111* days estimated for Difficulty to retarget. Could this kill PTS?

[barwizi]

new KGW called DGW
coder please have a look here:

https://bitcointalk.org/index.php?topic=421615.8760

chaeplin wrote

"
Yay name is "DarkGravityWave"

https://github.com/evan82/darkcoin/commit/50189ca2a010728bc559a4f46568267bf13b7ff7
"

Its not tested how can you trust it?

o.k. then forget KGW or DGW

only a suggestion:
implement a short time-retarget after 16 or 32 blocks with maximum swing of +- 50 %
regarding the difficult-average of the last 128 blocks

25% swing makes for smoother transitions.  but i think this is a better idea .

[Schwede65]

new KGW called DGW
coder please have a look here:

https://bitcointalk.org/index.php?topic=421615.8760

chaeplin wrote

"
Yay name is "DarkGravityWave"

https://github.com/evan82/darkcoin/commit/50189ca2a010728bc559a4f46568267bf13b7ff7
"

Its not tested how can you trust it?

o.k. then forget KGW or DGW

only a suggestion:
implement a short time-retarget after 16 or 32 blocks with maximum swing of +- 50 %
regarding the difficult-average of the last 128 blocks

[jsidhu]

Ok i will switch it

[bitbadger]

new KGW called DGW
coder please have a look here:

https://bitcointalk.org/index.php?topic=421615.8760

chaeplin wrote

"
Yay name is "DarkGravityWave"

https://github.com/evan82/darkcoin/commit/50189ca2a010728bc559a4f46568267bf13b7ff7
"

Its not tested how can you trust it?

This is true.  However, we know that standard KGW is tested and failed.  This exploit has been used on several coins including DarkCoin and Version.

[jsidhu]

new KGW called DGW
coder please have a look here:

https://bitcointalk.org/index.php?topic=421615.8760

chaeplin wrote

"
Yay name is "DarkGravityWave"

https://github.com/evan82/darkcoin/commit/50189ca2a010728bc559a4f46568267bf13b7ff7
"

Its not tested how can you trust it?

[Schwede65]

new KGW called DGW
coder please have a look here:

https://bitcointalk.org/index.php?topic=421615.8760

chaeplin wrote

"
Yay name is "DarkGravityWave"

https://github.com/evan82/darkcoin/commit/50189ca2a010728bc559a4f46568267bf13b7ff7
"

[testz]

Its just interim so KGW should suffice it does what we need it to do... Whats the verdict do I go test now or wait if we arent even gonna do it?

I downloaded client from the ANN thread and it doesnt sync so you may want to updte that link if said clien is old and 1.0.0 has new seeds. Even the add nodes dont work.

No, for now we will keep all as is. But if you finished KGW implementation why not take few additional hours and test it.  :)
Please post a link to "ANN thread", today the official one is: http://invictus.io/bitshares-pts.php

[bitbadger]

Implementing KGW or changing the retarget interval will require code to be changed, and everybody to download new clients.  Plus the stigma of a "hard fork".

A matching bounty on each block mined, up until the difficulty reset, would require just a bit of accounting work.  (And ~3500 PTS.)  The bounties could even be paid out over time, say every 100 blocks or something.  Pools would have to make some changes, but they would just have to keep track of amounts credited to miners, and double them when they receive the bounty payment.

Of course, that doesn't solve the problem long-term; the next DAC snapshot announcement will undoubtedly bring in a lot of new mining power, and after the snapshot, price will drop and the miners will go away again, and the loyal miners will be stuck with the high difficulty again.

[testz]

BitShares-PTS it's an investment cryptocurrency and peoples except speculators doesn't care about network speed. One more fork give us more problems and repute disadvantages.

All the more reason to do something if you can.  These investors on a normal network would be getting 3x the number of shares and here you are saying, "FU, not our problem. Go ahead and waste your time and energy mining for 1/3 the payout/dividends in the future, but keep mining to keep the network limping along."

In our current case any solution requires hard fork. I can't consider current situation as a huge problem. The miners rewards doesn't connect to the network speed, with same diff they get same amount of PTS per day. The miners left the network because the exchange rate was drop.
Maybe I missed something important but I really doesn't see this as a problem.

Unfortunately KGW has some security disadvantages and probably in our case better to drop diff retarget from 4032 to 2016 blocks but again this requires hard fork.

[NewMine]

BitShares-PTS it's an investment cryptocurrency and peoples except speculators doesn't care about network speed. One more fork give us more problems and repute disadvantages.

All the more reason to do something if you can.  These investors on a normal network would be getting 3x the number of shares and here you are saying, "FU, not our problem. Go ahead and waste your time and energy mining for 1/3 the payout/dividends in the future, but keep mining to keep the network limping along."

[jsidhu]

Its just interim so KGW should suffice it does what we need it to do... Whats the verdict do I go test now or wait if we arent even gonna do it?

I downloaded client from the ANN thread and it doesnt sync so you may want to updte that link if said clien is old and 1.0.0 has new seeds. Even the add nodes dont work.

[Markus]

Wouldn't the easiest fix be to lower the retarget interval from 4032 blocks down to somewhere like 100-300 blocks?

[testz]

IMHO we don't need to do anything about this problem, transaction time is same as Bitcoin and still acceptable, we have 4 mining pools, the possibility of 51% attack is very small. Network has enough nodes, clients syncs fast.

BitShares-PTS it's an investment cryptocurrency and peoples except speculators doesn't care about network speed. One more fork give us more problems and repute disadvantages.

[testz]

Is there nodes running the testnet? Ip?

Or do I have to do a test in a box approach.. creating a new testnet?

For Ip you can look into sources for testnet dns seeds.
If you want to test your KGW at any case you need to setup your disconnected testnet network, because regular testnet will not accept your blocks.

Also give me a list of good nodes on live.. preferabble dns.. Ill update the dnsseed list so client connects without add adding nodes manually. The ann thread has some nodes but they dont connect. This is unacceptable.

You don't need add nodes manually dns seeds already updated in BotShares-PTS 1.0.0.0.

[bitbadger]

there are two economic problems with fixing this with fees alone:
1) retarget occurs every ~2000 blocks, which now is ~3500PTS in mining revenue. Current hashrate is down to <50% of what it was previously. However, offering 500 PTS will only increase payouts by ~14%. Therefore, you will need a lot more than 500 PTS to entice miners back.

Here's an idea: double the block rewards until we hit the retarget.  Invictus can do this without changing a line of code or forcing anybody to download anything.  Simply make it publicly known, and then transmit (from their pool of PTS) the value of an extra block reward to the address of every miner who finds a block.

In essence, a bounty on every found block, for the value of that block.