4 ms·

I feel like the article definitely has some merit but the timing of it (days before a TBD announcement on token trading), the inclusion of very obvious non-uniq

by sushid 9y ago

I feel like the article definitely has some merit but the timing of it (days before a TBD announcement on token trading), the inclusion of very obvious non-unique points (i.e. not scalable, anything Bancor can do Ethereum can do - uh, why not include the definition of Turing complete here) and the false assumption that Bancor's use case is for a groundbreaking ERC20 token that's going to be traded on every exchange makes the author's claim dubious. A more likely use case for Bancor is some SMB (store/restaurant) or content creator issuing tokens to be used for membership reward/points or for voting on a new video/project idea, not for Augur/Steem/Gnosis 2.0 to issue an ICO. If you take this gross assumption of the equation, his whole argument about slippage or "trailing the market", etc are gone. You're issuing some small time tokens. There no reason to always buy back from the market and there are no big players trying to "deplete your reserve." The author makes it sound like I'm selling some popular cryptocurrency at a fixed price and am being left to bleed dry when in actuality, I'm issuing my_random_token for friends and customers. I didn't review the code but if what he's saying is correct, there is a reason for alarm regarding their reimplementation of basic arithmetic functions(!!!) and user overpaying. By trying to make this into a laundry list, it comes more off as a FUD argument, very similar to what Bitcoin maximalists were doing post the DAO hack to Ethereum. (Disclaimer: my Bancor holding is about 0.5% of my total Crypto portfolio).

MichaelGG 9y ago
>(store/restaurant) or content creator issuing tokens to be used for membership reward/points Why would a restaurant use this for membership point when a glorified Access db would work just as well?
- Noxchi 9y ago
  We are witnessing the dotcom bubble version 2: crypto bugaloo
  - edoceo 9y ago
    We're gonna rock down to electric revenue. And then we'll take it higher!!
- zanny 9y ago
  Seriously, trustless networks are only good for arbitrage. Everything else doesn't need trustless when its all your own thing you control. Its just a lot of overhead and complexity you don't need then.
- sushid 9y ago
  You can say that about nearly all of these projects. Why Steem when Reddit/HN? Why Augur/Gnosis when we have sports and political betting websites? Why Sia/Maidsafe when we have Dropbox/Box/GDrive? And for the record, an SMB was just an example of a small-time issuer as opposed to a massive ICO worth millions. But the fact is, we continue to iterate and rewrite things in technology. Why Dropbox when I have a thumb drive? Why CRM/Salesforce when I have a Rolodex? Why noSQL when... SQL? No one is saying blockchain technology will cure cancer or is imperative in bringing about a change.
  - MichaelGG 9y ago
    Some of those offer actual benefits. How does a restaurant benefit from using this tech to keep track of rewards points? Other than marketing "it's like food, but with blockchain!"?
    - sushid 9y ago
      Look, again, it's just an example of a small scale use of Bancor's token issuance. Not a how-to-guide for all SMB owners. Using Bancor does not mean that the SMB is doing this all from scratch. After all, how many restaurants do you see implementing their own POS system with in-house engineers? And to be honest, how airtight do you think a mom and pop shop's MS Access DB backup strategy would be? I would trust "food with blockchain" over "food with please-don't-die-on-me production DB."
      - MichaelGG 9y ago
        OK, but even a tiny example should make sense. OK so use Google Sheets instead of Access? Or anything? I don't get where the distributed, trustless, properties of blockchain tech make any sense for a centralized DB with only one source of truth (the restaurant).
        sushid 9y ago
        Again, I get where you're coming from. "But what's the point?" is usually an argument that can be made for any new technology if you don't believe in the features that make it exciting and new. That doesn't mean people can't or shouldn't explore in a yet-another-noSQL DB, a new MVC framework, a new low-level language, etc. This hypothetical store doesn't have to use blockchain technology, but it could and reap the benefits or potential harms of doing things the new way.
maldeh 9y ago
A more likely use case for Bancor is some SMB (store/restaurant) or content creator issuing tokens to be used for membership reward/points or for voting on a new video/project idea, not for Augur/Steem/Gnosis 2.0 to issue an ICO. I question if their massive ICO was primarily backed by the marketing promise of restaurants issuing their own currencies. Even if that were the case, I'd still be concerned if this protocol can only be safely used for small-time operators, while any token with any semblance of scale on the chain is thoroughly gameable. On a side-note, Emin Gün Sirer made a lot of noise about re-entrancy exploits present within the DAO in the months preceding the hack ( http://hackingdistributed.com/2016/05/27/dao-call-for-moratorium/ http://hackingdistributed.com/2016/05/27/dao-call-for-morato...), so I wouldn't take his concerns about the Bancor codebase lightly, and may even be inclined to give him the benefit of the doubt vis a vis FUD.
btown 9y ago
Sure, the OP may very well have a conflict of interest. At the very least, he has the conflict of interest of wanting to draw attention to his own post. Hanlon's Razor would suggest that the OP simply went overboard in finding criticisms, where only a few would do. But independent of intent, the logical arguments about trading phenomena make sense, and what's worse, the code smell is very, very real. If you want to see the source for that screenshot, here it is in all its gory detail: https://github.com/bancorprotocol/contracts/blob/f198f7d66580578ba06730f415d4276bbf6195c2/solidity/contracts/BancorFormula.sol#L221 https://github.com/bancorprotocol/contracts/blob/f198f7d6658... Oh, they say it's exposed for unit testing! Let's grep the codebase for fixedExpUnsafe... nothing. Oh, and it's in the same file as the business logic; the only test for that file is for the end-to-end logic. "Forget the unit tests, the integration test will catch it." Yeah... Not the type of code I'd bet my business on.

by sushid 9y ago

MichaelGG 9y ago

>(store/restaurant) or content creator issuing tokens to be used for membership reward/points Why would a restaurant use this for membership point when a glorified Access db would work just as well?

Noxchi 9y ago
We are witnessing the dotcom bubble version 2: crypto bugaloo
- edoceo 9y ago
  We're gonna rock down to electric revenue. And then we'll take it higher!!
zanny 9y ago
Seriously, trustless networks are only good for arbitrage. Everything else doesn't need trustless when its all your own thing you control. Its just a lot of overhead and complexity you don't need then.
sushid 9y ago
You can say that about nearly all of these projects. Why Steem when Reddit/HN? Why Augur/Gnosis when we have sports and political betting websites? Why Sia/Maidsafe when we have Dropbox/Box/GDrive? And for the record, an SMB was just an example of a small-time issuer as opposed to a massive ICO worth millions. But the fact is, we continue to iterate and rewrite things in technology. Why Dropbox when I have a thumb drive? Why CRM/Salesforce when I have a Rolodex? Why noSQL when... SQL? No one is saying blockchain technology will cure cancer or is imperative in bringing about a change.
- MichaelGG 9y ago
  Some of those offer actual benefits. How does a restaurant benefit from using this tech to keep track of rewards points? Other than marketing "it's like food, but with blockchain!"?
  - sushid 9y ago
    Look, again, it's just an example of a small scale use of Bancor's token issuance. Not a how-to-guide for all SMB owners. Using Bancor does not mean that the SMB is doing this all from scratch. After all, how many restaurants do you see implementing their own POS system with in-house engineers? And to be honest, how airtight do you think a mom and pop shop's MS Access DB backup strategy would be? I would trust "food with blockchain" over "food with please-don't-die-on-me production DB."
    - MichaelGG 9y ago
      OK, but even a tiny example should make sense. OK so use Google Sheets instead of Access? Or anything? I don't get where the distributed, trustless, properties of blockchain tech make any sense for a centralized DB with only one source of truth (the restaurant).
      - sushid 9y ago
        Again, I get where you're coming from. "But what's the point?" is usually an argument that can be made for any new technology if you don't believe in the features that make it exciting and new. That doesn't mean people can't or shouldn't explore in a yet-another-noSQL DB, a new MVC framework, a new low-level language, etc. This hypothetical store doesn't have to use blockchain technology, but it could and reap the benefits or potential harms of doing things the new way.

maldeh 9y ago

A more likely use case for Bancor is some SMB (store/restaurant) or content creator issuing tokens to be used for membership reward/points or for voting on a new video/project idea, not for Augur/Steem/Gnosis 2.0 to issue an ICO. I question if their massive ICO was primarily backed by the marketing promise of restaurants issuing their own currencies. Even if that were the case, I'd still be concerned if this protocol can only be safely used for small-time operators, while any token with any semblance of scale on the chain is thoroughly gameable. On a side-note, Emin Gün Sirer made a lot of noise about re-entrancy exploits present within the DAO in the months preceding the hack ( http://hackingdistributed.com/2016/05/27/dao-call-for-moratorium/ http://hackingdistributed.com/2016/05/27/dao-call-for-morato...), so I wouldn't take his concerns about the Bancor codebase lightly, and may even be inclined to give him the benefit of the doubt vis a vis FUD.

btown 9y ago

Sure, the OP may very well have a conflict of interest. At the very least, he has the conflict of interest of wanting to draw attention to his own post. Hanlon's Razor would suggest that the OP simply went overboard in finding criticisms, where only a few would do. But independent of intent, the logical arguments about trading phenomena make sense, and what's worse, the code smell is very, very real. If you want to see the source for that screenshot, here it is in all its gory detail: https://github.com/bancorprotocol/contracts/blob/f198f7d66580578ba06730f415d4276bbf6195c2/solidity/contracts/BancorFormula.sol#L221 https://github.com/bancorprotocol/contracts/blob/f198f7d6658... Oh, they say it's exposed for unit testing! Let's grep the codebase for fixedExpUnsafe... nothing. Oh, and it's in the same file as the business logic; the only test for that file is for the end-to-end logic. "Forget the unit tests, the integration test will catch it." Yeah... Not the type of code I'd bet my business on.