Have a concise name for the cryptographic protocol described in RFC003

thomaseizinger commented 5 years ago

Problem

On several occasions, we learned that "rfc003" is not a good name for a protocol. It is just not descriptive about what the protocol actually does. Now that we are starting to market COMIT, the names of protocols become public facing, hence they should be good. This will be a very impactful change but I think the added clarity is worth it.

Most importantly, if we go for https://github.com/comit-network/RFCs/issues/121, we are going to need a good name for the protocol string. /comit/rfc003 is kind of odd because the protocol by itself would be described in a different RFC (maybe?).

Goal

Have a concise name for the cryptographic protocol described in RFC003. This is just about deciding on the name.

Recommendation

grab Thomas but don't grab self!
Take a whiteboard and list all important key words that make up the protocol (e.g. HTLCs, 2 party, 4 transactions, etc)
Make up a sentence or a phrase that describes the protocol
Reorder the phrase until you can form a good-sounding acronym from the initial letters of the words.
Don't create a spike doc about it

bonomat commented 5 years ago

https://www.fantasynamegenerators.com/ancient-greek-names.php :P

D4nte commented 5 years ago

Also drop the "swap" terminology, or at least, do not include in this name and drop it slowly?

bonomat commented 5 years ago

If I remember correctly, we said some time ago we want to wait for the next protocol before we rename it so that we don't get a clash with the meaning of the acronym.

// Edit: what I want to say is: when thinking about RFC003, also keep in mind that we will have a HTLC to LN, Scriptless Script for BTC-ETH, Scriptless Script for BTC-GRIN, ... some time soon™️.

If

Now that we are starting to market COMIT, the names of protocols become public facing, hence they should be good. This will be a very impactful change but I think the added clarity is worth it.

is a main concern, then I propose to consult the marketing team before making a final decision.

thomaseizinger commented 5 years ago

consult the marketing team before making a final decision.

The TenX marketing team? Or who are you referring to? It is not my main concern to be honest :) Personally, I care more about the name being good by itself (when being used within the codebase, in responses, protocol names, etc) and less of why it is important now :D I wanted to list reasons that other people can hopefully relate to to get support for pushing for this now :)

bonomat commented 5 years ago

thought so :P

tcharding commented 4 years ago

Suggest we just use this thread and come up with a name instead of using a white board. If we all add any terms that are relevant then any of us can do the rearranging and make suggestions. As the newest, and one with the least domain knowledge, I'll kick it off.

tcharding commented 4 years ago

Hashed Time Lock Contracts
SHA-256
two party
Atomic
Single round trip message requirement
Non-custodial
Token free
On chain
Linkable
Large/verbose contract
Block time limited

bonomat commented 4 years ago

Imho an unambiguous name has to contain HTLC in it somehow. Everything else could be any (atomic swap) protocol.

Now, when coming up with a different name, keep in mind that we will have different protocols in the future:

Different scriptless Script-based protocols (think GRIN-BTC, BTC-ETH, ...)
HTLC-based protocol to swap into LN
HTLC-based protocol to swap into Raiden ...

D4nte commented 4 years ago

on chain HTLC Layer 1 htlc

tcharding commented 4 years ago

Layer 1 htlc

Is there any reason not to use these three words? Does this clash with any near future protocols?

bonomat commented 4 years ago

Done in https://github.com/comit-network/spikes/pull/47 @da-kami & @thomaseizinger : please create follow-up tickets.

comit-network / RFCs