During : What are the downsides of ?
@cwebber : Well, they can be delegated really easily (but so can ACLs by proxy so ACLs are misleading)

me (in my head) : noooo, ACLs are difficult to delegate in practice and that's what counts. BUT ocaps can also be made difficult to delegate by requiring each action to be signed using a key that is also used for other things !

@cjd @cwebber this is an important point for me.. Preventing delegation might be equally mathematically impossible, but I care more about how impossible it is IRL :)

@lain @cjd I think I can express how the ideas from Horton (I'm not saying adopting Horton as a protocol itself, but the underlying idea) can help us achieve what we want, maybe even more powerfully than what we currently have. But I have to write it up.

The good news is that I *am* writing it up, this week and next with the paper on secure user interfaces coming out of RWoT.

BTW @lain I am taking your comments to heart from the ocap session about needing to explain more clearly what this can give us that an ACL approach can't. I'll work on that.

And actually, I was pleased with your point that "we can adopt stamps even if we don't adopt ocaps"... if even only that were to happen, that would be a useful contribution from the ocappub to the greater fediverse even if it didn't use ocaps-for-actors. :) Incrementalism!


@cwebber @lain @cjd Question about stamps: It sounds like proof-of-work. How do you imagine they will be generated?

@gargron Stamps were left ambiguous so that proof-of-work was one option, but I think it's less interesting than the "tradeable" stamps, and it turns out that pretty much everyone at APConf agreed that the PoW version isn't as interesting as the tradeable "getting paid" version, so let me unpack how "untradeable stamps" works.

By the end of explaining the plan I really like @lain seemed convinced, so I'll explain in this thread and ask lain to clarify if I missed anything so far.

@gargron Ok, so first of all, no blockchain or PoW for tradeable-stamps required. Instead, we move to the idea that anyone, or any community, can establish a "live" fiat ~currency like token system. Literally, anyone can make it, and how much "value" it has is up to the users. From chuck-e-cheese tokens to porn cams, lots of sites / places have these kinds of "not really money but kinda" tokens, and so do games like WoW and etc.


@gargron Let's ignore exactly how it works but just accept a few facts:
- anyone can make them
- they can be very local or very global
- people decide how much value they have
- they can be traded, bought, sold
- the entity in charge of them is responsible for how much they inflate them by issuing them
- they are NOT on a blockchain, though if someone chooses to back them by a blockchain (the way US dollars used to be redeemable for gold) they can, but it's not strictly necessary


@gargron @lain So how to build it? Two paths: the ocap folks demonstrated you can make "money" basically from ocaps alone (see "simple money" on erights.org/elib/capability/od )

But, that doesn't provide all the properties we want, even though it is easy to make and issue them.

@gargron @lain So what should "tradeable stamps" be? What about Digicash? GNU Taler is a FOSS implementation (not sure if it's the one you want, but it is one), and it's actually an old design, but it was patented and heavily litigated. Well, the patents have expired now, and it has some good ideas:

- You withdraw tokens
- Once you withdraw them, the server doesn't know which ones you have
- You get receipts for transactions
- Buying is anonymous
- But selling is not!

@gargron @lain So those last two (buying is anonymous but selling is not) are highly interesting. You *could* set up the Digicash ~bank to operate in an anonymous manner, it's kind of a choice (I won't explain that fully here), but that means two paths open up:

- "legit, by the books" currencies which actually are fully taxable
- more IOU-like per-community tradeable token systems (which, of course people will trade for real money, but...)

@gargron @lain Anyway, NO blockchain involved, NO proof of work necessary. Transaction times are small and fast. It does mean "trusting" the digicash bank, but that's why it's important to have an open market and let anyone trade the stamps; eventually, some reasonably well trusted stamp providers will emerge.

And... this lets us open up a path towards an alternative to Facebook Libra and Weechat that's actually respectful of user freedoms and individual autonomy.


@cwebber @Gargron @lain
The concerns I raised with this model were:
1. Financializing human interactions can lead to much scarier emergent behaviors because "internet is serious business" becomes reality.
2. If the stamp is legally classified as a financial asset and an admin has any power over it, the admin becomes a bank (see: E-gold). WoW avoids this by having lawyers lobbying to prevent it's classification as financial. Blockchains avoid this because admins have no power.

@cjd @gargron @lain I'll argue again that blockchains haven't succeeded where you're claiming, and I believe I know that because I'm close to several people who work on blockchains.

a) coindesk.com/new-irs-warning-l
b) the people I know who are working on blockchains are being careful to distinguish between tokens that are primarily for a "useful purpose" (eg filecoin) vs ones that are meant to be money exactly for financial oversight reasons

@cwebber @Gargron @lain
I think you're referring to the split between "utility tokens" and "security tokens"

The genesis of this split was because so many ICOs were trying to paint themselves as a decentralized blockchain but really they were pre-mining the coins and then going around convincing people to buy their them order to have future value which starts to look like a "Security Offering". Securities are cool but that's a whole nother layer of regulation (and for good reason).

@cjd @gargron @lain Aha... you may be right, I might have misunderstood the term "utility token"?

@cwebber @Gargron @lain
So "utility token" is a token which is useful on it's own (e.g. ethereum is useful for getting global consensus on the result of executing certain code)

I think there's an unreasonable tenancy of people to avoid things labeled "blockchain" because people think there are limitations which are not actually fundamental to the tech.

1. A blockchain doesn't need work -> PoS, DPoS, private
2. A blockchain doesn't need a native token/coin -> Liquid

@cwebber @Gargron @lain
3. A blockchain doesn't need to grow forever -> pruning and checkpointing

So the positive side of using something which is at least nominally a blockchain is that there's plenty of legal guidance to inform you whether you're crossing into the scary world of "you are now a financial institution".

Using something like Liquid allows everyone can issue their own token rather than trying to convince the fedeverse to get consensus on one coin which is IMO a non-starter.

Sign in to participate in the conversation

Server run by the main developers of the project 🐘 It is not focused on any particular niche interest - everyone is welcome as long as you follow our code of conduct!