Feathercoin 2019 the plan?

Janar

And how can you put in the top 5, what can people do, what are your solutions?

AcidD

@kelsey its great to see you around here again.

kelsey

Hey AcidD, cool you’re still around :)

kelsey

@janar said in Feathercoin 2019 the plan?:

And how can you put in the top 5, what can people do, what are your solutions?

Well I think once the dust settles many of the 1001 scam projects will die, opportunity then for some of the original coins that still remain strong to rise up the rankings.

AcidD

Keeping on topic.

Feathercoin needs to get used… in order for it to get used it needs to be accepted where people will use/spend it.

before we can do any of that I think the team wants to use the cryptowinter to iron issues with the difficulty adjustment algo. EHRC is not really the way forward I dont think…there are better algos out there right now and currently the team are testing some of them.

once our blocktime is at a stable 1 min regardless of wild hashrate fluctuations, I think we’ll have a better chance of getting accepted as a form of payment.

j_scheibel

@acidd said in Feathercoin 2019 the plan?:

HRC is not really the w

I mentioned on another thread an unrealistic dream i had of dealing with hashrate. its worth mentioning if you really do want something possibly better than ehrc. its unrealistic though as i cant imagine how you implement it.

instead of using a fluctuating difficulty. just accept the single the best submission any mining address gives you inside of the 1 minute window (maybe make 50 second window so there is time resolve conflicts over the network.) the problem is of course resolving what could be billions of terrible possibly deliberately bad submissions. i could stop that a little by only allowing 1 submission (signed) per wallet. this at least helps the nodes quick throw out noise.(first signed submission for an address is the only one accepted) as to dealing with who wins, i figure there must be a similar mechanism already in place. though it probably hasnt ever had to deal with this much traffic

with this, difficulty goes away and is something you can see from looking at the winning solutions in the past but has no baring on anything short of a point of interest.

j_scheibel

@j_scheibel said in Feathercoin 2019 the plan?:

at the winning solutions in the past bu

perhaps minor improvement to that might be to broadcast the current best difficulty network wide after 15 seconds then 30 and then 45. again just to cut down submissions and/or help with filtering.

AcidD

@Wellenreiter @Bushstar - what do you guys think of the above idea ?

Wellenreiter

Well, we discussed the ‘accept any block coming in within the one minute ± x seconds’
I see severe problems here.

as the different nodes receive the blocks at different times due to the forwarding or may be having been offline, we need to use the timestamp contained in the block for the consensus and acceptance check.
This timestamp can be manipulated easily.-> NoGo!!!
We were discussing not a broadcast, but more general some ‘out of block chain’ mechanism to announce the difficulty’. I was thinking about changing from a pure POW coin to a POW/POS coin, but not with alternating POW/POS blocks as Trezarcoin does, but a more or less independent POS chain, where a block is generated every 2nd minute and announces the POW difficulty in addition to the POS parameters. The POS block also must be used as the previous block for the next POW block so the chains must be linked.
You see there are diverging requirements, that need to be addressed.
Whatever we choose to implement a ‘out of BC’ announcement of the POW difficulty requires a massive change of the code and will take a year or more to be coded, tested and implemented.

Implementing a totally independent announcement in a decentralized environment is a challenge, where I don’t have a solution, unless we would implement a ‘master node’ that would be a centralized approach, which is against the philosophy of a decentralized coin.
Even, if we would find a way to implement an ‘announcement channel’, there are big problems to solve.

an IRC like channel requires central servers
how to decide on the difficulty, if several nodes announce a difficulty at the same time or within the defined announcement window?
how can we sure, that all nodes agree on the announced difficulty? There is a high risk of a forged block chain here.

Regaring EHRC:
The diff algo is not as bad as people think and tell.
Any diff calculation algorithm mitigates the influence of changing hashing power on the time to block. Also depending on the coin’s base hash rate, long term changes in base hash rate and type and severity of attacks/hash rate changes require different settings to keep the time to block as close to the desired one as possible.
the current EHRC settings were designed at a time where the available hash rate was less than 10% of today’s hash rate and the current baseline for FTC of ~ 4 Gh/sec was seen during attacks only, while the base line was around 1 Gh/sec.

I’d say that the parameters need adjustment and have found a promising set of parameters.

AmDD

Id give my vote to the POW/POS idea. I know thats a big change and a lot of work but I think it gives us a number of advantages.
It solves the difficulty issue as stated above.
adds a layer of security as a 51% now requires more than just a lot of hashrate
it gives people a second mining option for income
in order to stake you need to run a node, more nodes helps distribute the network

Wellenreiter

Even when implemenrting a POW/POS solution as described, we still need a POW algorithm. There are two problems that together cause long gaps between blocks.

the difficulty is high, but the hash rate is low, due to a large pool (e.g. nicehash) withdraws it’s hash power.
currently there is no way to announce a new and lower difficulty before the next block is found, what takes a long time due to 1)

POW/POS mixture as described mitigates 2) while a good pow algorithm hopefully mitigates 1)

The short or very short time to block caused by the increase of hash rate has no problem like 2) and a much smaller probelm like 1).
as the blocks are found very fast the pow algorithm can adjust and announce new difficulties very fast.

What also needs to be remembered, is, that a time frame with long or very long time 2 block is required leverage the overall time 2 block. Without time 2 block rates > 1 minute the long average of all blocks generated would get smaller and smaller for every peak in hash rate causing an overproduction of coins.