Jump to content

Professor Hantzen

Bronze Member
  • Content Count

    529
  • Joined

  • Last visited

  • Days Won

    1

Professor Hantzen last won the day on February 22 2017

Professor Hantzen had the most liked content!

About Professor Hantzen

  • Rank
    Advanced

Contact Methods

  • Website URL
    https://twitter.com/phantzen

Profile Information

  • Gender
    Male
  1. I"m not sure how being able to view the current list of proposed transactions allows for front-running? The most an actor with that information can do is submit a reactionary transaction into the same proposed list - however, they cannot determine the execution order of the finalised, validated set, so would not success at front-running by this method essentially be random? I guess it would be possible for a bad validator (to whom you are choosing to submit transactions to directly) to provide a fake response, and submit their own transaction in your place, silently delaying yours until the next consensus round. However, you would notice this and be able to switch to another validator instead (or just run your own from the outset).
  2. Professor Hantzen

    XRP Ledger transaction volume surpasses total trading volume

    Historically, payment volume often far exceeded transaction volume. Its easy for even a single actor (with at least two accounts) to spam up millions and billions of payment volume at very little expense sending the same funds back and forth. Transactions however usually will incur significant risk if one were to attempt the same as they have to deal with price/orderbook fluctuations and/or spread, and/or IOU fees. As such, it's likely payment volume will peak out greater at times and with no impactful cause. The balance between the two is therefore not a reliable indicator. Maybe in a few years when the landscape has changed - but at the moment the ledger is mainly populated by people testing things out. Any single entity conducting a relatively cheap on-ledger payments stress-test could for example cause a peak in payments volume. Or an actor with badly written code (something we've seen a few times in other areas).
  3. Professor Hantzen

    secp256k1 endomorphisms

    @caroma If you enable this and are able to improve rippled performance, maybe post a github issue with your findings (https://github.com/ripple/rippled/issues) and see if this can be made the default?
  4. That depends on the use-case and how it's set up. In some situations there will be an advantage - possibly significant. In others, there will be none. In the case of a malicious actor executing a predefined set of transactions for a market maker for example, the net loss may be minimal - a bunch of sells or buys executed and resulting in a unintended configuration of funds and the payment of fees, but no actual loss. However, the same malicious actor gaining access to a hot wallet could do whatever they wish with whatever funds that wallet controls.
  5. Professor Hantzen

    Stock Node vs. Validator

    @nikb @mDuo13 ?
  6. Professor Hantzen

    Bitstamp selects trading system from Cinnober

    Assuming its Fair Fair Fair...
  7. Professor Hantzen

    Does XRP Have a Name Problem?

    @Stormchaser417 It's not called or pronounced "zerp". That's an unofficial nickname given by some community members (personally, I can't stand it - for similar reasons to you - but I believe people are free to call things whatever they want and from that angle I support it 100%). Any official reference always pronounces each letter "X-R-P", for example, when spoken about by it's inventors. "XRP", pronounced as an acronym, is the name.
  8. This is a great question, and a correct answer may be as varied as there are exchanges. I can't say anything for sure, but in my general experience, exchange order-matching (in any market) can be a dark and unfair place. In a more demonstrable sense, you'll find that when things get busy on most exchanges, they slow down processing orders or even die completely. This suggests many may simply keep the various clocks in different areas synchronised, and make a (nanosecond perhaps) timestamp when receiving a trade. Then they execute in a distributed queue which, when large enough, causes delays in order processing as trades are sent back and forth (I'm sure we've all experienced an exchange order taking aeons to return a result, especially during high market activity). They likely also have various contingency plans which may violate the expected order (and so also fairness) to get things moving and/or mitigate liquidity problems. The metric you want to be tracking is not ping, but order confirmation time. Run a timer from submitting a trade to receiving a response. Try that from different locales and see what you get. It will likely be much larger than the ping in general, and (hopefully...) more consistent regardless of where you are testing from. Only test locales simultaneously, and across similarly-liquid markets, as different times of day and market activity will likely affect results.
  9. Professor Hantzen

    Why. ledger close interval decreasing?

    If you look at the charts above that one, transactions trended down over the same period of time (Sept-Oct), so maybe ledgers were able to be closed more frequently as there wasn't as much to get done. This isn't always how the network responds - if you look at other times, transactions going up coincided with faster closes. Possibly it has this time because of the type of transactions that decreased. At the bottom of that page, you can see transactions by type and result. There is a large drop in OfferCancel's over that period, and a fall-off in the many tecPATH_DRY's that were coming through. Perhaps both of these dropping together resulted in less load on the consensus process and so a faster rippled.
  10. Professor Hantzen

    XRP UNITED - rP0w3rP01nT.xrp

    This is certainly a concern. I think the marketing is spot-on for who they're going for, but they do need to address the trust issue in a big way to attract more than this crowd. Though I believe people who will respond to this style of marketing are a huge part of the current crypto demographic, they are not necessarily the part with the most money or the longest retention.
  11. Professor Hantzen

    XRP UNITED - rP0w3rP01nT.xrp

    Great video, but the voiceover volume is too low relative to the music. You should also cut the last minute of the video. Videos with longer runtimes can be discouraging for people to view, and this is effectively 1:40 not 2:40. Certainly its annoying for the viewer to have to skip through to discover there is no further information. Other than these two criticisms though, its really great - I think you're doing awesome work.
  12. If you need to sign transactions dynamically and frequently via a script, there is really nothing you can do to improve security beyond that of the machine running the script. The secret will have to be stored in memory somewhere, so your script can use it. You could use a secure element, but if the script can access the secure element, so can someone with control of the machine. If you have enough similarity in your transaction dynamics, you could theoretically sign batches of valid, labelled transactions offline (by sequence number), and manually transfer them to the online machine at regular intervals. Your script could then select the appropriate transactions throughout the time period. The danger then is limited to a malicious actor choosing to execute only the full set of transactions for that period of time. As this was going to happen anyway - and depending on what your script is doing - this may minimise any potential damage somewhat.
  13. Professor Hantzen

    Bot trading with Gatehub

    Yeah, as soon as traditional banking is involved things get slow and costly. Hopefully sometime soon, that won't be necessary anymore.
  14. Professor Hantzen

    Bot trading with Gatehub

    Sure - that's what I was (perhaps badly...) describing. But bear in mind that if you don't want to do more than one round, then you can only make the profit for one round. Let's take 3.5%, or $350 on $10,000 as your target. First off, you're going to encounter slippage from going that deep into market depth and that will vary every few minutes or seconds. For example, at the time of writing, there's actually only $200 "available" - so how will you decide when to trade? And what's your guarantee the market won't move underneath you moments before you place the two trades and have you grabbing a loss right at the outset? With further unknown SWIFT expenses ranging from $50 to $150 or more at least on entry (and possibly also on exit, even if you're only doing one round), plus two lots of $20 (0.2%) or so for the trades themselves, then paying tax on whatever profits you may make, seems like a pretty large chunk of hassle and risk to try and net only slightly more than your costs, assuming the risks don't wipe those profits out completely. Even if you succeed, and get say $50 - $100, is that a worthwhile payoff for tying up $20,000 of capital for an unknown length of time? (Possibly weeks or more if you really get on the wrong end of "SWIFT"). After you've done that one round, you'll have a pile of XRP on the wrong side (easily fixed), and a pile of USD on the other wrong side (not so easily fixed, as you have to use SWIFT and incur that non-fixed expense). If you're then thinking of doing it again, you need also bear in mind that your actions may affect the market, the depth is not very solid with the Gatehub IOU and volumes are low. If you do decide to do it a few times, chances are your repeat buying of $10,000 worth of XRP on the Gatehub side will drive the price in that market up to parity (that amount is sometimes more than the entire trading volume for that pair in a day). Profitable, low-risk arbitrage opportunities don't typically stick around for long, typically seconds if not milliseconds as there are already multiple bots trying to exploit them (and when one bot gets it first, the second, third bot etc grabs a loss instead). If you see what looks like an arbitrage opportunity laying dormant for days or weeks, ask yourself first - why hasn't someone written a bot to exploit this? Why did the bot-writers look at this and think "nah!". If you're still interested, carefully look at all the risks involved, play out every step on paper first - when you map it out, each steps risks and costs will almost always add up to the difference you're looking at grabbing, that's literally why the difference is there. Unless you've found something particularly exotic or unknown - in which case, definitely don't post about it on a forum, letting me know about it...
  15. Professor Hantzen

    IOU decimal digits precision problem

    Can you give the transaction hash? tesSUCCESS is a response regarding the process of submitting a transaction - it only means that the transaction has been *submitted* successfully - it doesn't necessarily mean the transactions intentions were honoured. There was an amendment a while back called TickSize which allowed each IOU to set its own precision maybe have a look at that for more detail on how this aspect of the ledger works.
×