elective-stereophonic
elective-stereophonic
[Ardor] It's coming... singapore
Please login or register.

Login with username, password and session length
Advanced search  

News:

Latest Nxt Client: Nxt 1.11.15

Pages: 1 2 3 [4]  All

Author Topic: [Ardor] It's coming...  (Read 23355 times)

gh

  • Full Member
  • ***
  • Karma: +19/-2
  • Offline Offline
  • Posts: 202
    • View Profile
    • NXT Chart
Re: [Ardor] It's coming...
« Reply #60 on: January 22, 2017, 07:07:36 pm »

.
« Last Edit: February 05, 2017, 05:34:54 pm by gh »
Logged

blackyblack1

  • Hero Member
  • *****
  • Karma: +165/-82
  • Offline Offline
  • Posts: 1764
    • View Profile
Re: [Ardor] It's coming...
« Reply #61 on: January 22, 2017, 08:02:23 pm »

Does this mean we won't be getting low fees on child chains?
Fees can be zero if there is a bundler willing to sponsor the childchain. Fees for the bundler on Ardor will be minimum 1 ARDR, if I have understood it correctly.
Another option when fees can be zero is when forger is a bundler. Forger should decline all external bundlers and bundle all childchain transactions when he is ready to forge a block.
Logged

gh

  • Full Member
  • ***
  • Karma: +19/-2
  • Offline Offline
  • Posts: 202
    • View Profile
    • NXT Chart
Re: [Ardor] It's coming...
« Reply #62 on: January 23, 2017, 06:58:26 am »

.
« Last Edit: February 05, 2017, 12:09:27 pm by gh »
Logged

blackyblack1

  • Hero Member
  • *****
  • Karma: +165/-82
  • Offline Offline
  • Posts: 1764
    • View Profile
Re: [Ardor] It's coming...
« Reply #63 on: January 23, 2017, 08:18:25 am »

Does this mean we won't be getting low fees on child chains?
Fees can be zero if there is a bundler willing to sponsor the childchain. Fees for the bundler on Ardor will be minimum 1 ARDR, if I have understood it correctly.
Another option when fees can be zero is when forger is a bundler. Forger should decline all external bundlers and bundle all childchain transactions when he is ready to forge a block.
You probably have to use the API to achieve that (forging your own bundler to avoid fees) and I don't think it will be easy to achieve. On January 12 ScripterRon wrote the following (in another thread): There is also a bundle API.  You can check if you are due to forge the next block and then use the API to manually bundle transactions instead of running an automatic bundler.  But the forger also run automatically, so you can't guarantee that your transaction will be selected unless you overpay the fee, even though the forger is running on your local node. Of course, your bundler can set its rate to undercut the competition, which would result in users using your rate when submitting child transactions.  The other bundlers would not process the transactions unless they also lower their rates to compete with you.
I think the only good way to achieve this is by changing the code of the Ardor forger itself. It should not break consensus but I did not get a confirmation that such a change will be accepted in the main Ardor repository. So the forgers willing to get the best profitability will be forced to use 3rd party code not signed by the main dev team.
Logged

ScripterRon

  • Hero Member
  • *****
  • Karma: +75/-2
  • Offline Offline
  • Posts: 523
    • View Profile
Re: [Ardor] It's coming...
« Reply #64 on: January 23, 2017, 03:13:37 pm »

So I can create an account by sending 0.0001 IGNIS and it'll cost the same as sending to an existing account? Does this mean we won't be getting low fees on child chains?
The fee for a child chain transaction depends on the bundler rate.  For example, say a transaction costs 1 ARDR.  If the IGNIS bundler rate is 1.0, then processing the transaction on the IGNIS chain will require a fee of 1 IGNIS.  However, if the bundler rate is 0.1, then processing the same transaction on the IGNIS chain will will require a fee of 0.1 IGNIS.  It is totally up to the account running the bundler as to its rate.

There can be multiple bundlers for the same chain and they can have different rates.  The server will tell you the best (lowest) rate but you can specify any rate you want when submitting a transaction.  All bundlers with the same or lower rate will process the transaction and include it in the child block transaction (a child block transaction contains one or more child transactions).  Since multiple bundlers can process a transaction, there is an additional bundler parameter available.  The account running a bundler can say that it will pay more than the minimum to get its bundled transaction accepted in a block.  For example, the bundler could specify it will pay an additional 0.01 ARDR for each ARDR in the child block transaction fee.  So, if the total for the bundled transactions is 2.0 ARDR, the bundler would pay 2.02 ARDR for the child block transaction.  The forger would then select the transaction paying 2.02 ARDR over the transaction paying 2.00 ARDR when creating a block (there is a limit of one child blocktransaction per chain per block).
Logged

ScripterRon

  • Hero Member
  • *****
  • Karma: +75/-2
  • Offline Offline
  • Posts: 523
    • View Profile
Re: [Ardor] It's coming...
« Reply #65 on: January 23, 2017, 03:19:13 pm »

I think the only good way to achieve this is by changing the code of the Ardor forger itself. It should not break consensus but I did not get a confirmation that such a change will be accepted in the main Ardor repository. So the forgers willing to get the best profitability will be forced to use 3rd party code not signed by the main dev team.
Even if you forge a block, that does not mean that your block will end up in the block chain.  If a better block is created by another forger, that block will replace your block.  So you end up playing the probabilities.  If you are running a Nxt 1.x node, take a look at nxt.log.  Check the number of rollbacks that occur.  This indicates a better block was found and it replaced the current block(s).
Logged

blackyblack1

  • Hero Member
  • *****
  • Karma: +165/-82
  • Offline Offline
  • Posts: 1764
    • View Profile
Re: [Ardor] It's coming...
« Reply #66 on: January 23, 2017, 04:26:20 pm »

I think the only good way to achieve this is by changing the code of the Ardor forger itself. It should not break consensus but I did not get a confirmation that such a change will be accepted in the main Ardor repository. So the forgers willing to get the best profitability will be forced to use 3rd party code not signed by the main dev team.
Even if you forge a block, that does not mean that your block will end up in the block chain.  If a better block is created by another forger, that block will replace your block.  So you end up playing the probabilities.  If you are running a Nxt 1.x node, take a look at nxt.log.  Check the number of rollbacks that occur.  This indicates a better block was found and it replaced the current block(s).
You are wrong. In a proposed solution the forger does not risk anything. If his block is rolled back he does not get the childchain fees and does not loose his Ardor coins. So basically the result is the same as skipping his turn.
If bundler and forger are running separately the risk of loosing Ardor fees occurs.
Logged

Brangdon

  • Hero Member
  • *****
  • Karma: +229/-25
  • Offline Offline
  • Posts: 1389
  • Quality is addictive.
    • View Profile
Re: [Ardor] It's coming...
« Reply #67 on: January 23, 2017, 08:13:40 pm »

So I can create an account by sending 0.0001 IGNIS and it'll cost the same as sending to an existing account? Does this mean we won't be getting low fees on child chains?
The fee for a child chain transaction depends on the bundler rate.
Yes, of course. I'm interested in the minimum ARDOR fees, and the trade-off between spam deterrence and encouraging usage. Since the cost to the network of a transaction depends on whether or not it creates a new account, shouldn't the minimum ARDOR fees be higher in that case?
Logged

CryptKeeper

  • Hero Member
  • *****
  • Karma: +78/-5
  • Offline Offline
  • Posts: 1235
    • View Profile
Re: [Ardor] It's coming...
« Reply #68 on: January 24, 2017, 07:52:30 am »

So I can create an account by sending 0.0001 IGNIS and it'll cost the same as sending to an existing account? Does this mean we won't be getting low fees on child chains?
The fee for a child chain transaction depends on the bundler rate.
Yes, of course. I'm interested in the minimum ARDOR fees, and the trade-off between spam deterrence and encouraging usage. Since the cost to the network of a transaction depends on whether or not it creates a new account, shouldn't the minimum ARDOR fees be higher in that case?

Regarding this, I've posted a proposal about a feeless child chain a while ago:

https://nxtforum.org/general/a-hybrid-powpos-approach-for-feeless-transactions-on-ardor-child-chains/msg224054/#msg224054

tl;dr
Proof-of-Work (POW) could be integrated into an Ardor child chain wallet, preventing blockchain spam in a feeless system.
Logged
Follow me on twitter for the latest news on bitcoin and altcoins!
Vanity Accounts Sale :-)

Jean-Luc

  • Core Dev
  • Hero Member
  • *****
  • Karma: +816/-81
  • Offline Offline
  • Posts: 1610
    • View Profile
Re: [Ardor] It's coming...
« Reply #69 on: January 24, 2017, 11:20:29 pm »

I think the only good way to achieve this is by changing the code of the Ardor forger itself. It should not break consensus but I did not get a confirmation that such a change will be accepted in the main Ardor repository. So the forgers willing to get the best profitability will be forced to use 3rd party code not signed by the main dev team.
Even if you forge a block, that does not mean that your block will end up in the block chain.  If a better block is created by another forger, that block will replace your block.  So you end up playing the probabilities.  If you are running a Nxt 1.x node, take a look at nxt.log.  Check the number of rollbacks that occur.  This indicates a better block was found and it replaced the current block(s).
You are wrong. In a proposed solution the forger does not risk anything. If his block is rolled back he does not get the childchain fees and does not loose his Ardor coins. So basically the result is the same as skipping his turn.
If bundler and forger are running separately the risk of loosing Ardor fees occurs.

Ardor fees from child block transactions are split back in 1:1:1:1 ratio with the previous 3 block forgers.
Logged
GPG key fingerprint: 263A 9EB0 29CF C77A 3D06  FD13 811D 6940 E1E4 240C
NXT-X4LF-9A4G-WN9Z-2R322

blackyblack1

  • Hero Member
  • *****
  • Karma: +165/-82
  • Offline Offline
  • Posts: 1764
    • View Profile
Re: [Ardor] It's coming...
« Reply #70 on: January 25, 2017, 09:05:47 am »

I think the only good way to achieve this is by changing the code of the Ardor forger itself. It should not break consensus but I did not get a confirmation that such a change will be accepted in the main Ardor repository. So the forgers willing to get the best profitability will be forced to use 3rd party code not signed by the main dev team.
Even if you forge a block, that does not mean that your block will end up in the block chain.  If a better block is created by another forger, that block will replace your block.  So you end up playing the probabilities.  If you are running a Nxt 1.x node, take a look at nxt.log.  Check the number of rollbacks that occur.  This indicates a better block was found and it replaced the current block(s).
You are wrong. In a proposed solution the forger does not risk anything. If his block is rolled back he does not get the childchain fees and does not loose his Ardor coins. So basically the result is the same as skipping his turn.
If bundler and forger are running separately the risk of loosing Ardor fees occurs.

Ardor fees from child block transactions are split back in 1:1:1:1 ratio with the previous 3 block forgers.
Wow. Now that was really unexpected.
Logged

lurker10

  • Hero Member
  • *****
  • Karma: +168/-33
  • Offline Offline
  • Posts: 1334
    • View Profile
Re: [Ardor] It's coming...
« Reply #71 on: January 25, 2017, 10:50:28 am »

Bitcoin can handle 3 tps, Ethereum 15 tps. I am quoting someone re Ethereum, didn't verify this.

Realistically how many tps can Ardor handle on all child chains? Some ballpark, order of magnitude estimated figure could do.
Logged
Run a node - win a prize! "Lucky node" project jar: NXT-8F28-EDVE-LPPX-HY4E7

Jean-Luc

  • Core Dev
  • Hero Member
  • *****
  • Karma: +816/-81
  • Offline Offline
  • Posts: 1610
    • View Profile
Re: [Ardor] It's coming...
« Reply #72 on: January 25, 2017, 07:31:20 pm »

Since the cost to the network of a transaction depends on whether or not it creates a new account, shouldn't the minimum ARDOR fees be higher in that case?
I have thought about that but haven't implemented it yet, maybe.
Logged
GPG key fingerprint: 263A 9EB0 29CF C77A 3D06  FD13 811D 6940 E1E4 240C
NXT-X4LF-9A4G-WN9Z-2R322

Sebastien256

  • Hero Member
  • *****
  • Karma: +169/-24
  • Offline Offline
  • Posts: 2823
  • ^LOOK UP^ = Nxt community!
    • View Profile
Re: [Ardor] It's coming...
« Reply #73 on: January 26, 2017, 03:53:55 am »

Since the cost to the network of a transaction depends on whether or not it creates a new account, shouldn't the minimum ARDOR fees be higher in that case?
I have thought about that but haven't implemented it yet, maybe.

I think now is the time to do it, before launch. resistance is to come if fee are increased later. Just an opinion.
Logged
Please drop your ideas concerning Nxt and/or NRS in this topic -> List of feature request for Nxt and/or NRS (with the full list in OP).

qq2536007339

  • Hero Member
  • *****
  • Karma: +42/-9
  • Offline Offline
  • Posts: 514
    • View Profile
Re: [Ardor] It's coming...
« Reply #74 on: January 26, 2017, 03:55:00 am »

Yes, of course. I'm interested in the minimum ARDOR fees, and the trade-off between spam deterrence and encouraging usage. Since the cost to the network of a transaction depends on whether or not it creates a new account, shouldn't the minimum ARDOR fees be higher in that case?

I'm also interested in the minimum ARDOR fees,I think should set less than 0.05 ARODR,otherwise won't take the advantage of punch.
Logged
你送我阿朵,我是要的。ARDOR-DJ68-PG7W-4JEU-2LU5T

blackyblack1

  • Hero Member
  • *****
  • Karma: +165/-82
  • Offline Offline
  • Posts: 1764
    • View Profile
Re: [Ardor] It's coming...
« Reply #75 on: January 26, 2017, 10:30:19 am »

I think the only good way to achieve this is by changing the code of the Ardor forger itself. It should not break consensus but I did not get a confirmation that such a change will be accepted in the main Ardor repository. So the forgers willing to get the best profitability will be forced to use 3rd party code not signed by the main dev team.
Even if you forge a block, that does not mean that your block will end up in the block chain.  If a better block is created by another forger, that block will replace your block.  So you end up playing the probabilities.  If you are running a Nxt 1.x node, take a look at nxt.log.  Check the number of rollbacks that occur.  This indicates a better block was found and it replaced the current block(s).
You are wrong. In a proposed solution the forger does not risk anything. If his block is rolled back he does not get the childchain fees and does not loose his Ardor coins. So basically the result is the same as skipping his turn.
If bundler and forger are running separately the risk of loosing Ardor fees occurs.

Ardor fees from child block transactions are split back in 1:1:1:1 ratio with the previous 3 block forgers.
Wow. Now that was really unexpected.
I have made some calculations. The modified forger (forger + bundler) is 25% in the worst case more profitable than the regular bundler. Eg regular bundler spends 1 ARDR to get 10 IGNIS fee and modified forger spends 1 ARDR and get 0.25 ARDR back as a forging reward. But this is a worst case when childchain fees are close to 1 ARDR in value.
Now there is probability that childchain fees are much higher than 1 ARDR eg 100 IGNIS fees. Regular bundlers would bid until somewhere of 10 ARDR fee for forger. But modified forger would pay the same 1 ARDR fee and get the 100 IGNIS fee to himself for just 0.75 ARDR penalty.
Someone can notice that the regular forger is even more profitable since it would simply take 10 ARDR from bundlers to himself. But remember that the forging fee is divided 1:1:1:1 ratio so he would only get 2.5 ARDR to himself.

TLDR: modified forger can get 9.25 ARDR fee instead of 2.5 ARDR fee if the fees in the block are high enough.
Logged

blackyblack1

  • Hero Member
  • *****
  • Karma: +165/-82
  • Offline Offline
  • Posts: 1764
    • View Profile
Re: [Ardor] It's coming...
« Reply #76 on: January 26, 2017, 01:39:12 pm »

And one more thing. With a constant rate of transactions modified forger bears almost no risk of loosing his Ardor. So still the opportunity of (almost) zero fees transactions exists.
Logged

rubenbc

  • Hero Member
  • *****
  • Karma: +187/-3
  • Offline Offline
  • Posts: 1304
  • I'm here and you?
    • View Profile
Re: [Ardor] It's coming...
« Reply #77 on: January 26, 2017, 02:02:15 pm »

Thanks for share your minds @blacky, your description has helped me a lot

Enviado desde mi MI PAD 2 mediante Tapatalk

Jose

  • Global Moderator
  • Sr. Member
  • *****
  • Karma: +77/-1
  • Offline Offline
  • Posts: 363
    • View Profile
Re: [Ardor] It's coming...
« Reply #78 on: January 26, 2017, 07:51:00 pm »

...(a child block transaction contains one or more child transactions)...

Can a bundler choose how many child transactions are included in a single child block transaction?
Could a bundler set the minimum number of child transactions to be included in a child block transaction and not generate/process that child block transaction until that queue is filled up?
Logged
Pages: 1 2 3 [4]  All
 

elective-stereophonic
elective-stereophonic
assembly
assembly