Transcript for #bitcoin-dev 2018/02/14

05:57 stoopkid hi, is there any mechanism for trustless exchange between omni assets and btc that only uses standard btc transaction features?
06:05 Dizzle stoopkid: I doubt it
06:05 Dizzle but don't actually know
06:05 windsok I think Omni has a DEX http://omniexplorer.info/dex.aspx
06:06 Dizzle windsok: yea, but I think that's implemented in the Omni layer instead of the Bitcoin contracts alone
06:07 windsok but all Omni transactions are bitcoin transactions
06:09 windsok http://omniexplorer.info/lookuptx.aspx?txid=6dcd8eedf2f01480833ecbeb7998c09d38faf708b8f9fe0af5251...
06:09 windsok vs
06:09 windsok https://btc.com/6dcd8eedf2f01480833ecbeb7998c09d38faf708b8f9fe0af525165c3eafd094
06:10 windsok uses OP_RETURN
06:11 stoopkid how does the distributed exchange work i'm having trouble finding concrete technical info
06:14 windsok I'm pretty sure it's the same or very similar to how counterparty works. They might have better docs https://counterparty.io/docs/
06:14 windsok Note that the way both omni and counterparty are implemented does not really scale.
06:14 gmaxwell Today on the internets: people investing in altcoins by people too inept to setup their own p2p network. :P
06:17 stoopkid windsok: what i was originally thinking was that you should just be able to use a 2-of-2 multisig address between the counterparties with presigned "refund" tx's before they send their assets there to make sure they can still pull them out in case the exchange fails to go through
06:18 stoopkid since it's just on the btc blockchain it should basically be like exchanging bitcoins for bitcoins, unless there happen to be any peculiar details of the omni protocol that get in the way of that
06:19 stoopkid ideally it would be a scriptless script (to w/e extent omni could allow it) to minimize potential points of failure since i'm not really a security expert
09:08 RainMan28 hey gmaxwell, saw your comment on reddit regarding bitpay and BIP70. Do you think their payment protocol is a bad idea?
09:15 stoopkid anybody have any references on what goes in the data for omni class C tx's?
09:18 gmaxwell RainMan28: well I've seen several people with long lists of complaints and shortcomings in BIP70, and pretty much none of them appear to be addresses except the send on one side thing. And at least in my brief glance at what they posted, they didn't answer the criticial questions that caused bip70 to have the always broadcast to the network design: things like how do you handle the case where the far
09:18 gmaxwell side never sends to the network, or how do you distinguish a cancellation vs a new payment.
09:18 gmaxwell though I didn't look at it carefully.
09:18 gmaxwell I also don't bother wasting my time on any bitcoin protocols that don't want to write a bip and take public comment, -- waste of time.
09:20 RainMan28 ah gotcha
09:20 gmaxwell maybe it turns out fine, --- would be hard to be that much worse than BIP70 IMO. :)
09:21 RainMan28 Oh yeah, I definitely empathize with them having a need for it when the fees were really high. Having an ATM network, we experienced a lot of the same issues they were experiencing when customers would try to sell their BTC for cash. Sending the wrong amounts and then requiring refunds, and losing $30 in sending and $30 in getting a refund, etc, etc
09:22 gmaxwell yes, indeed, though BIP70 just didn't do anything helpful there.
09:22 RainMan28 BIP70 was more about how to make a direct payment to their node vs having it be broadcast across the network?
09:23 gmaxwell That is why they said they forced BIP70 but bip70 doesn't do that.
09:23 RainMan28 there are so many BIPs lately, it is hard to keep up
09:23 gmaxwell Every BIP70 implementation that I'm aware of sends to the network.
09:24 gmaxwell (There was a big argument between gavin and sipa when it was proposed, sipa and others wanting it to be specified to not send to the network; but to do that you have to address how a bunch of corner cases are handled and gavin didn't want to deal with that.)
09:24 RainMan28 oh very interesting
09:25 gmaxwell like what happens if the txn is never sent to the network? when do you return the txouts to the wallet? What if the far side rejects it? How do you know this and know when you make a replacement to be sure to conflict the original?
09:25 gmaxwell If the reason they'd reject it is fees, they probably should be able to communicate a minimum feerate in the request or you might loop forever. :P
09:34 RainMan28 oh those are good points
09:38 Megumiin Bitpay did relent and they now offer a non-payment protocol payment flow
09:39 Megumiin But it looks like it might be limited to non-javascript enabled browsers
09:40 gmaxwell Megumiin: oh thats great to hear, (first statement)
10:15 stoopkid hello, can anybody provide a reference on the data structure used in omni class C transactions?
10:40 wumpus Megumiin: awesome news? how do they limit it to non-javascript-enabled browsers?
10:40 Megumiin <noscript> tags most likely
10:41 wumpus in that case, blocking javascript for just the bitpay site will work
10:41 Megumiin not exactly userfriendly
10:42 wumpus but better for privacy, I guess
10:42 gmaxwell someone could publish a browser extension. :P
10:42 Megumiin I assume they only added it because of their tor issues
10:42 Megumiin can confirm it's noscript
10:43 wumpus ah yes, instead of fixing the fetching of payment requests over tor
10:44 gmaxwell pay-bitpay-bits
10:45 wumpus not that the only problem was over tor, bitcoin wallet for android had trouble using/paying the payment requests from bitpay even over plainnet
10:45 Megumiin but this doesn't change that for them
10:45 Megumiin since they still run js
10:45 wumpus right
10:46 wumpus discoverability is an issue too - who thinks of disabling js, when they have a payment issue
10:47 wumpus at least I would never have discovered that if it wasn't said here
13:01 wumpus I thought I was going to be able try it out, but the VPS provider has moved to coingate instead of bitpay
13:41 wumpus maybe next time I order pizza, if takeaway is still with bitpay
13:43 Megumiin better order fast before bitpay bleeds all customers to alternatives
13:48 wumpus heh
22:51 Azelphur So, trying to use bitcoinds rpc interface to generate an address, bitcoin-cli help seems to have no mention of getnewaddress (or anything like it), the docs say they exist, but on my installation, they don't. Has this functionality been removed? starting to wonder if the ArchLinux package has been messed with, or something.
23:43 Azelphur Figured it out, archlinux systemd script starts bitcoind with -disablewallet
23:56 mlz lol