Discussion about the transaction fees

Nagalim · September 8, 2015, 10:28pm

I’m voting 0.02 nbt. I don’t think we should spike this too quickly because it could do irrevocable damage to our brand.

willy · September 9, 2015, 5:40am

I think the data feeds of @cryptog and @cybnate are still voting for 0.01, right?

Sentinelrv · September 9, 2015, 5:42am

Is there a way to tell how many people are following these feeds? I wonder how much impact they’re having on things like this.

willy · September 9, 2015, 5:46am

Mh. Analysing httpd logs?
It’s http traffic after all, but I’m not sure the two can analyse such things on github, which is hosting their data feed files.

tomjoad · September 9, 2015, 2:36pm

Actually, there is a way to track this. If each data feed provider inputted a unique motion hash into their voting (such as 0B5AB73448AFDF1080598ED9A083DF4EE94925F2 for “This is Cybnate’s data feed.”) we could track in real time the level of support each data feed provider has. This motion hash would be unique to an individual data feed provider. This approach would guard against voter centralization because the network would be aware when a particular data feed provider is being used by too many users on the network.

masterOfDisaster · September 9, 2015, 2:58pm

Right!

The quoted part from @mhps is important, as we were dealing with a test motion hash…

cryptog · September 10, 2015, 7:23am

Yes I am voting for 0.01 right now.

Cybnate · September 12, 2015, 10:05am

Same here

Nagalim · September 12, 2015, 1:37pm

Any reason for voting for 0.01? Why not 0.02?

masterOfDisaster · September 12, 2015, 1:42pm

…or 0.05…

Nagalim · September 12, 2015, 1:45pm

I could argue 5x is a big increase and most exchanges have set at around 0.1 nbt withdrawal fee. 0.02 would hardly be noticed and would double our fee income.

masterOfDisaster · September 12, 2015, 3:09pm

I have no detailed information about NBT are used at the moment.
Looking at the volume at exchanges I need to assume that hedging (BTC) volatility is still a major part.
Further I assume that there are not only little amounts of NBT per person used for the hedging.
For moving dozens or hundreds (or thousands?) of NBT from exchange to exchange to arbitrage it plays little role whether the fee is 0.01 NBT per kB or 0.05 or 0.1 or 1.

It would be very helpful to have an option at the [blockexplorer][1] that lists transactions in a selected time frame with options to sort it, showing the accumulated NBT fee and the average fee per transacted NBT volume.

Lacking this information I can only suppose the current user profile is moving big volumes of NBT. HEre are some examples. I just scrolled down in the blockexplorer through transactions of the last 24 hours and picked transactions:

Then there are a few transactions like this with little volume:
https://blockexplorer.nu/blocks/523423/1: 13.5787 NBT
https://blockexplorer.nu/blocks/523422/1: 47.6423 NBT
https://blockexplorer.nu/blocks/522865/1: 56.7867 NBT

Oh, and I’ve found micro transactions:
https://blockexplorer.nu/blocks/523595/1: 0.0369 NBT
https://blockexplorer.nu/blocks/523152/1: 0.04 NBT
https://blockexplorer.nu/blocks/523052/1: 0.0582 NBT
https://blockexplorer.nu/blocks/522935/1: 0.0474 NBT
https://blockexplorer.nu/blocks/522883/1: 0.1074 NBT

This list of NBT transactions in the last 24 hours isn’t complete, but close to it (except for the hourly 21/22 NBT transactions).
From what I gather from it I had no trouble suggesting a fee of 0.1 to 1 NBT per kB.
It wouldn’t hurt when moving big amounts, which seems to be the major use case (at least in the last day).
I have no idea (and forgive me that I’m too lazy to do that manually and too unskilled to automate it) how it looks in the last 7 days, last 30 days, whatever time frame.

**hint, hint **
It would be very helpful to have an option at the [blockexplorer][1] that lists transactions in a selected time frame with options to sort it, showing the accumulated NBT fee and the average fee per transacted NBT volume.

[1]: https://blockexplorer.nu

Nagalim · September 12, 2015, 3:23pm

You are using an inherent concept of volume dependence here without rigorously defining the change address. I wrote a program that can be used to find upper and lower bounds on a volume-dependent fee. I also wrote a program that finds how many fees have been burnt, but I had a hard time distinguishing fees from custodian burns.

Anyway, the answer is that we can totally get more out of our volume. However, if we go too much we will kill the initially curious and not very crypto-wealthy who are vital to marketing efforts.

masterOfDisaster · September 12, 2015, 3:26pm

I’m aware that some of the transaction might be above 1 kB - if that’s what you are pointing to.

That’s the reason why I’d like to have convenient access to

I fully agree to this:

And for that reason I chose a very wide range (0.1 to 1.0 NBT per kB) when suggesting a tx fee

Nagalim · September 12, 2015, 3:27pm

Not blockchain volume, $$ volume. You want to know how much $$ was sent but you don’t know which address was the change. You’re overestimating your numbers.

tomjoad · September 12, 2015, 3:30pm

I agree. The current method of setting transaction fees allows zero ability to price discriminate based on transaction size, and so it makes sense to keep fees reasonably low to attract a high number of new users to our network, even if it means giving up short-term profit. Although, it would be nice to see even 0.02 NBT voted for just to send a signal that we are able to change our Tx fees.

And I still think something like @woodstockmerkle’s chart of scaling variable transaction fees should be the next major developer priority once B&C Exchange development is complete.

masterOfDisaster · September 12, 2015, 3:30pm

Ah, now I get it (I think…).
I listed the volume of the UTXO that was part of the tx, but would need to know how much was sent and how much was change, right?

Nagalim · September 12, 2015, 3:35pm

For instance this guy:
https://blockexplorer.nu/blocks/523659/1
You probably called that a 20 nbt Tx, but what if it was a 0.03 nbt tx and the 20 nbt was the account balance?

@tomjoad I would love to see a volume dependent tx fee passed, no matter what function we decide on. As long as there’s a votable proportionality constant, I think we could do just about anything and it would work out. The change problem is a powerful issue, but not undefeatable.

masterOfDisaster · September 12, 2015, 3:52pm

You are right. I just tried to find some information on which tx fees could be based on. But it seems like the tx fee is in the same category as the park rate interest: no reliable data to base a decision on

Nagalim · September 12, 2015, 4:07pm

we can find an absolute minimum average Tx size and standard deviation by assuming the biggest output is the change. Maximum is found by assuming no change. In any case, if there’s a TxOut address that’s also a Txin, assume that’s the change address.