Topic: Bitcoin download speed singularity inevitable with Bitcoin Classic (Read 1334 times)

I`ve already had a thread talking about download speed singularity, but now let's apply this problem to Bitcoin Classic.

Block size download speed singularity = Block size is so big, that it takes more than 10 minutes to download 1 block, so that a new block will form while you barely downloaded the previous one. Which means that you can never download the blockchain, and it will become an isolated, centralized entity, as nobody could download it, but the ultra fast internet owners, or the blockchain breaks up.

This is similar to how our universe is expanding, the OBSERVABLE UNIVERSE is an inverse singularity, because nothing can escape it ,and everything is trapped in it. Therefore, if the block size were to become like 1 TB, then obviously it will become impossible to download it in 10 minutes and therefore bitcoin ends as we know it.

---

Ok now this is a remote situation, and I get it , blassic wants a  2-4-8 mb block size, and the situation with 1 TB block size will be far in the future.

However there is another singularity, that is much closer, and much more worrysome: The Blockchain Size Download Speed Singularity

The Blockchain Size Download Speed Singularity = The blockchain itself is so big, that it is impossible to download it all from the start, and by the time you download it, it grows even more. So in the previous issue, you already had the blockchain somewhat up to date and you only needed the new blocks, which was impossible to download. But in this example, you are a fresh user, and want to grab the entire blockchain, which it will become impossible, because the daily max bandwidth will be a tiny fraction of the entire blockchain size.


This means that if the block size becomes say 64mb in the foreseeable say 6-7 years that would mean that it grows 9 GB/DAY, and if you want to download it from the start, a blockchain of 1-2 petabyte size by then, it would become impossible to download it, or very unfeasable.

This means that only existing nodes can hold the bitcoin, and there will be no new nodes, and it will become instantly centralized.

---

So if the Classic people want eventually 64-128-256 mb blocks, then this singularity will become inevitable, and it would destroy decentralized bitcoin as we know it.

This is another huge flaw Classic has, but it has been swept under the rug, so I`m here to expose this flaw.

Keep this in mind when you are talking about internet speeds and running a fully functional Bitcoin node. If you don't limit the number of connections to your node, it will often eat all of your available upload bandwidth. Once your upload bandwidth is saturated, your download bandwidth becomes practically useless.

I have been running a full node since 2011. I have top tier internet speeds (far, far better than the averages posted above) in a well developed location. Sometime in the past year or so I've had to limit the number of connections in order to conserve bandwidth for other regular internet uses. This is obviously with the 1MB anti-spam limit in place.

My concern is how much more bandwidth is my node going to require if we remove this anti-spam limit without addressing the efficiency of the network. My concern is if a Bitcoin fanatic like myself isn't able to run a fully functional node with his top tier home internet, aren't we harming decentralization? I understand that there are many efficiency improvements in the pipeline, but as of right now I still have to gimp my node and limit the amount of connections it will accept.

I've seen you talking about Netflix and modern video games, and the bandwidth required to use those services, several times in the past and I just want to remind you that the Bitcoin network needs peers who upload large amounts of data as well.

average internet speeds of 1999:
256kbit down 64kbit up (32kbyte down 8kbyte up) = (19meg down per 10 minutes 4.8megUP per 10 minutes)
(usually throttled down, but yes its possible to download an 5mb MP3 in 5 minutes at those speeds)

average internet speeds of 2015:
5mbit down 750kbit up (625kbyte down 93kbyte up) = (375meg down per 10 minutes 55megUP per 10 minutes)
(usually throttled down, but yes its possible to watch 3gb netflix in an hour (500mb in 10 minutes at those speeds)

so we can see a 10x growth in less than 20 years..
and the dooms day of large blocks such as the 1tb block the OP is prophesying. but lets say that in we grw to 4mb in 4 years, 8mb in 4 years and so on and so on.

then the average speed in MANY decades WILL cope.

i really love the blockstreamers and altcoiners doing all they can to shout out that onchain bitcoin's cant cope and to instead use their offchain/altcoins

lesson one to fudsters:
dont try grabbing a doomsday prophacy of 20+years to denounce possibility of a smaller growth now.
coz thats like trying to say that child abortion should happen because obesity and cancer will kill most people in 50 years so there is no point in anyone growing up

lesson two
try to us logical and rational and HONEST information about CURRENT situation to denounce CURRENT debate about CURRENT proposals

I agree about the liability statement.  Sure it's great for some full nodes (oracles) to exist with history since the beginning of time, but that's too much burden for average user as time goes on.

It seems to me that it should be possible to start/run a new node with only:
  1) The utxo set starting at block X in the past.  where X is at user's discretion.
  2) the full TX history since block X.

Normally X should be something like a month in the past or at least  a week to avoid problems with reorgs.  If a reorg did occur that is deeper than node history then I suppose it should just stop and issue an error message.

Yes it requires some amount of trust since one is not downloading entire history.   But it seems like the hash of each block and each corresponding utxo set could be distributed in enough places (peers, oracles) that one could be relatively certain.  I mean it would be faster to connect and check with eg 500 peers/oracles than to download and verify the entire blockchain.

I believe this should be significantly faster for initial use than even a pruned node because pruned nodes still require a  sync from genesis block for verification.

So what's wrong with this idea?  I imagine that it or something like it has already been proposed.   Does it have a name?

Can the blockchain still be downloaded via http and/or torrent?    I seem to recall that used to be possible, with some sort of published hash for verification...?

The trick with any scheme like this (download or on usb/cd) is establishing trust that the blockchain hasn't been tampered with.   There must be a simple way to verify via 3rd party tool(s).

bitcoin can 'parse' validate fine on a 1.2ghz raspbeery Pi.. (basically phone technology). even a 2005 desktop is 2 times faster(3ghz dualcore)
and as for the cheapest desktop today. well thats atleast 4x faster (3ghz quarcore)

again not an issue

oh never thought about that but yes they could actually transform this into a business, by selling usb stick with a pre-filled blockchain

The Blockchain Size Download Speed Singularity = The blockchain itself is so big, that it is impossible to download it all from the start, and by the time you download it, it grows even more.

Thanks for exposing this flaw.
Can't we just have bitcoin2mb

For now,
Just 2mb (with input cap?) nothing else.

Ok you want more nearby issues, fine.

What about parsing speed, you know that the blocks all have to be parsed as well, which requires not only good internet bandwidth, but also good processor speed.

So this issue could make all services slower, if they dont have the best available processors.

There is already a problem with node count decline, and more people will stop using full client, because not only the client is downloading the chain slow, but it also slows down PC while it calculates it.

Not to mention what the miners have to go through with huge blocks, its a lot of computing power, that could be more evenly spread out via a function of time, rather than cramming up a lot of block space which may or may not be processed in 10 minutes.

Perhaps but imagine how silly would that be. People would sell Bitcoin Blockchain part 1 , part 2 , part External HHD drives in stores each 100 petabyte.

How silly and ackward would that be to go into a tech store and buy bitcoin blockchain filled external hard disks, because the blockchain would be so big that you could not download it.

Of course eventually that will happen, but with 1 mb block this scenario would be in like 200 years, not 20.

whats silly is you using a doomsday scenario of 20+ years time to avoid expanding bitcoin this year.

please go to an abortion website and tell everyone abortions are good because it avoids the pain of their child having a heart attack or getting hit b a car in their 30's. because to me thats what the white noise of your rhetoric sounds like

Perhaps but imagine how silly would that be. People would sell Bitcoin Blockchain part 1 , part 2 , part External HHD drives in stores each 100 petabyte.

How silly and ackward would that be to go into a tech store and buy bitcoin blockchain filled external hard disks, because the blockchain would be so big that you could not download it.

Of course eventually that will happen, but with 1 mb block this scenario would be in like 200 years, not 20.

i don't see the issue with the blockchain size, which is growing by something like only 10gb-20gb per year

and you only need to download it once, then it can be easily given to your friend with a simple usb stick, there is no need for each person in the planet to download it

i don't see the issue with the blockchain size, which is growing by something like only 10gb-20gb per year

and you only need to download it once, then it can be easily given to your friend with a simple usb stick, there is no need for each person in the planet to download it

i don't see the issue with the blockchain size, which is growing by something like only 10gb-20gb per year

and you only need to download it once, then it can be easily given to your friend with a simple usb stick, there is no need for each person in the planet to download it

So the internet may grow 20x in the next years, but it may only grow 2x. I already have 20 mb/s internet, optical fiber.

So there is not much room left, yes maybe the cable structure can be improved ,and the network servers to be more efficient, but it will eventually hit a limit.

This is actually a pretty good argument.

We can't have infinitely scaling blocks, but something that would slowly increase the block size once blocks get over 92% full wouldn't be bad.

I would like to point out that fiber-optic cables are going to be increasing the download speeds of normal people over the next decade, so maybe having the blockchain as >1Tb isn't something that would be completely detrimental? Sure they wouldn't be able to implement it quickly, but it could be implemented.