Quote from: brg444 on September 16, 2015, 06:42:09 PM

Quote from: adamstgBit on September 16, 2015, 06:27:38 PM

Quote from: Holliday on September 16, 2015, 06:19:33 PM

Quote from: adamstgBit on September 16, 2015, 04:41:36 PM

bottom line, we need improvements, But still the answer is No you shouldn't aim for running full nodes while running a full household worth of devices. I believe we should keep the limit such that a typical home connection can handle it with about 80% of the its bandwidth being utilized. this will be the upper limit which should never be crossed. and at the same time offer simi-full nodes that do something useful with only 10-20% of a typical home connection.

some improved incentives will also help, what's the point of running a "super node" that relay GB's of data per day? maybe miners would be willing to pay a small fee to connect to this node so that it gets blocks faster? or something!

Again, my connection and my hardware are no where close to "typical". You are, two days later, still syncing from 1 year behind? I sync the entire chain in 14 hours. My other household uses (frequent web browsing, occasional gaming and streaming video) take nowhere near the resources of Core.

If a Bitcoin enthusiast like myself, who spends money on hardware specifically to run a node, who has top tier internet speeds, can't run a full node at full capacity (without a dedicated line), I would say that the decentralization of the network has been harmed. (Of course, I am currently running multiple full nodes, but not at full capacity. My concern is increasing the data a node should share before making the system as efficient as possible.)

If you want to take off the training wheels, let's talk about it after we've got the pedals, chain, handlebars, and shifters in fantastic working order.

You envision a network where there are multiple implementations of the software based on the resources at hand. This isn't necessarily a bad thing, unless it raises the bar so much that the end user can no longer access the full censorship-proof nature of Bitcoin.

I envision a network that is able to function in a possible future where governments force ISPs to audit or even censor Bitcoin traffic. I don't think we will have truly censorship-proof money until the Bitcoin network is not only able, but actually functioning, on a global wireless mesh network that is entirely out of the reach of any entity that would wish to control it.

I want Bitcoin to function in a worst case scenario. This isn't a war to see who can provide the cheapest, most convenient transaction for buying a glass of wine, this is a war for financial freedom.

its only taking days to sync because i need to work all day on my computer and i cant have core eating 90% cpu while i work.
also i turn off my machine at night because it's hot in my room.

i think your exaggerating the resources core takes and that's why i'm sycning a full node, once the full node is synced i will see first hand if this hurts my streaming.

agreed we need improvements done before we up the limit.

i think your pushing the "war" thing a little to far, right now mostly all governments are OK with bitcoin.

How can you be so naive?

By exaggerating do you suggest Holliday is lying? What he's saying sounds plausible to me and certainly not anecdotal, there have been many reports of node owners experiencing similar problems.

Planning for "the war thing" is how you build security systems: assuming worst-case scenarios. Today's reality might be quite different from tomorrow's. Do you really expect governments of the world to let Bitcoin take over the world without putting up a fight?

honestly I expect them to embrace bitcoin as the banksters monopoly money fails them.

and i think Holliday is simply unlucky and is someone bombarded with requests from alot of node needlessly
again proving that the network is setup really ineffectively

look at what this guy is saying on the matter.

Quote from: uxgpf on September 16, 2015, 06:04:39 PM

I'm running one. Currently 31 connections, 5.5 kB/s down 9.2 kB/s up.

[edit] Killed the node while fiddling with it...so now only 15 connections. Anyway, here's some more numbers.

Code:

Every 10.0s: bitcoin-cli getinfo Wed Sep 16 21:21:33 2015

{
   "version" : 110000,
   "protocolversion" : 70010,
   "blocks" : 374827,
   "timeoffset" : -1,
   "connections" : 15,
   "proxy" : "",
   "difficulty" : 56957648455.01000977,
   "testnet" : false,
   "relayfee" : 0.00001000,
   "errors" : ""
}

5 snapshots from download and upload bandwidth:

Code:

$ for i in {1..5}; do awk '{if(l1){print ($2-l1)/1024"kB/s",($10-l2)/1024"kB/s"} else{l1=$2; l2=$10;}}' <(grep wlan0 /proc/net/dev) <(sleep 1; grep wlan0 /proc/net/dev); sleep 1; done
2.47266kB/s 9.90723kB/s
5.45605kB/s 10.3809kB/s
0.604492kB/s 0.650391kB/s
7.7959kB/s 8.41016kB/s
3.51953kB/s 10.4092kB/s

how can they get such drastically different results