Quote from: Chris_Sabian on September 15, 2015, 10:28:09 PM

Quote from: adamstgBit on September 15, 2015, 10:09:55 PM

Quote from: brg444 on September 15, 2015, 09:37:16 PM

Quote from: adamstgBit on September 15, 2015, 08:56:51 PM

Quote from: Holliday on September 15, 2015, 08:30:05 PM

Quote from: VeritasSapere on September 15, 2015, 04:59:27 PM

Well I have an average connection for Europe and i could easily support two gigabyte blocks from home, eight megabyte blocks would not be a problem at all. This person in Florida needs to either get a new internet provider or update his client by the sounds of it. lol

Did you read my posts in this thread? I've had to drastically reduce the connectivity of my full node (latest Core release) running on it's own dedicated hardware (modern quad-core CPU, 16GB RAM, SSD) in order to keep the home network functional for other daily use demands.

My node will happily eat as much upload speed as I give it and I have top 10% home internet speeds (probably better). It can bring simple web surfing to a standstill if I let it.

Nodes don't just accept blocks. Most of my bandwidth use is on the upload side (sharing data with other peers)! Larger blocks will obviously have a direct impact on the amount of data shared.

Do you run a node or are you just guessing?

should bitcoins traffic limits be based on what 10% of typical home connection can handle?

maybe 20% ? maybe 50%? surly anything demanding 20-10% isn't going to impact number of full nodes.

in anycase i feel that there can be much improvement as to how data is shared across the network to dramatically decrease bandwidth use for full nodes, which should allow for proportionally bigger blocks.

I know someone that turns on his full node like once a week, just to download the lastest blocks and then turns it off again, so that when he makes a TX it doesn't take him long to sync up first. is this node useful? should ALL users be asked to run a full node?

Bitcoin traffic should be limited based on capacity of anonymous bandwidth growth.

Quote from: brg444 on September 15, 2015, 06:23:45 PM

Assuming yearly 20% increase blockchain size & 10% reduction in bandwidth costs, after 15-20 years no new nodes can enter system except maybe huge datacenter operations. https://www.youtube.com/watch?v=TgjrS-BPWDQ&feature=youtu.be&t=7331

what if new node run with pruned versions of the blockchain?

maybe we could have a sort of yearly checkpoint, we no longer ask nodes to download data that is >1year old and base verification of off these yearly checkpoints.

Doesn't that violate the whole point of being a node? Having a complete copy of the blockchain? Not saying that is bad but just bringing up a point.

It is absolutely bad. A checkpoint implies someone somewhere is holding the full history. Who would that be?