<gingeropolous> "is it possible to roll jamtis..." <- Is your question can a jamtis style address be used with the ringct algorithms "with no additional functionality beyond current" or "with additional jamtis functionality"?

with no additional functionality beyond current.

<gingeropolous> " with no additional functionali..." <- I find that to be an interesting problem. I will look into it and either propose a way to do so, or learn more about why there is no reasonable way to do it.

Isn’t that contradictory? It wouldn’t be jamtis without jamtis functionality.

My interpretation of the question is can we publish valid jamtis style addresses, but feed it into rct machinary to spend/receive. The *minimum* benefit there being that when seraphis is implemented, everyone can keep using the valid jamtis address they had been using. But more functionality will be available. And *perhaps* some additional jamtis benefits could be used as well.

no that is impossible

seraphis addresses have generators not present in rct addresses

UkoeHB: The jamtis scheme has three public keys. Two of them use G. One uses U and X (the one with the spend key). Couldnt rct be adapted to use U and X instead of G in places where the spend key is used?

if you made a really big effort with new cryptography, yes probably

but it isn't possible with current proofs

Suppose a sender used K1 from jamtis in place of the public spend key in RCT. What if the sender publishes two ephemeral public keys: rU and rX. I think the reciever could remove the U portion. Then go on to use rX in place of rG in standard RCT algorithms with an additional proof showing they removed the U portion.

You could in theory send a RCT output to a jamtis address by only using the keys K2 and K3. But the recipient would have to use a non-standard key derivation to detect such outputs.

That would be easier as they both only use G. I am imagining thats what ginger was getting at. Could a small tweak allow jamtis addresses to be used prior to Seraphis?

The follow up questions would be "why", "how best to", and "should". At a minimum, users could adjust to the new address format. But, perhaps much more could be accomplished without getting complicated? Could view tags be used? Could RIDs be used?

gingeropolous: ^ did I read your question right? Or have I gone off to another tangent entirely?

Hello everyone.