chaser

joined 1 year ago
[–] [email protected] 3 points 5 months ago* (last edited 5 months ago)

they've done worse every time I used them: they jacked up the price well after my deposit was detected by them. I ended up losing several percents compared to the price they promised.

it's a scammy service and they invest a lot into event sponsorships (for conferences, meetups) to gain visibility. they also leave useless comments on GitHub issues, which also looks like self-promotion. just like this useless Lemmy post. I'll never use them again.

[–] [email protected] 3 points 6 months ago* (last edited 6 months ago) (2 children)

as someone who has studied both, I would recommend LUSD (v1) over dai. LUSD was launched 3 years ago, so it stood the bear test. the minimum collateralization ratio of 110% applies to individual troves as long as the total system collateralization is over 150%. once that's breached, troves are required to have 150% minimum. the Achilles heel is the oracle. if Chainlink pulls the rug, which they can, it's over (sadly, Tellor is used by Liquity in a way that it can't protect against a Chainlink apocalypse). Maker is somewhat better in this because they use Chronicle, which is ran by more trustworthy people, but I'm almost sure they haven't made their contracts immutable. if that is the case, then the same attack vector exists there.

as you'll see, neither of these are the solution we're looking for, and they both run on the no-privacy, hypercomplex, captured, constantly changing Ethereum blockchain, so... fuck.

but dai for a long time has not been what the market thinks it is. avoid it.

[–] [email protected] 3 points 7 months ago (1 children)

Rucknium brought up your concern during today's Monero Research Lab meeting, some people commented on it: https://libera.monerologs.net/monero-research-lab/20240410#c361656

[–] [email protected] 2 points 7 months ago* (last edited 7 months ago)

there is:

IRC: irc://irc.libera.chat/#no-wallet-left-behind

Matrix: https://matrix.to/#/#no-wallet-left-behind:monero.social

[–] [email protected] 3 points 9 months ago

nice catch! I will avoid them until they secure their infrastructure.

[–] [email protected] 3 points 9 months ago (1 children)

there is nothing that requires a perpetuals market to have a lower volume than the whole spot market for the underlying asset. perpetuals are derivative assets, the reason they exist is to enable trading regardless of access to the underlying asset's supply.

side note: Binance will shut down their XMRUSD perpetuals with the delisting, but they will keep running their XMRUSDT perpetuals (USDⓈ-M XMRUSDT):

Please note that users may continue trading USDⓈ-M XMRUSDT and ANTUSDT Perpetual Contracts.

(https://www.binance.com/en/support/announcement/binance-will-delist-ant-multi-vai-xmr-on-2024-02-20-f73b083ba6834771b07dbe5319917ae5)

[–] [email protected] 4 points 10 months ago

on the major existing decentralized exchange, Bisq, you already have a publicly observable price feed for XMRBTC. you can calculate XMRUSD based on that and a BTCUSD price feed that you trust.

DEXs that are in the making, namely Serai and Haveno, both plan to have at least one XMR pair with an Ethereum-based dollar-pegged coin. I suppose, although I'm not sure, that at least in the case of Serai the trades on that pair will have publicly observable prices.

Haveno will also have non-blockchain, actual fiat (cash, wire, payment apps) pairs with XMR.

you can take sources like this and calculate an average or a median. current price aggregators, like coin listing sites, will probably do the same, so it's likely that even in a DEX-only future you'll get your prices from the same sources as today.

[–] [email protected] 4 points 10 months ago (1 children)

heads up about the screenshots -- when you need to obscure text (or anything else) on a picture, don't use blurring, pixelization or similar effects. they can be brute-forced and the original text can be retrieved, with trivial computing power. always use color fill to obscure data, and don't make it exactly as big as the data is (~visual padding).

[–] [email protected] 1 points 1 year ago* (last edited 1 year ago)

this concept has been discussed before as "return addresses". it would marginally increase the transaction size and thus the storage requirements, but it's possible and seems reasonable. it looks like no one pursued it since knaccc formalized it. nevertheless it's on the roadmap ("Future" tab).

https://old.reddit.com/r/Monero/comments/b0gjud/monero_return_addresses_who_would_use_this_feature/

https://github.com/monero-project/research-lab/issues/53

view more: ‹ prev next ›