Announcing the Thunder Network Alpha Release

[Lauda]

Wow this sounds amazing, i paticullarly liked the "better than visa scale" comment.  more than a 100000 TPS  

This seems like a bit 'overkill'. Bitcoin is far away from needing 100 000 TPS. I'd be very satisfied if we had a capacity of 1000 TPS at the moment as there would never be some kind of backlog due to inadequate amount of space (unless there was some kind of attack) for quite some time.

I believe this could potentially open doors to new avenues of attack that wouldn't be possible on the main blockchain.

There might be attack vectors for the off-chain system, but that is rather unavoidable. However, I'm certain that those will be discovered and prevented on time (or at least most of them).

[110110101]

Wow this sounds amazing, i paticullarly liked the "better than visa scale" comment.  more than a 100000 TPS  

This seems like a bit 'overkill'. Bitcoin is far away from needing 100 000 TPS. I'd be very satisfied if we had a capacity of 1000 TPS at the moment as there would never be some kind of backlog due to inadequate amount of space (unless there was some kind of attack) for quite some time.

I believe this could potentially open doors to new avenues of attack that wouldn't be possible on the main blockchain.

There might be attack vectors for the off-chain system, but that is rather unavoidable. However, I'm certain that those will be discovered and prevented on time (or at least most of them).

I'll obviously have to read more and maybe try this out soon, but it was my understanding that the speed of 100 000 TPS was more based on available nodes? With fewer nodes, fewer transactions? If that is the case then this technology seems pretty scalable to whatever is required by the network.

[franky1]

for me i just see LN and TN as a offchain solution like coinbase/bitstamp.

you have to send funds to a special address (just like depositing funds to a service)

then you have to log into a client (just like logging into a service) and then move funds within that system(service)..

the stupid things is that not everyone is going to be sending funds to the same destination when they do transactions within LN/TN and so the settling of funds WONT be aggregating from 100,000 inputs with 100,000 outputs.. into 1input 1output

instead these settling transactions will be HUGE, yet still limited to the onchain capacity of todays ~2500, near future segwits ~3800 or next years roadmap of ~7600 input/outputs per 10 minutes

infact many people are suggesting that the ONCHAIN blocks will start to show just ONE large transaction from a LN/TN that will cause issues for people wanting to use onchain transactions. which will also cause issues because the headache of getting the LN lock transaction accepted will mean people will end up forced to lock in more funds on a more long term bases into LN/TN for the chance of actually transacting, rather than for people to voluntarily just use LN/TN now and again, due to the fact that the onchain blocks are always crammed with a LN/TN transaction bloating the block.

imagine it instead of 2500 transactions from 2500 different people.. there will be 1 transaction with 2500 input/outputs that have been aggregated from maybe 3500 different people. some may say thats a benefit.. but if every block is filled with this single transaction.. then there wont be much left for people to use onchain transactions.

knowing that the average person makes ~42 payments a month (statistics of both bitcoin and visa) a system such as LN/TN is not the ultimate solution for scaling purely because of the inability to aggregate lots of peoples different destinations into just a few.

the 2 times LN/TN would be useful. is:
for something like coinbase where people move funds between deposit addresses but the ultimate settlement transaction is coinbases single coldstore address.

gamblers who send funds to the same 'bet' address multiple times a day, as it would appear as just one input/output(per user) at settlement as oppose to multiple input output per day onchain multiplied by users

in my eyes LN/TN will only be useful where there is an ultimate static destination that receives alot off volume per day or week from the same collection of users, rather than a system anyone can use to buy anything from anywhere really quickly

[davinchi]

I don't really get the overall idea here, but what caught my attention are the scale and extremely cheap payments. 100k tx/s with only a few thousand nodes is a really great achievement. And with cheap payments,this could entice more people into the bitcoin world.

[Lauda]

I'll obviously have to read more and maybe try this out soon, but it was my understanding that the speed of 100 000 TPS was more based on available nodes? With fewer nodes, fewer transactions? If that is the case then this technology seems pretty scalable to whatever is required by the network.

Correct, just read this part carefully:

Scale: According to our tests so far, we can achieve better-than-Visa scale (100,000 TPS) with only a few thousand nodes on the network

This is why we need exact numbers. 'A few thousand nodes' could be a lot of numbers. I'm interested in how many nodes are required to achieve 1000 TPS.

I don't really get the overall idea here, but what caught my attention are the scale and extremely cheap payments.

Start with reading the Lightning Network wallpaper and finding ELI5 explanations.

And with cheap payments,this could entice more people into the bitcoin world.

Bitcoin is already cheap; I don't see why this would make a difference as far as that is concerned.

[RoseMann]

Seeing is believing my friend ---> " Scale: According to our tests so far, we can achieve better-than-Visa scale (100,000 TPS) with only a few thousand nodes on the network " .... If that is possible, a

company like VISA can just buy your company and double their capacity and kill Bitcoin. Where are these nodes or were they tested with virtual machines? I hope they give us a live demonstration

on how this will work, and some proof of these statements.  

Visa can buy the blockchain.info, or the thunder network, but there will other providers appear. It cannot buy them all.