Topic: Please do not change MAX_BLOCK_SIZE (Read 13023 times)

If it is a crap then why do you use it?
IMHO he stopped developing it when he realized that the network was already strong enough, to resist any possible corruption within the development team that he left the code to.


As for the rest of your post, your proposals of how to handle the problem are just stupid - no more comments.

Or you could just stop cramming bloat inside the damn blocks, and keep a separate useful chain of transactions, and the useless chain of blocks out of the TX chain.

You only need a diff 1 for TX chain. It is only the blocks which need the true diff as proof of work. How many useless and repeat transactions are included in a block, but simply ignored because an earlier duplicate exists in a previous block?

This whole system is setup bass-ackwards. Only miners should be concerned with the super-bloated chain. If the TXs are confirmed, then they are confirmed. They only have to be linked to a block, not crammed inside of it. Oh yea, they do for this crappy moronic setup. lol. The only setup where something like 0 is 2K, and 500000 is 2K, and we get "charged" for sending K, without being told what K is, or what the K-value is, prior to sending, or why a tx we are sending is 100K or only 2K. That is what happens when you let programmers try to use the real world... They fail.

Screw-it... Make max-block-size a full GB... it won't matter what it is, because you either download 10000 small ones, or 1 large one, it is still the same freaking volume of crappy data-packing and useless bloat.

Do you guys even know what g-zip is? Or, I don't know... byte-reduction by not sending bloated "human-readable" expanded values, where real values are just as functional.

Stop sending "0100101010010101001" and "sud3U" when you only need to send "FU"

Overkill where it isn't needed, and not enough refinement where it is, where it is already possible. Because you keep building off one mans poor crappy code. He stopped developing this crap because it was crap.

I don't know why you came out with 500%, but well.. let's go into it.

For the last 24h, the average block size has been ~88.6KB - this gives you a potential for 1000% growth, yet without touching MAX_BLOCK_SIZE.

In fact the potential is much, much bigger, because when you eliminate free transactions, by forcing the fee market to work, it will make all the small payments (0.01 BTC or less) unprofitable. And now such a tiny transactions seem to be taking a huge amount of the blocks. And who is paying for this? We are all paying for this - and if you increase the bock size, we will be paying for it even more.

Vote for option 'E' and hope someone is smart enough to figure out what needs to be done for dynamic, deterministic block management.

Still, I believe that the period between 0 conf and 1 will hamper adoption in the physical world for immediate payments.  It would be nice for Bitcoin to be as pedestrian as technically possible.


Sergio had an idea worth debating.  More of an indirect approach.
https://bitcointalksearch.org/topic/m.1561612

Explain.

IMO we need at least 5MB MAX_BLOCK_SIZE to handle a 500% increase in Bitcoin usage.

+1

Please do not raise the max block size. It hurts projects I am working on.

Paypal is a payment processor, not a currency. Bitcoin will beat Paypal if it beats the fiat currencies that Paypal operates on.

Payment processors appeared on the market because at some point in time there was a need to send fiat money quickly and cheaply. Sending bills and coins in envelopes was obviously not practical and the market developed a solution.

Let the same need to appear in Bitcoin, and the solutions will also develop themselves, allowing bitcoins to be sent cheaper and quicker, while still keeping it lightweight and decentralized.

And let me repeat it one more time: increasing MAX_BLOCK_SIZE is not a solution for anything!
Though, it has a good potential to create a disaster.

But it is an outstanding opportunity for off-chain solutions.  A lot of Bitcoin holders already use potentially off-chain solution such as an on-line wallet for spending money.  I have and do even while I run one or more full clients.  A solution where I could use such a service to make micro-transactions would be a very interesting to me, and I am much more likely to actually perform such transactions if someone offered a well engineered system which made it safe and easy to do.

In fact, since I first studied Bitcoin in mid 2011 I've never personally felt comfortable using it for piddly little shit because I was bothered by the scaling issues since day one.  OTOH, the big draw for me has always been de-centralization, and it is clearly not a priority for a lot of people in the community.  In fact, it does not really seem to be a central concern for the Bitcoin Foundation to my shock and horror.

Luckybit, this is the same argument made by Shiresilver
https://bitcointalksearch.org/topic/m.1602604
Please see the counter-arguments which follow in the thread.

tl;dr: Microtransactions cost the network more than they are worth.

Yes they are and I'll explain why. The problem with the web is content isn't monetized. All content must be monetized. All websites must be paid for somehow. If we use a website like suicidegirls for example, currently it does not accept Bitcoins. They expect us to pay a monthly fee of $5 which ends up on our bank statement to be scrutinized. They have to charge a fee because they have to pay their models somehow, so they charge a membership fee.

Micropayments change everything. First we'd have privacy, nothing would have to be on our bank statement. Websites would charge transaction fees and micro fees to fund their site. A nash equilibrium would be reached between end users and site owners allowing the site to make a profit while allowing end users access to content.

Microtransactions would allow end users to pay for their online activities but also allow end users to get paid for their online activities. These microtransactions are the major advantage that cryptocurrencies have over fiat currencies. The other major advantage is privacy. Why not leverage both these advantages to produce a fully monetized web experience which benefits end users by allowing end users to effectively pay for content with their revenue from mining or whatever they do to earn coins and allow content producers to get paid for producing content?

If you don't have a way for content producers to get paid then you dont have a way for masses of people to earn coins. Anyone can make a blog, get paid via micropayments, and earn coins just like exchanges can all charge fees and pay for their operation that way. Anyone can start a site or forum and earn coins, now you have people actually building stuff, working for coins, spending coins, all which are good for a currency but which will only be possible with microtransactions. These small transactions which are a fraction of a cent are enough to monetize the entire web.

It does not have to be equal to fiat money values either. Assume for a moment that every screen name, every account you make, every post you make, every feature you use, all is priced in some satoshi. What if you make money for every post you make but you pay for every feature you use? This would create an economy even on forums such as this. A few bucks worth of satoshi would unlock portions of the site reserved for people who pay for VIP access. A few bucks worth could be earned just by posting quality content on the site. If every url click could have a price attached to it, the entire web can be monetized easily with each click costing a fraction of a cent but over the course of a day those cents add up. Not only this but it would mean exclusive content for people who have the coins and people who don't have coins would have to buy them or not access certain portions of certain sites at all.

What a coincidence! I can't remember the last time I paid someone with plutonium yet that is valued at $10,990 per gram.
http://www.nbl.doe.gov/htm/lists/plutonium_certified_reference_materials_price_list.htm


No, they are not, unless you mean small transactions equivalent to fiat paper currency values..

If Bitcoin is to be a currency and not just hoarded then it must support microtransactions. That is the only way it can beat Paypal. Everything on the web can be monetized once microtransactions are implemented properly.

I believe everything which isn't monetized should be monetized because nothing is actually free. Thinking we can access websites for free and get free content forever is an illusion. Micropayments solve the problem because Bitcoins are valuable. Everything has some value which can be priced in Bitcoins and should be and that even includes stuff like our screen names and accounts on sites but it can also be our ability to access any link on a site. If you have to pay x Bitcoins to access certain links on the site this pays for the site and you'll certainly have enough satoshi to comfortably pay for it.

I'll answer some of your points, though I snipped them for brevity.

The Megauploads thing (among many others) demonstrate that state and corporate interests are already significantly merged.  For all intents and purposes it is fair to say that the NSA and various enforcement agencies will do the bidding of the large financial corporations if and when it is deemed to be in the interest of said corporations.

Bitcoin could survive, and possibly thrive, by being absorbed into the large financial and internet companies and in that way avoid attack.  I dis-favor that approach.

Bitcoin could also survive if it remains a tiny dot, and/or is useful to a sliver of those with political clout.  I distrust that even if this were the desired trajectory it would be within the power of the solution to modulate these factors.  If it is useful, and it is, demand could grow rapidly and in waves.

If Bitcoin attempts to remain de-centralized by forming low data-rate distributed core and a multitude of decoupled second tier solutions, I believe it would have a chance of surviving in something akin to it's present form even under significant attack.  So, to address your concern about my preferred strategy needing to scale, it is not the case.  You choose the data rate which gives a light enough footprint to achieve your distribution goals, then let second tier providers build on top of it.  With economic incentive I have high confidence in the market to produce robust solutions at a rapid rate.

I cannot envision a solution which you seem to be eluding to involving high-cap players and legal structures being what I would consider robust or desirable.  I strongly prefer that Bitcoin remain completely decoupled from any legal structures.  I have nothing against cooperating with state-level legal frameworks and I think that individual operators should usually do so, but I strongly feel that it should be optional and the core solution should avoid being hamstrung by such constructs.

In the case of Megaupload it was not a threat to the US government, it was a threat to certain US businesses that corrupted the US government to align it's national interest with their business interests. We voters did not vote to take down Megauploads to protect the business model of their competitor. You're right they were able to kill it and you're right Bitcoin will probably be attacked in some countries at some point.

But Bitcoin is not completely decentralized, nothing we know of can scale and be completely decentralized. The math isn't going to work. How can it work for Bitcoin? If we look at how Bitcoin is now it's not completely decentralized. We have mining pools and companies.

Just use Bitcoin to purchase VPN service. What do you mean by fire up a VPN to make a micro transaction? What does it matter if we did have to run a VPN? Microtransactions are a thing that cryptocurrencies can do better than any other.

Bitcoin does not threaten all central governments. The NSA is not threatened at all by it.
I do see your point that you cannot count on jurisdiction and that is a gamble but isn't that the gamble we take right now already? It seems we are already taking that sort of gamble and a 51% attack has not occurred.


The solution you see in your vision just does not seem possible to do and have it scale up to the levels Bitcoin has to scale up without changing Max Block Size. So unless Bitcoin is going to basically just settle for being unable to scale up it's going to have to make a concession. If it does not then Litecoin or some other coin probably will and will surpass Bitcoin because of it. Microtransactions are essential to the success of Bitcoin and cryptocurrencies in general.

How exactly would you do this and allow Bitcoin to both scale up and do microtransactions?

I have some possible solutions to prevent highly capitalized entities from operating Bitcoin natively but they all would require complex legal schemes to pull off such as cooperatives, syndicates, L3C's and benefit corps. It is possible in theory but in practice there is the concern that if done wrong it would become just another highly capitalized entity.

Breaking away from POTS modems which are, unfortunately, probably not sufficient for anything but a solution designed to target them which Bitcoin never was...

Anyone who has much experience with system design at the network level will also understand that in order to have a smooth functioning system, one needs significantly in excess of a simplistic number resulting from a division operation and two gross numbers.

I've already described seasonal and diurnal traffic patterns which are certain to impact a currency system significantly. 

More generally, if one is running almost any system near capacity, it is an arduous and frusterating task to catch up after inevitable failures of one sort of another.  I've seen poorly spec'd systems and/or systems which have grown unexpectedly which are *never* able to catch up.  If one is lucky (or skilled) the system can perform in a degraded capacity while catching up, but this is not likely to be the case for Bitcoin due to it's design.

Mix in a determined attacker with significant resource who can deploy, at least, denial of service attacks (and likely much more), and we have a solution which is very subject to outage.  My opinion for a long time has been to deploy some 'marketing' resources to making it broadly understood to the end-user public that even in the event of a long duration outage, Bitcoin is not 'dead' and the value held is still as safe as an individual's ability to retain their secret keys as sole property.

Anyone that is still using a modem probably won't have a digital connection, so they won't get 56kbps, it's more like 3.5-4KB/s, 35kbps or so.

Not that it really matters, since even at 56kbps it wouldn't come close to keeping up.  At the "real" optimal speed you'd get, about 4KB/s, it'd take over 2 hours.

Well, I forgot that the block chain can be compressed... so if you were downloading with zmodem or something instead of the bitcoin client, it'd probably be a bit faster

+1

Retrospecting the thread, I'd like to bring in some numbers.
At the moment we have 6353442 unspent outputs with 11287864 BTC in them, from which 4799281 (75%) outputs are below 0.01 BTC, and they carry 1515 BTC (0.01%) of the money.
The average block size, for the last 24 h, has been 125.0 KB.

So it makes me wonder, again; why do we actually need to increase the 1MB block limit, within the next year or so?
What numbers do he have to support even worrying about it?

As for my measurement, if we do the dust filtering right, and allow the fee system to work out some fair tx fees, we wont need to touch MAX_BLOCK_SIZE at least for the next 3 years. And if there will be a war (which I hope not) the 1MB limit might easily survive for 10 more years...

Keep bitcoin node lightweight - please.

The Internet is anything but decentralized and it is no trouble whatsoever for the authorities to shut down centralized services.  Nor is it required that there is a low threshold for collateral damage.  Witness Megauploads as an example.  In that case the service threatened mainly the dominant political player (the US) yet it was still possible to completely kill it and do so overnight.

It would also be fairly trivial, I believe, for a great deal of grief to be heaped upon those end-users attempting to use Bitcoin natively, and do so at the ISP and even the backbone level.  Do you really want to fire up a VPN to make a micro-transaction?  ...even if it remains completely trivial to use you favorite VPN which is far from certain and, I believe, even unlikely...

The internet will not go away absent a Mad-max scenario.  It is to critical.  But it would not be very hard to make it much less free.  Something like Bitcoin has the potential to threaten pretty much all central governments (who are pretty universally cronies of the modern economic systems which have developed) so I think that counting on jurisdiction hoping is a risky gamble.

A low data-rate solution operated by low capitalized independent specialists is possible under almost any realistic regime of increased network control.  And second-tier providers have a lot more flexibility to adapt to on-the-ground and evolving threats.  This is why I see such a structure as having a realistic possibility of enduring in most possible future scenario.  A solution where a set of highly capitalized entities operating Bitcoin natively is, to me, a very ginger solution which relies on the kindness of it's adversaries for survival.

 edit: characterize specialists, end-users, and spelling.