicity etc, then the fall-back “plan” is a painful and panicked hard-fork. Think: duck going through the water-wheel.

Q: When is a hard-fork not a hard-fork?
A: When it is planned so long ahead that 100% of all software instances are upgraded before it takes effect.
In practice, this can’t be achieved but targeting 90+% is desirable and sufficient for a smooth transition.

Q: When is a hard-fork most painful?
A: When it is a rapid decision, forced by circumstances, and everyone has to upgrade in a very short time.
The more delay before acting causes a bigger downside when a fork becomes forced.

Q: What about keeping the 1MB to put upward pressure on fees?
A: That is “Blockchain Economics”. Applying to the blockchain a false economic model of reality, i.e. an economic model which has a hard limit such as the zero-bound in Central Bank interest rate targeting which attempts to set the price of money in a fiat system, but fails in a deflationary economy.
When the rubber hits the road of this reality the result will be a rapid plateau in total fees and then an inexorable decline as users abandon the use of Bitcoin in favour of alternatives, and new cryptocurrency users keen on the future of money are forced to use alternatives from the get-go.

Q: OK, but I’m a determined fan of Blockchain Economics. When is it safe to try it?
A: It’s never safe because of the Trade-Off, but the best time is trying a soft-limit while the hard-limit is higher (safety valve), and when the network is small but viable, and no one is paying attention, such as the situation in 2012, Unfortunately, the time for this has gone because the Bitcoin market capitalisation is in the billions of dollars and the world’s press is closely watching.

Q: What is the Trade-Off?
A: A fundamental of Bitcoin is its confirmation cycle which averages 10 minutes.
When blocks have extra space the transaction counts per block are highly variable, but any given unconfirmed  transaction may expect confirmation in 10 minutes.  
When blocks are full transaction counts per block are stable, but the expected confirmation time for a unconfirmed transaction becomes highly variable.

Q: Is the benefit of very stable block sizes worth the cost of highly variable confirmation times?
A: No!

Q; When is a flooding or spamming attack on Bitcoin most effective and cheapest to execute?
A: When blocks are nearly always full. It is least effective and most expensive when blocks normally have lots of unused space.

Q: Can Bitcoin be a reserve / settlement currency with high-value users, primarily a store of value like an electronic gold?
A: Yes, but not when it can only handle 0.01% (a ten-thousandth) of the world’s transactions. Not when another cryptocurrency is handling even 1%, let alone 10%, 50% or 99.99% of the world’s transactions. Bitcoin can only become a true reserve currency and electronic gold when it scales. Otherwise this remains a dream, a mirage.

Q: What about bandwidth considerations?
A: Satoshi put the 1MB in place when there were no lightweight nodes and all users had to run a full node. That time is nearly five years ago. Global improvements in bandwidth mean that the overhead he allowed for, by selecting 1MB in 2010, is more like 4MB in 2015.

Q: What about technical concerns such as growth of the UTXO set?
A: The smartest way to meet a challenge is to leverage it as an opportunity.
Example : Aligning the demand for blocks larger than 1MB with an incentive to maintain a cleaner UTXO set.
The complexity of this change needs to be weighed against the simplicity of a fixed limit (e.g. 20MB). Sometimes simplicity outweighs a leveraged gain.
Also, the UTXO set could be kept cleaner, as a stand-alone change: allowing free transaction space based upon reducing utxo (negative delta) instead of being based upon the number of days destroyed, which was to encourage old coins being spent, something less important.

Q. What are the implications for decentralisation, particularly: full node counts?
A. Full node counts have been in decline for several years, mostly because of the growth in SPV wallets and lightweight nodes, and mining pools (for minimizing variance), also the increase in blockchain size for users who won’t spend a few $100 on TB disks.
Many node owners are long-term investors and believers in Bitcoin as a major currency and payment system of the future. That is why they hang in there. They are waiting to see it scale and will run non-mining, non-commercial nodes to do it. If the status quo is destabilized (e.g. unstable confirmation times) then full node owners will switch off faster.
Full node counts would be improved by greater adoption and the future use of node payment services where non-mining nodes get micro-payments via off-chain channels for servicing the network.
Keeping the 1MB, because it is the least conservative action, will accelerate the decline in full node counts.

Q: What if someone makes a load of gigablocks?
A: They can’t. If the 1MB disappeared, the message size limit of about 32MB means that this is the maximum block size. Gigablocks are not possible until all nodes support set reconciliation ("bandwidth reduction scheme") such as in IBLT, and block segmentation software.

Q: Why not wait for off-chain solutions which can handle 100x the on-chain volume, like Lightning Networks?
A: There is no time for that, LN is still in the documentation and technical specification stages, also the LN opinion is that 1MB is too small for them in the event that a lot of payment channels need to be closed quickly. 32MB or even 20MB would be sufficient for a long time.

molecular
Donator
Legendary
*
Offline Offline

Activity: 2772
Merit: 1019



View Profile
May 28, 2015, 06:16:29 AM
 #2

well-written Q&A, thank you.

I take minor issue with calling IBLT block transmission scheme "compression". It gives the false impression that storage requirements would somehow be decreased ("oh, we can ZIP the blockchain?"), which is not the case. Maybe call it a "bandwidth reduction scheme"? Not elegant, but I fail to find a good expression.

Let's move that cow!

PGP key molecular F9B70769 fingerprint 9CDD C0D3 20F8 279F 6BE0  3F39 FC49 2362 F9B7 0769
anderson00673
Sr. Member
****
Offline Offline

Activity: 308
Merit: 250



View Profile
May 28, 2015, 06:45:38 AM
 #3

Thank you sir there is some excellent information in there.  Did you get that somewhere or make it yourself?  Either way great set of q/a.  I feel like this should be stickied.  Anyone else?
pandher
Legendary
*
Offline Offline

Activity: 952
Merit: 1000


Stagnation is Death


View Profile WWW
May 28, 2015, 08:10:20 AM
 #4

If Bitcoin fails to scale, its days are numbered
solex (OP)
Legendary
*
Offline Offline

Activity: 1078
Merit: 1002


100 satoshis -> ISO code


View Profile
May 28, 2015, 08:48:26 AM
Last edit: May 29, 2015, 12:34:51 AM by solex
 #5

well-written Q&A, thank you.

I take minor issue with calling IBLT block transmission scheme "compression". It gives the false impression that storage requirements would somehow be decreased ("oh, we can ZIP the blockchain?"), which is not the case. Maybe call it a "bandwidth reduction scheme"? Not elegant, but I fail to find a good expression.

Let's move that cow!


Thanks!
For a long time I tried to avoid referring to IBLT as "compression", trying "bandwidth reduction scheme".

Thank you sir there is some excellent information in there.  Did you get that somewhere or make it yourself?  Either way great set of q/a.  I feel like this should be stickied.  Anyone else?

Thanks also. It's my own summary, reworking the concepts behind the pros and cons which have been debated for a few years.

Edit: added off-chain Q&A item.

Sheldor333
Sr. Member
****
Offline Offline

Activity: 406
Merit: 250


View Profile
May 28, 2015, 08:49:21 AM
 #6

You wouldn't download a cow.  Grin

Kyraishi
Hero Member
*****
Offline Offline

Activity: 952
Merit: 513



View Profile
May 28, 2015, 08:58:32 AM
 #7

You wouldn't download a cow.  Grin

That depends.....
How about two cows, or from two cows?
To be more specific, from tucows.com ?  Grin

@OP excellent Q&A

Thanks  Cheesy

solex (OP)
Legendary
*
Offline Offline

Activity: 1078
Merit: 1002


100 satoshis -> ISO code


View Profile
May 29, 2015, 11:13:52 PM
Last edit: May 29, 2015, 11:32:33 PM by solex
 #8

Gavin is 100% right and showing leadership to break the analysis paralysis.

sourceforge
Quote
What do other people think?


If we can't come to an agreement soon, then I'll ask for help
reviewing/submitting patches to Mike's Bitcoin-Xt project that implement a
big increase now that grows over time so we may never have to go through
all this rancor and debate again.

I'll then ask for help lobbying the merchant services and exchanges and
hosted wallet companies and other bitcoind-using-infrastructure companies
(and anybody who agrees with me that we need bigger blocks sooner rather
than later) to run Bitcoin-Xt instead of Bitcoin Core, and state that they
are running it. We'll be able to see uptake on the network by monitoring
client versions.

Perhaps by the time that happens there will be consensus bigger blocks are
needed sooner rather than later; if so, great! The early deployment will
just serve as early testing, and all of the software already deployed will
ready for bigger blocks.

But if there is still no consensus among developers but the "bigger blocks
now" movement is successful, I'll ask for help getting big miners to do the
same, and use the soft-fork block version voting mechanism to (hopefully)
get a majority and then a super-majority willing to produce bigger blocks.
The purpose of that process is to prove to any doubters that they'd better
start supporting bigger blocks or they'll be left behind, and to give them
a chance to upgrade before that happens.


Because if we can't come to consensus here, the ultimate authority for
determining consensus is what code the majority of merchants and exchanges
and miners are running.


--
--
Gavin Andresen

If BitcoinXT needs to blaze the trail then two further considerations are:

1) Make the 20MB expire after 5 years so that the limit becomes the default of 32MB.

2) After the desired mining majority for the fork is achieved, allow a short grace period, e.g. 10 days, before >1MB blocks are allowed, which gives time for laggard full nodes to upgrade.

Q7
Sr. Member
****
Offline Offline

Activity: 448
Merit: 250


View Profile WWW
May 30, 2015, 09:28:24 AM
 #9

Well, there is always a dilemma, if we don't do it now then when? As we move further down the road, the risk associated would further increase. But then if we don't do it, the question is, are we fully prepared for the eventual increase in number of transaction, years down the road? We are talking about bitcoin reaching mainstream adoption right, and isn't that what every bitcoiners who are holding on to the coins envisioned for?

NorrisK
Legendary
*
Offline Offline

Activity: 1946
Merit: 1007



View Profile
May 30, 2015, 09:31:34 AM
 #10

Can't they make it a sliding scale? Say increase the limit with every difficulty adjustment, based on the size of the blocks the month before?
coinits
Legendary
*
Offline Offline

Activity: 1582
Merit: 1019


011110000110110101110010


View Profile
May 31, 2015, 12:43:43 AM
 #11

I am a nobody in the grand scheme of things but will offer up an opinion anyway. If you mess with the blocksize of bitcoin people are going to balk and this coin will take a substantial hit. It has been around too long and too well established to start doing tweaks to it. Adapt to the perceived shortcomings or create a separate (forked) coin.

Jump you fuckers! | The thing about smart motherfuckers is they sound like crazy motherfuckers to dumb motherfuckers. | My sig space for rent for 0.01 btc per week.
molecular
Donator
Legendary
*
Offline Offline

Activity: 2772
Merit: 1019



View Profile
May 31, 2015, 05:56:05 AM
 #12

I am a nobody in the grand scheme of things but will offer up an opinion anyway. If you mess with the blocksize of bitcoin people are going to balk and this coin will take a substantial hit. It has been around too long and too well established to start doing tweaks to it. Adapt to the perceived shortcomings or create a separate (forked) coin.

I was at a wedding and talking to an old acquiantance there. So what's new with bitcoin, he asked. He was genuinely interested, the ever-falling price didn't really scare him and he was thinking about buying in and becoming a user.

I said there's currently a big debate about increasing the maximum amount of transactions we can process from 7 to 140.

His reaction: "what do you mean, you can only process 7 transactions per second?!? That's clearly nowhere near enough. Bitcoin is going to fail if you don't solve that problem."

I had to agree. 7 is not enough. I don't think they guy will become a bitcoiner before the cow has been moved.

To be clear: I'm all for researching and implementing a real solution to the problem. But it's going to take a couple of years to do that. Let alone roll it out in a meaningful way.

I was expecting the next ten-bagger (10-fold bitcoin ramp) sometime 2016. I think I will have to revise that expectation. If we're going to go with a 1 MB limit it's not going to happen the way I imaged. Public discussion will be about the scalability of Bitcoin and how slow and expensive it is to send money. There'll be no "to the mooon!"-posts on reddit, there'll be "fuck, my money is stuck again!"-posts. It'll drive people away and it puts an effective cap on growth. "Oh, but we've got the lightning network in the pipeline, in about 3 years you'll be able to send 1 million tx per second, so don't worry", is not going to cut it... at all.

We need to buy time.

PGP key molecular F9B70769 fingerprint 9CDD C0D3 20F8 279F 6BE0  3F39 FC49 2362 F9B7 0769
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!