The Future of XCP Protocol Fees

[davestaxcp]

This question of changing the XCP fee system has been asked or presented a good deal of times...

---

But the theory historically of XCP fees in general is quite plain:

Initial XCP creators destroyed BTC for XCP to exist.
Using XCP means destroying its supply.
Destroying XCP was historically commonly referred to 'anti-spam' fees (and with good reason)

---

To initiate certain functions on the protocol you need XCP:

• Issue token (Named)

only one Named asset of any specific name can exist (makes sense)

Fee has always been a flat 0.5 XCP fee.

• Sweep Address

referenced as 'anti-spam fee' by John Villar https://github.com/CounterpartyXCP/CIPs/blob/master/cip-archive/cip-0020.md

is a dynamic fee structure related to how many assets (and assets ownerships) user is sweeping: CounterpartyXCP/counterparty-core#1220 (comment)

this should be estimated and shown to the user by the protocol and by wallet infrastructure provider: CounterpartyXCP/counterparty-core#2636

• Fairmints (XCP-420 standard- see: https://x.com/droplister/status/1968154337571406232 + https://www.xcp.fun/

uses XCP to fairmint, but if goal is not reached XCP is returned

also introduces idea for requiring XCP to start fairmints

• Dividends

uses dynamic fee

ex: https://tokenscan.io/tx/3112997

this should be estimated and shown to the user by the protocol and by wallet infrastructure provider: CounterpartyXCP/counterparty-core#2636

• Attach Asset to UTXO

uses dynamic fees > CounterpartyXCP/counterparty-core#2893

---

With this in mind, are there any functions which should require an XCP 'anti-spam' fee?

• Start Open Fairmint with custom options (as issuer, not user 'minting' the open fairmint)
• Counterparty Inscriptions (taproot)
• Future Market Makers? > DEX Market Makers (?) counterparty-core#2749

---

Changing fee dynamics or even the theory of XCP fees has been discussed in the past:

• "the introduction of XCP fee on every meaningful Counterparty action" > Enhancing the XCP Ecosystem: Towards a Stakeholder-Driven Network #129
• Using dynamic XCP fees for more than just sweep and attach > Use Dynamic Fee for More than Just attach (?) counterparty-core#2893
• adding fees for specific protocol functions 'numerics' in this circumstance > https://forums.counterparty.io/t/xcp-fee-on-numeric-assets/2445
• probably many other places... feel free to reference anything I missed

---

I think the reason this issue CounterpartyXCP/counterparty-core#2893 was so contested is because of the theory of the dynamic fees specifically presented in that post.

The main contentions there was:

• adding fees to features that do not already have them (users would be angry and it would cause friction)
• adding a dynamic fee structure that is EXPENSIVE in high usage environments (especially assuming the price of XCP rises during more usage and more new users interacting) ... we have already suffered too much for high BTC fee environments
• adding fees in a manner that doesn't take into account individual features and those individual theories of 'anti-spam' in those circumstances

---

I personally am open to adding dynamic Fees to existing features like Issue New Asset
... though making it cheaper on a protocol level could anger previous issuers if the fee becomes too cheap
... it could also incentivise the opposite of the thory of the fee in the first place = 'anti-spam fee'

I personally think each feature we add or change a fee to should be discussed individually.

Would love to hear the community's thoughts on the fees we already have and features that might benefit for the theory of 'anti-spam functionality'

Do we still agree with this original thesis?

---

edit: Dividends do have dynamic xcp fees... but we barely notice them! (i didnt!)

[droplister]

I think if the supply is changed and XCP was applied consistently as a gas following the type of logic attach utxo uses where it is zero fees when usage is low makes sense. The tx fees on chain have been 0.1 sats/vB for months, which is like 25 sats for a tx, so I wouldn't say we're suffering from Bitcoin tx fees. And it was very clear on Ethereum that there was plenty of appetite to pay high fees during high usage, like $300-1000 a tx on Ethereum during big spikes in use. I don't think that's the target or goal here, but evidence of actual behavior.

But if we can get to a place where you don't need to know Feature A costs X and Feature B costs Y and the wallet just says this is the protocol fee for this tx, it would be inline with how other things work and make XCP demand more robust which makes the DEX more robust.

[davestaxcp]

@adamkrellenstein I did finally check up and changed this post because Dividends do have a dynamic fee.

• Nobody has ever brought up (that I've seen) that Dividend fees are excessive or annoying.... probably because a user that owns a named asset probably has enough xcp dust to use this feature without thinking

how does the protocol treat Dividend fees and how are those fees derived?

for example Sweeps use this model:
CounterpartyXCP/counterparty-core#1220 (comment)

[adamkrellenstein]

Yep, it's very small — 0.0002 XCP per asset holder. Only for non-BTC dividend assets (BTC dividends are fee-free). Was enabled with a protocol changed on block 330,000.

There's also an API endpoint to estimate fees before composing: /v2/addresses/<address>/compose/dividend/estimatexcpfees?asset=MYASSET