How does block size harm decentralization?

[c_atlas]

From what I understand, a larger block size would mean more transactions (or data in general) can be included in each block effectively increasing the number of transactions the network can handle in a finite time frame (at least 1 block long). Consequently, the size of the blockchain will increase at a greater rate since more data is being added everytime a block is accepted. Eventually, this would mean the blockchain would be too large to store so only dedicated companies will be able to have full nodes, resulting in the centralization of mining.

Wouldn't that happen eventually anyway? Sure it'll take longer but unless the chain is pruned how will people be able to store it?

The only people running full nodes now are enthusiasts who like verifying everything themselves (though it's getting harder since the blockchain is already pretty big), pool operators, and mining farms or services like NiceHash/Blockchain.com

This is not a post in support of a change in block size or for sidechains, I'm not qualified to take a side in that debate at all lol. What are the consequences of a block size increase and more specifically how it would be harmful for bitcoin's decentralization?

[Seymour Locksmiths]

I think the idea is that advances in networking technologies will mean we can transfer more data in the same amount of time as. As long as networking capacities grow quicker than the blockchain then we will be able to avoid the problem you have described.

[franky1]

the debate is that those that want people to be pushed off bitcoin and into other networks will say bitcoin cant scale and blockchains ar broke because they want people to use commercialised networks/services.

the myth presented about bitcoin is the whole "gigabytes by midnight" fud that is being pushed by those commercial lovers, to make is sound like bitcoin has to remain at 1mb base block or jump to gigabytes.

the funny part is that progressive SCALING (step by step growth) like what WAS allowed 2009-2015 (0.25mb, 0.5mb, 0.75mb,1mb) pretty much went stagnant at 2015

even the supposed 'segwit' has not achieved a scaling solution ONCHAIN and we are still stuck at the implied limit of 600k transactions a day, while segwits true purpose is to be used as the gate way tx format for an alternative network

even though if you do the math of sending a full block to another user 1block of 1mb legacy transactions every ~10 minutes is1.66kb/s (even dialup can handle that)

so yea those against BITCOIN adoption/scaling are not really true 'bitcoiners' as they prefer to be stuck shouting out that bitcoin cant scale beyond 1999's technology

the only reason to keep to the implied 600k tx a day, is purely to incentivise commercial services to advertise their alternative networks/services while also pushing the fee's of bitcoin up to keep users from wanting to utilise the bitcoin network

[c_atlas]

I think the idea is that advances in networking technologies will mean we can transfer more data in the same amount of time as. As long as networking capacities grow quicker than the blockchain then we will be able to avoid the problem you have described.

I don't think the network is the problem, sure initial download of the blockchain would take a really really long time (especially if you verify every tx signature,  but that can be skipped anyway), but day to day broadcast of blocks and transactions would be fine.

My understanding is that it's just a disk storage issue.

even though if you do the math of sending a full block to another user 1block of 1mb legacy transactions every ~10 minutes is1.66kb/s (even dialup can handle that)

so yea those against BITCOIN adoption/scaling are not really true 'bitcoiners' as they prefer to be stuck shouting out that bitcoin cant scale beyond 1999's technology

Again I don't think it's a network issue. If the block size was 8mb like in BIP-101 then it would be 13.33kb/s. I'm certain modern day CPUs can handle the tx verification even if we increased the # of txs a lot, and there really doesn't seem to be a network issue unless I'm totally missing something; so it must be storage.

[franky1]

I think the idea is that advances in networking technologies will mean we can transfer more data in the same amount of time as. As long as networking capacities grow quicker than the blockchain then we will be able to avoid the problem you have described.

I don't think the network is the problem, sure initial download of the blockchain would take a really really long time (especially if you verify every tx signature,  but that can be skipped anyway), but day to day broadcast of blocks and transactions would be fine.

My understanding is that it's just a disk storage issue.

even though if you do the math of sending a full block to another user 1block of 1mb legacy transactions every ~10 minutes is1.66kb/s (even dialup can handle that)

so yea those against BITCOIN adoption/scaling are not really true 'bitcoiners' as they prefer to be stuck shouting out that bitcoin cant scale beyond 1999's technology

Again I don't think it's a network issue. If the block size was 8mb like in BIP-101 then it would be 13.33kb/s. I'm certain modern day CPUs can handle the tx verification even if we increased the # of txs a lot, and there really doesn't seem to be a network issue unless I'm totally missing something; so it must be storage.

its not even about storage
1mb base block =52gb a year(4mb=~200gb)
i can get a 256gb microsd card(size of a fingernail)
i can get a couple terrabyte hard drive for less than a week of groceries cost

funny thing is those worried about 1mb+=bad, before 2017 suddenly shut up when it became convenient to allow 4mb to allow the gateway to another network..

yet millions-billions do facetime/stream HDTV,they play online games and upload the gaming while narrating while streaming HD content all the same time.
and happily would take up hundreds of gigabytes of data. just so they can run around a map shooting people

but yet strangely the rhetoric of same data, same speeds becomes suddenly impossible in regards to bitcoin

[squatter]

Wouldn't that happen eventually anyway? Sure it'll take longer but unless the chain is pruned how will people be able to store it?

The chain can be pruned. Whether full nodes are storing the entire blockchain isn't the major issue. Storage space is a bottleneck and can stop people from running full nodes, but bandwidth and latency are more pressing issues.

The only people running full nodes now are enthusiasts who like verifying everything themselves (though it's getting harder since the blockchain is already pretty big), pool operators, and mining farms or services like NiceHash/Blockchain.com

Exchanges and basically any service who credits accounts based on the blockchain need to run full nodes.

I'd say that the more people that run full nodes, the better. Without full nodes enforcing the consensus rules, miners and economically important nodes can collude to change the protocol. If the chain will be hard forked, SPV nodes have zero say about it. Only full nodes are relevant in hard fork attempts.

This is not a post in support of a change in block size or for sidechains, I'm not qualified to take a side in that debate at all lol. What are the consequences of a block size increase and more specifically how it would be harmful for bitcoin's decentralization?

Larger blocks cause propagation delays. This increases orphaning rates, which disproportionately hurts smaller miners and will eventually force them off the network via unprofitability. Miners can sidestep these problems by centralizing -- pooling hashpower, colocation. Less mining entities, each with larger market share = bigger threat of transaction censorship, government targeting of miners, 51%/selfish mining attacks, political hard fork attacks, etc.

Larger blocks also cause increased bandwidth requirements. Data upload limitations are the primary bottleneck here. People who don't have access to unlimited internet data will become increasingly likely to shut down their nodes as it begins to interfere with streaming, gaming, file sharing, etc. The less full nodes there are, the easier the consensus is to attack.

[pooya87]

block size does not harm decentralization, how block size is increased can do that. for example if you increase it suddenly to a much bigger number now it will end up centralizing bitcoin but if it is increased slowly with the advancement of hardware and internet speed then it can't do much harm.

the most important thing to remember is that size of the blockchain is only one of the problems. the bigger problems is the hardware limitations. if we have 10 MB blocks for example you will have to verify 10 MB worth of transactions every 10 minutes on average and also be able to receive and send that much every 10 minutes. additionally miners will start facing problems, specially those with slower internet speeds who have to compete with those with high speed connections and that leads to miner centralization.
so the numbers you are reciting above (13.33kb/s, etc) are only when you assume people download 1 block per 10 minute but that doesn't make any sense. not for miners not for others. and it is not just downloading blocks, a node does a lot of other things that takes traffic including relaying transactions in its mempool

[squatter]

block size does not harm decentralization, how block size is increased can do that. for example if you increase it suddenly to a much bigger number now it will end up centralizing bitcoin but if it is increased slowly with the advancement of hardware and internet speed then it can't do much harm.

There's still a larger problem here that most people don't discuss. Even if hardware advancement allows us to significantly increase block sizes over time, that doesn't address whether doing so is compatible with Bitcoin's hard cap on supply.

Without inflation, the system needs fee revenue to continue incentivizing miners. The block size limit is the only means we have to enforce scarcity of block space, which guarantees fee revenue. Otherwise, Bitcoin's Byzantine fault tolerance may be threatened as block rewards decline in value. You can't have a network worth many billions or trillions of USD where miners have no incentive to secure the network.

[pooya87]

block size does not harm decentralization, how block size is increased can do that. for example if you increase it suddenly to a much bigger number now it will end up centralizing bitcoin but if it is increased slowly with the advancement of hardware and internet speed then it can't do much harm.

There's still a larger problem here that most people don't discuss. Even if hardware advancement allows us to significantly increase block sizes over time, that doesn't address whether doing so is compatible with Bitcoin's hard cap on supply.

Without inflation, the system needs fee revenue to continue incentivizing miners. The block size limit is the only means we have to enforce scarcity of block space, which guarantees fee revenue. Otherwise, Bitcoin's Byzantine fault tolerance may be threatened as block rewards decline in value. You can't have a network worth many billions or trillions of USD where miners have no incentive to secure the network.

in my opinion higher fees will kill bitcoin so if block size scarcity is enforced in the future*, remember that bitcoin was meant to be a decentralized currency not something people only trade or store value in so they don't care if they pay a high fee to transfer it. it may also nullify what you are saying since people would stop using bitcoin and consequently the price would drop and there wouldn't be any more profit for them miners anymore.
bigger capacity also means more transactions, and that means more transaction fees in total.

* note that i am saying in the future, as of today and in the close future we don't need block size increase or even fees to replace the block reward since it is quite high still and will remain high for many years.

[franky1]

block size does not harm decentralization, how block size is increased can do that. for example if you increase it suddenly to a much bigger number now it will end up centralizing bitcoin but if it is increased slowly with the advancement of hardware and internet speed then it can't do much harm.

There's still a larger problem here that most people don't discuss. Even if hardware advancement allows us to significantly increase block sizes over time, that doesn't address whether doing so is compatible with Bitcoin's hard cap on supply.

Without inflation, the system needs fee revenue to continue incentivizing miners. The block size limit is the only means we have to enforce scarcity of block space, which guarantees fee revenue. Otherwise, Bitcoin's Byzantine fault tolerance may be threatened as block rewards decline in value. You can't have a network worth many billions or trillions of USD where miners have no incentive to secure the network.

what your not realising is that you cant have a network that is core demandd to stick with under 600k transactions a day as that then just makes the network only useful to under 600k people wanting to use/monitor/care about bitcoin daily.

people end up using other networks that are cheaper and able to process more than 600k transactions a day and bitcoin is left stranded as the coin no one wants to return to because the costs become too high..

LN thunderdome: btc and LTC iou's may enter, only LTC iou's may leave

secondly the REAL incentive to pools is not for the next few decade requiring fee's. the block reward is sufficient enough
(in short dont even bother bringing 'fee's for pools' into a scaling debate unless its a few decades in the future)

thirdly fee COST USERS funds, not incentivise users.
arguing that things need to be done for user benefit, then push that argument aside to then say fe's need to rise to benefit pools. is just playing games and empty excuses as to why scaling should be stalled/stagnant

again forcing utility to remain at 600k a day transactions helps no one, incentivises no one and actually makes peopl want to care less about bitcoin if they cant use it as and when they like.

so if it costs something just to monitor 600k peoples transactions as a user.. and then costs a user $1+ just to use it themselves. then its the fee's that will push people away far sooner.

the real goal would be to
allow more transactions
allow the combined (lower fees) to accumulate to be more as a total.

thus making bitcoins blockchain USEFUL for users. and also cheap to use for users.thus users continue to be full nodes

which after years of progressive growth would then become enough to cover costs for pools eventually(when pools actually need it)

[Kakmakr]

A gradual increase in the Block size was a old method to scale, since then new innovations like second layer applications <Lightning Network> was introduced to reduce the burden on the Blockchain to prevent a scenario where the Blockchain grows too fast for people to be able to affordably run nodes with big enough storage space. <It is also significantly faster than the Blockchain>

People like Roger Ver and supporters of Block size scaling will learn the hard way, that it is not a good strategy for the long-term and definitely not if it goes mainstream. <Currently BCash has huge Block sizes and only a small percentage of blocks are used to capacity>

[franky1]

A gradual increase in the Block size was a old method to scale, since then new innovations like second layer applications <Lightning Network> was introduced to reduce the burden on the Blockchain to prevent a scenario where the Blockchain grows too fast for people to be able to affordably run nodes with big enough storage space. <It is also significantly faster than the Blockchain>

LN is NOT a bitcoin network. it is a thunderdome network for multiple coins.
for those that dont get the movie reference.. its a fortknox custodial banking system of not 100% control

which once you enter it. you wont want to leave and obtain BTC on exit. because it will become too impractical/expensive(fees) to close out and withdraw bitcoin from the locks

people will end up just atomic swapping to altcoin ious(12 decimal unconfirmed LN payments) within LN and exit using LTC or other coins that have been modified to have gateways to LN but have more transaction counts and cheaper fee's

oh by the way. those wanting to be full nodes for LN will end up actually harbouring multiple chains to allow channels to atomic swap. so if bitcoin for instance is too much to handle. imagine full node running multiple chains at once..
stick that scenario in your play book and run with it.. itll shock you in a hard way

LN's design is to take utility AWAY from bitcoin

People like Roger Ver and supporters of Block size scaling will learn the hard way, that it is not a good strategy for the long-term and definitely not if it goes mainstream. <Currently BCash has huge Block sizes and only a small percentage of blocks are used to capacity>

^ blah blah blah social drama distraction.

im guessing your the kind of person that only has one plate, one knife and fork and one shelf in your kitchn because you only plan on minimal reserves.

smart people have extra wardrobe space in their bedroom. a full plate and cuttlery set in the kitchen so that they can cope with change and growth/expansion rather then race around at the last minut when theings happen

EG when you buy a computer do you buy a computer with the minimalist of features or do you try to buy a computer that has more than you need so that yor covered for future changes?

[Kakmakr]

A gradual increase in the Block size was a old method to scale, since then new innovations like second layer applications <Lightning Network> was introduced to reduce the burden on the Blockchain to prevent a scenario where the Blockchain grows too fast for people to be able to affordably run nodes with big enough storage space. <It is also significantly faster than the Blockchain>

LN is NOT a bitcoin network. it is a thunderdome network for multiple coins. which once you enter it. you wont want to leave and obtain BTC on exit. because it will become too impractical/expensive(fees) to close out and withdraw bitcoin from the locks

people will end up just atomic swapping to altcoin ious(12 decimal unconfirmed LN payments) within LN and exit using LTC or other coins that have been modified to have gateways to LN but have more transaction counts and cheaper fee's

oh by the way. those wanting to be full nodes for LN will end up actually harbouring multiple chains to allow channels to atomic swap. so if bitcoin for instance is too much to handle. imagine full node running multiple chains at once..
stick that scenario in your play book and run with it.. itll shock you in a hard way

People like Roger Ver and supporters of Block size scaling will learn the hard way, that it is not a good strategy for the long-term and definitely not if it goes mainstream. <Currently BCash has huge Block sizes and only a small percentage of blocks are used to capacity>

^ blah blah blah social drama distraction.

No franky1, I know you are 100% against the LN and I respect that, but you should also be able to see that Block size scaling is not the way to go, if we want Bitcoin to go mainstream. If the LN can only handle most of the micro payments, then we have already made significant progress. <Let's give it some time to mature and to show it's advantages over Block size scaling. >

Yes, every method has it's advantages and disadvantages, but stagnating on one method <because it was Satoshi's original idea> is not very innovative or creative at all. We have talented developers with lots of new ideas that might improve on Satoshi's original implementation. 

[franky1]

No franky1, I know you are 100% against the LN and I respect that, but you should also be able to see that Block size scaling is not the way to go, if we want Bitcoin to go mainstream. If the LN can only handle most of the micro payments, then we have already made significant progress. <Let's give it some time to mature and to show it's advantages over Block size scaling. >

Yes, every method has it's advantages and disadvantages, but stagnating on one method <because it was Satoshi's original idea> is not very innovative or creative at all. We have talented developers with lots of new ideas that might improve on Satoshi's original implementation.  

3 years after we hit the implied tx count limit and bitcoins network has not improved.. instead "use this other network and stop using bitcoin".... (facepalm) thats a 200 year old business model which we all know how it ended

gold is limited utility, lock it up and use these promissory notes, oh and by the way it will be expensive if you want your gold back so here have some silver and nickel if you dont wanna play with promissory notes

[squatter]

There's still a larger problem here that most people don't discuss. Even if hardware advancement allows us to significantly increase block sizes over time, that doesn't address whether doing so is compatible with Bitcoin's hard cap on supply.

Without inflation, the system needs fee revenue to continue incentivizing miners. The block size limit is the only means we have to enforce scarcity of block space, which guarantees fee revenue. Otherwise, Bitcoin's Byzantine fault tolerance may be threatened as block rewards decline in value. You can't have a network worth many billions or trillions of USD where miners have no incentive to secure the network.

in my opinion higher fees will kill bitcoin so if block size scarcity is enforced in the future*, remember that bitcoin was meant to be a decentralized currency not something people only trade or store value in so they don't care if they pay a high fee to transfer it. it may also nullify what you are saying since people would stop using bitcoin and consequently the price would drop and there wouldn't be any more profit for them miners anymore.

bigger capacity also means more transactions, and that means more transaction fees in total.

* note that i am saying in the future, as of today and in the close future we don't need block size increase or even fees to replace the block reward since it is quite high still and will remain high for many years.

Right now, the cost per confirmed transaction (based on actual mining expenditure) is much higher than the fees users are paying. Miners are subsidizing that cost because their revenues are temporarily being subsidized by inflation. However next year, inflation will drop to 12.5% of the original reward and it quickly drops off after that. When should we expect fees to begin making up the difference? The design can't remain completely dependent on fiat speculation forever. At some point, users need to pay the actual cost of transactions if they expect miners to continue securing the chain. The sooner users get used to higher fees, the smoother the transition to a deflationary Bitcoin will be.

If we keep increasing block size, "more transactions" doesn't imply a requisite increase in fees. That all depends on how limited block space is. If we increase block size beyond demand because technology advancement allows it (Like Bitcoin Cash), fees will drop towards zero. We need full blocks and a constant transaction backlog, otherwise there is no fee pressure.

Now, what evidence do you have that "higher fees will kill Bitcoin?"

[nutildah]

This is not a post in support of a change in block size or for sidechains, I'm not qualified to take a side in that debate at all lol. What are the consequences of a block size increase and more specifically how it would be harmful for bitcoin's decentralization?

I think you hit on the answer earlier in your post: bigger blocks = a bigger blockchain = less people running full nodes = more centralization.

Somewhat humorously, BCH blocks never come close to getting filled and are on average much smaller than BTC blocks. BSV block size limit is even bigger and has even fewer transactions.

I think both the on- and off-chain scaling solutions are a much better approach than increasing the block size limit, for the very reason that its a bad idea to let the blockchain become exponentially larger. Right now its growing at a linear rate and it should be kept that way.

Of course there are those who absolutely insist on "staying true to Satoshi's white paper," which itself is not perfect and merely an outline for the idea of what went on to become bitcoin. I see increasing block size as merely the twiddling of an ancient parameter and rather a stagnant approach to a technology still largely under development.

[bots1]

the most important thing to remember is that size of the blockchain is only one of the problems. the bigger problems is the hardware limitations. if we have 10 MB blocks for example you will have to verify 10 MB worth of transactions every 10 minutes on average and also be able to receive and send that much every 10 minutes. additionally miners will start facing problems, specially those with slower internet speeds who have to compete with those with high speed connections and that leads to miner centralization.
so the numbers you are reciting above (13.33kb/s, etc) are only when you assume people download 1 block per 10 minute but that doesn't make any sense. not for miners not for others. and it is not just downloading blocks, a node does a lot of other things that takes traffic including relaying transactions in its mempool

Your opinion makes sense, even today mining still uses simple hardware. So that mining difficulties can still be overcome by cheap and many shops. I think why "segwit" in August 2017 changes the size of transactions from 1MB to weight units and separates between shipping transactions with signatures. If in the future the block size becomes larger, the possibility of mining hardware also does not use simple tools, consequently the full chain will only be owned by large companies.

[Artemis3]

This is not a post in support of a change in block size or for sidechains, I'm not qualified to take a side in that debate at all lol. What are the consequences of a block size increase and more specifically how it would be harmful for bitcoin's decentralization?

I think you hit on the answer earlier in your post: bigger blocks = a bigger blockchain = less people running full nodes = more centralization.

Somewhat humorously, BCH blocks never come close to getting filled and are on average much smaller than BTC blocks. BSV block size limit is even bigger and has even fewer transactions.

I think both the on- and off-chain scaling solutions are a much better approach than increasing the block size limit, for the very reason that its a bad idea to let the blockchain become exponentially larger. Right now its growing at a linear rate and it should be kept that way.

Of course there are those who absolutely insist on "staying true to Satoshi's white paper," which itself is not perfect and merely an outline for the idea of what went on to become bitcoin. I see increasing block size as merely the twiddling of an ancient parameter and rather a stagnant approach to a technology still largely under development.

Ironically here i am, weeks into syncing the blockchain and run a full node, but I'm barely halfway there (84g). I like bitcoin core solution, and the market likes bitcoin core as well regardless of what fork supporters say.

This is not "a war" though. It is about CHOICE. If you have a "better" idea, you are free to fork and see if others follow you. So far what others have proposed is clearly not good enough to take over the crown. Bitcoin as it is handled by the core team remains The King.

LN provides neat things like (true) decentralized exchange, you could theoretically directly exchange litecoin for bitcoin and back without using any web site (wallet to wallet). LN is optional, you may use it or not. Either way, providing this option is already offloading the on chain network enough for it to scale. Each shop could run its own LN node, and why not, a full Bitcoin node along it (and while you are at it an Electrum server pointed to your node).

Perhaps someone will sell some raspi pre-installed with everything so its plug and play and shops can have their solution ready, or same as there is a linux distro for "miners", a linux distro foe "nodes" (or the same distro could do both).

Anyway, enlarging blocks is dumb, I wish newbies stopped requesting that as a "solution" for a "problem" that is no more... Spend that energy pushing for faster wallet LN adoption...

[c_atlas]

Larger blocks cause propagation delays. This increases orphaning rates, which disproportionately hurts smaller miners and will eventually force them off the network via unprofitability. Miners can sidestep these problems by centralizing -- pooling hashpower, colocation. Less mining entities, each with larger market share = bigger threat of transaction censorship, government targeting of miners, 51%/selfish mining attacks, political hard fork attacks, etc.

Larger blocks also cause increased bandwidth requirements. Data upload limitations are the primary bottleneck here. People who don't have access to unlimited internet data will become increasingly likely to shut down their nodes as it begins to interfere with streaming, gaming, file sharing, etc. The less full nodes there are, the easier the consensus is to attack.

What's the largest contributing factor to increased propagation delay? Is it transaction verification, a bandwidth issue, maybe something else? Verifying more transactions would slow down block creation, but I don't think it would by too much. As for bandwidth, miners don't send out blocks to every node, they send them to their peers who then propagate them further, so I'm not quite sure I understand how bandwidth is a problem since there are plenty of nodes transmitting and receiving data.

block size does not harm decentralization, how block size is increased can do that. for example if you increase it suddenly to a much bigger number now it will end up centralizing bitcoin but if it is increased slowly with the advancement of hardware and internet speed then it can't do much harm.

Without inflation, the system needs fee revenue to continue incentivizing miners. The block size limit is the only means we have to enforce scarcity of block space, which guarantees fee revenue. Otherwise, Bitcoin's Byzantine fault tolerance may be threatened as block rewards decline in value. You can't have a network worth many billions or trillions of USD where miners have no incentive to secure the network.

Agreed, but why would users want to pay massive fees to cover the miners expenses, especially if they can use other networks for far cheaper (I know I know "they get a secure decentralized network" but still, does this limit the types of transactions that will be capable on the bitcoin network? Is this where LN comes in?). Is it fair to assume that up til now (and maybe for a few more years) miners have been taking the hit in terms of fees with the belief that their accumulated BTC will be worth more in the future than their current mining operation has cost them (i.e a net positive in the future)?

Ironically here i am, weeks into syncing the blockchain and run a full node, but I'm barely halfway there (84g). I like bitcoin core solution, and the market likes bitcoin core as well regardless of what fork supporters say.

I installed core and was downloading the blockchain a few months ago on my laptop to try to use JSON-RPC with a python/bitcoin library, I had no idea it was over 200GB and eventually while bitcoind was getting blocks and verifying transactions I got a message saying my laptop was out of space 

On top of that I didn't have an internet connection at my place so I was downloading it at Starbucks and stuff so it took a couple weeks to get that far, yet it only took a few seconds to delete nearly the whole blockchain  I felt so stupid when I realized I didn't check how much storage my SSD had...

[mu_enrico]

Decentralization is an ambiguous word. I'm afraid this problem is more "political" than technical. It's about what kind of device (money) that you are willing to invest or include in the bitcoin network. The bigger the block, the more infrastructure you need to be able to process the block.

The current system where every user is a network node is not the intended configuration for large scale.  That would be like every Usenet user runs their own NNTP server.  The design supports letting users just be users.  The more burden it is to run a node, the fewer nodes there will be.  Those few nodes will be big server farms.  The rest will be client nodes that only do transactions and don't generate.

Source: https://bitcointalk.org/index.php?topic=532.msg6306#msg6306

Above is Satoshi's opinion. Core team might have a different opinion about this matter, but I'm not interested in politics. "Big blocker" and "small blocker" could be correct at the same time. Let's see what will happen in the next few years.