Topic: blockstream - wants to tax you and become the new Bitcoin oligarchy - page 2. (Read 8953 times)

Bandwidth.


Right but should we jump from 1 MB to 8 MB?


Transaction backlog is there and legitimate transaction gets rejected or take longer than usual.

except for the one just now.

bullshit!!!!!!! they never get filled! even with the spam from coinwallet.eu!
i don't scratch if i dont have an itch!

The idea of a jump (from 1 to 8 in this case) is to allow for temporary peaks in transaction volume.
The hard limit can only be raised in a hard fork and we don't want to fork every year or now and then.
The hard limit must be such that the network should feel uneasy operating at full capacity, but still workable.

In fact I'm more inclined to think that forking every 4 years as "Bitcoin halving celebration" is actually a good idea.
So my proposed solution would to be increase the hard limit to 8Mb (with soft limit of 4Mb for initial ramp up) and then re-evaluate every 4 years.

blocks sometimes do get filled up and we'll see more and more
of that as Bitcoin grows.  You want to wait till its a dire situation before expanding
the bandwidth?  wtf!!!

problem: when you dont qualify, you stfu..

The CEO of Intel suggested 'Moore's Law' is finally coming to an end
http://uk.businessinsider.com/intel-ceo-brian-krzanich-suggests-moores-law-is-over-2015-7?r=US&IR=T

Intel scientists find wall for Moore's Law
http://www.cnet.com/news/intel-scientists-find-wall-for-moores-law/


edit: bonus: The impending end of Moore's Law is not Intel's biggest problem: http://www.itworld.com/article/2949368/hardware/the-impending-end-of-moores-law-is-not-intels-biggest-problem.html

what? the miners don't have bills?

we don't fully use the 1mb and out of gavin(CIA)'s ass we jump to 8mb?

wtf!!!

What's a problem?  Storage space required to run a node? 

It is still a problem.

According to Gavin, storage space won't be the bottleneck in the future , but rather bandwidth.
"The (pruned) ledger only takes up 1 gig and will be further optimized."

BIP 101: Doubles maximum block size every two years.

IMHO, its not a good idea. CMIIW.


I know, maximum block size as 8 MB does not necessarily mean all blocks after that will be 8 MB. IMHO, its better for a gradual increase than a jump. However, I might support 8 MB but I am still opposed to BIP 101.

Moore's law? It looks like you aren't up to date with that one.
This explains it as well. My quote was from Wikipedia, as I failed at finding the article that I have recently read.

lol these people are so far away from reality, talking about things they will never grasp, it is dramatically staggering.

hopefully bitcoin will absorb this nonsense and rise from the xt ashes.

Intel conceded that Moore's law as applied to silicon lithography ran out of legs, the story came out a few weeks ago. Their next "tock" in their world renowned "tick-tock" development strategy has been re-scheduled to become a "tick" for the first time in history.

Exponential growth is not the same thing as infinite growth. Because exponential trends necessarily end when they hit the apex.

There wasn't a need for 1Mb limit 4 years ago either, we could have done away with 256Kb or 512Kb just fine. In fact, up until recently the hard 1Mb cap wasn't effective at all as it never was approached. But smaller soft limits were all hit and raised. That's how Bitcoin survived.

As this time we are discussing rising the hard limit (via hard fork) we need to future proof it a little bit. In this light, 8Mb seems reasonable as a new hard limit for the next 4 years, in conjunction with smaller soft limits that miners can agree to maintain internally to prevent network spamming and abuse. They can, for example, agree to not build on top of anything larger than 4Mb initially, however full nodes will still accept and propagate blocks all the way up to 8Mb. This way the infrastructure will be ready for soft expansion in the future.

Whether to expand beyond 8Mb, when and how is questionable and depends on many factors. Maybe a series of hard-forks down the road will be a more appropriate and responsible solution than trying to predict the technology, but anything less than 8Mb for the upcoming years might start hurting Bitcoin and deflect its user base towards other solutions. We, as a community, definitely don't want that.

The only reason that BIP101 is bad is that there is the possibility that it will be attempted to be implemented without consensus. There is no real reason why the network would not be able to support 8MB blocks on a technical basis and Moores law would imply that technology will advance quickly enough so that the network will continue to be able to support the increasingly large max block size and if it doesn't then then Bitcoin can be formed again.

Yes. We need to have bigger block size to handle a high number of transaction but we can't just increase to 8 MB because of that. Its always better to increase block size gradually. Now, there is no need for 8 MB and doubling every X years does not make sense. We need dynamic block size, of course, but a better one.

8Mb is what Bitcoin needs in order to have only twice the capacity of its closest PoW competitor. However a soft cap of 4Mb might be a good safety measure for the initial ramp up schedule. Whether doubling should occur every two or every four years and what an ultimate cap (if any) should be is all debatable.

Of course, its not better to do nothing. But is bad better than nothing? Is not it better to discuss more to find a better solution than to implement a bad one? 1 MB now is not a problem now except when blockchain is spammed by so-called "stress test". IMHO its better to wait a little than to implement BIP 101. No offence.

A few more thoughts on the matter...

Firstly, the stable and tested core has been proven to increase its soft limit from 250Kb, to 500Kb, to then 750Kb in order to survive this long by being a useful transactional medium with a single ledger. Artificially capping it to 1Mb now cannot be claimed as a safe solution going forward, because we are entering an uncharted territory here. Again a proven and workable solution that we currently have and value is known to have been increasing the cap, not maintaining it.

Secondly, the market for fees will emerge naturally as the network can timestamp a limited amount of transactions not because of an artificial cap, but due to physical limitations related to block propagation times and network stability. Whether we need an explicit limit baked into the protocol or allow the network to find its own equilibrium (with a help of outside competition in altcoin space) is a question open to debate. A series of planned workshops should help find a common ground with regards to this.

Thirdly, the idea of every monetary transaction on a single ledger might sound Utopian with present day technology and limitations, however, it is not entirely out of question. I, personally, don't think that every coffee purchase will be on blockchain within next 5 to 10 years, but it doesn't mean that we have to artificially cap the blockchain beyond what current technology allows. What transactions are in and what out will be decided by the market.