Jump to content

jbjnr

Bronze Member
  • Content Count

    406
  • Joined

  • Last visited

  • Days Won

    2

jbjnr last won the day on January 21 2019

jbjnr had the most liked content!

11 Followers

About jbjnr

  • Rank
    Regular

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Yes it would. But as they said around the time we were all analysing their sales - it's still good for independent researchers to check the figures - to keep them honest. When I first plotted their sales and correlated them to exchange volume, I got a figure of around 3% of daily volume being sold - which shocked me slightly as being so high. (for ref - posts (Nov 2018), (2018 Q4), (2019 Q1/Q2)) Later I started using different exchange volume data and their % dropped proportionally due to the higher volume numbers I was using, but they subsequently decided that their original exchange volume was probably contaminated by wash trading and switched to using a lower estimate (similar to the one I had been using originally) and dropped their sales correspondingly. I like to think that those of us tracking these things made a difference and helped to persuade ripple to reduce sales. I'm not sure that tracking ODL activity can impact anything much, but I'm sure that they will eventually release data of ODL amounts, corridors etc etc. Probably sooner rather than later - but - only if ODL turns out to save money for organisations like MGI. Once that has been established (and we still don't know if that's the case), they will no doubt want to tell everyone about it, but we're still very early in the game. One thing I think I'm seeing is that volatility does hurt ODL traffic, but I haven't really investigated this properly and will have to look into it further. I ought to have enough data now. The problem is that if ripple are subsidising MGI/FlashFX/others then it is much harder to be certain as traffic might continue even when loss making is occurring. I might need to look at trade data such as the stuff that @JASCoderhas been plotting to really investigate properly, but I don't have enough time to devote to this right now. My suspicion is that the ODL ripplenet interface includes a 'transfer when best' type option so that you can request a payment be made, but not immediately, so that if the rates fluctuate and are better than X or Y then the payment goes through, but if not, it goes via normal channels when some time limit elapses. This would explain spikes and dips in ODL traffic at certain times (similar to how an arbitrage bot works). Anyway. I'm sure we'll see figure from ripple soon. maybe the 2020 Q1 report will have more info. I'm hoping more channels open up soon, maybe the JPY/KRW stuff that utilityscan saw was the start of new corridor testing. fingers crossed - if they are new corridors, then my scripts will pick them up and they will be included in my plots and sums. (and yes - before anyone asks - I'm certain that increased ODL traffic will drive up price - we will just need a lot of it - and fewer economic collapses like the one happening right now).
  2. In general I'd say the answer is 'some'. The ODL tracker on https://utility-scan.com/#/dashboard looks at payments on the xrp ledger and compares them to trades on exchanges using the same amount. I think I read on this forum quite recently that they don't include it until they see 15 transaction that match the pattern. When those transactions are more than 24 hours old, they become part of the history, but don't affect new tests or numbers. However, there may be individuals who are doing ODL like activity - or random buy/sell pairs that appear to be matching that will appear as ODL activity on utilityscan.com - my own suspicion is that the numbers seen the other day including JPY, KRW etc were probably not ODL activity, but probably random trades that just happened to match the pattern. My own ODL tracker uses a different approach. I scan payments on ledger and do a very basic statistical check to see if the 'corridor' looks like ODL payments. By this, I mean are there regular payments from one account to another, using the same destination tag - are both accounts known as exchange accounts. Do the payments have sensible transaction sizes, do they fit a certain weekday/weekend pattern. When I find a corridor that 'looks' ODL-like, I then apply a filter to remove spurious corridors. for example, I see a great many ODL-like corridors that are probably arbitrage-bots between exchanges, casinos and other wallets that are 'unknown' - these I remove and what's left are then filtered down to 'candidates' and from there I keep only the ones that ripple have announced as trading partners, (bitso, bitstamp, coins.ph etc). When I plot the amounts on those corridors, I get numbers that match the ones shown on utility-scan.com - except for the JPY/KRW stuff - because I do not consider a corridor open until the trading pattern has stayed active for at least two weeks (I can increase or decrease that window). If a corridor is active during a period of time, and then goes inactive. I drop it from my list. This means that if I plot historical charts going back in time, traffic on these 'corridors' does not appear. In general, the amounts on these corridors is quite small by today's standards. If you look for example at this plot of weekly totals you will see activity USD/PHP and USD/MXN at low levels during the early part of last year - this was almost certainly testing of systems before they went properly 'live' later in the year. However, we have no proof of that and it might have just been normal traffic. My analysis relies on the assumption that if you were ripple/moneygram/other, it is unikely that you would mix ODL traffic with other money flows - when creating a new wallet is so easy and having different account tags on the exchange can be arranged. From an accounting point of view, it makes things much more straightforward to use certain wallets for certain operations and then summing/plotting/accounting becomes simpler. I should upgrade my scripts to keep 'old' corridors so that if ones come online, but then go dark again, then the summed traffic on those corridors is included. However, I am mostly interested in the main corridors and if old ones did exist and go dark, I would only care if they were big ones with a lot of traffic like the ones we see in the plot above. When I started tracking this back in 2018, I was having a hard time identifying 'corridors' because volumes were small, but now the signal stands out very clearly. And I ran my number (as mentioned in another thread the other day) against the numbers in the ripple q4 report and they %changes between quarters and other details (I forget what now) do match very closely what I see, so I am confident that my plots are basically correct. If I include or miss a handful of transactions, I'm not bothered - I'm interested in the overall trend and the main total. (I also see accounts that appear to be involved in rebalancing these corridors, but I am not including them in these plots). EDIT: One thing I forgot to mention is that the very early USD/PHP traffic last year was comparably speaking larger than the actual traffic taking place now. so it was probably setting up accounts as lots of large transfers that were 'too big' took place, however when real traffic started flowing, I kept the 'corridor' in my list. So, yes - there is almost certainly some spurious data.
  3. Thanks. I don't really have access to a web server that I can upload things like this to. My work has such facilities, but this doesn't really fall into the kind of category I can use for work related material. I don't run a web server from home and I'm not really very good at maintaining that kind of thing. Maybe one of my colleagues has a home server they'd be willing to host plots on...I'll ask.
  4. It's just an upgrade of my existing plots. I wanted to improve the display to show the relative strength of each corridor. It went from this to this and I added other resampling options to give weekly/quarterly/etc sums. I wanted to see if the numbers in the Q4 report matched the figures I get (they do). The final bar in the above plot is today starting at 00:00::00 UTC so it is a small bar as it's only 7am where I am. The weekly plot is going to be fine. we still have friday (usually good) + sat/sun (not good) to add and it looks like this, so this would be the 4th largest week so far - but I suspect we'll get close to #1 by sunday. I suspect that the global shutdown will reduce the amount of money being sent around - so I initially thought ODL will fall - but as only a small % of total traffic is using ODL, we might not see too big an impact. (I plot this data just for my for my own education as I am still waiting for the point where my model of ODL volume / xrp price to kick in).
  5. Only Thursday, but the Weekly (starts Mon on my graphs) ODL volume is holding up well - real shame that the tanking price is not supported at a higher level - this would have been the perfect week for xrp to show how it can resist the pull of BTC (TL;DR - it can't obviously!). We need an announcement from ripple that amazon, uber and apple are going all in on ODL ...
  6. Its an output from one of my ledger analysis scripts (in this case examining ODL activity).
  7. 100m retained for Feb-Mar 2020, 900m returned. No surprises. Unless like you me you are surprised that despite halting exchange based sales, they're still able to shift 100m per month (roughly). 2024-03-02 rp6aTJmW3nq1aKt3Jmuz4DPRxksT5PBjpH 2020-01-01 500.0 rsjFB8mPWqiZgPUaVh8XYqdfa59PE2d5LG 2020-01-01 300.0 2020-02-01 200.0 2024-04-02 rDqGA2GfveHypDguQ1KXrJzYymFZmKxEsF 2020-02-01 200.0 2020-03-01 300.0 rw2hzLZgiQ9q62KCuaTWuFHWfiX7JWg3wY 2020-02-01 500.0 2024-05-02 rDqGA2GfveHypDguQ1KXrJzYymFZmKxEsF 2020-03-01 100.0 rw2hzLZgiQ9q62KCuaTWuFHWfiX7JWg3wY 2020-03-01 500.0 +----------+---------------+------------+ | Month | Re-escrowed | Released | |----------+---------------+------------| | 2018-Feb | 900 | 100 | | 2018-Mar | 900 | 100 | | 2018-Apr | 900 | 100 | | 2018-May | 900 | 100 | | 2018-Jun | 900 | 100 | | 2018-Jul | 900 | 100 | | 2018-Aug | 900 | 100 | | 2018-Sep | 800 | 200 | | 2018-Oct | 800 | 200 | | 2018-Nov | 800 | 200 | | 2018-Dec | 800 | 200 | | 2019-Jan | 800 | 200 | | 2019-Feb | 800 | 200 | | 2019-Mar | 700 | 300 | | 2019-Apr | 700 | 300 | | 2019-May | 700 | 300 | | 2019-Jun | 700 | 300 | | 2019-Jul | 700 | 300 | | 2019-Aug | 800 | 200 | | 2019-Sep | 800 | 200 | | 2019-Oct | 800 | 200 | | 2019-Nov | 900 | 100 | | 2019-Dec | 1000 | 0 | | 2020-Jan | 900 | 100 | | 2020-Feb | 900 | 100 | | 2020-Mar | 900 | 100 | +----------+---------------+------------+ Just for the sake of curiosity, here's a plot of the total xrp in circulation (distribution) from the ripple server. Almost 44billion 'out there' (at least in principle).
  8. Please provide some evidence for this. I would like to see your source. (He has said many times that the buys and sells are matched).
  9. I applaud your sense of decency. It's one thing taking xrp from Mother Theresa ... Ultimately, attacks like this are good for the system in the long run, if Jed is planning on using Stellar to do bank to bank transactions, he'd better understand that this kind of thing can happen on an industrial scale on any exchange (whether it be mt Gox or the DEX).
  10. I do recall a conversation on this forum some time ago about people ready and waiting for Jed's daily sales. If it's them, then I suspect that they decided to try this little experiment out based on the predictability of the daily trades and some analysis of the bid/ask pattern (and based on the memo's it didn't go unnoticed by others). Nice job. I applaud them. Even if it is a little dangerous from the point of view of DEX manipulation (meaning it doesn't give the crypto space a good name/reputation - but let's be honest, the traditional markets are full of this as well).
  11. I get it from the ripple v2 data server and then run some analsysis scripts to extract things I'm interested in. I only download a tiny subset of the ledger as most of the data is rubbish and not useful to me.
  12. That is a pity. I assumed that the moderators of the forum were also maintainers of the code that makes it run and that it might be feasible to add the functionality, but I guess it's a non trivial task. Thanks for responding anyway. I have not reported anyone and I probably won't in future, using the ignore button used to be enough, but I guess I'll just stop reading (and contributing) to the forum soon.
  13. It's called a weekend. It happens quite regularly where I live. If you look at the progress of ODL since the tail end of last year until now, you'll see that it has transformed from a quite noisy signal into a clear pattern, but we are still very early in the evolution of ODL and should expect random fluctuations in traffic to take place from time to time. Growth at this pace is unlikely to be sustainable long term and some flat periods are to be expected. As the graphs I plot (and I believe the utility scan people probably do to) use UTC time zones, the monday data is usually a bit lower than midweek since some data in one timezone will appear during tuesday on these plots. Even now, when Monday is only half done, the volume is higher than most of pre-december.
  14. I disagree. A "group" should be dedicated to a special interest of some kind.. In the general sections open to everyone, I like to skim through the news and general discussion. I also skim the TA threads just to see if anything is predicted to happen in the near future even if I don't believe it to be reliable/factual/useful. Threads such as the AMA with Nicolas are interesting and have been mostly free of problems, but should someone post something offtopic, abusive or otherwise it would be nice to be able to remove them. I do not participate in that group and I am not aware of what is discussed there. I am only asking for "grown up" discussion in the main threads and I should not be forced off the forum by people who do not wish to (or are not capable of) keeping discussion relevant or on topic. The moderators have a hard time keeping up with the abuse on this forum, allowing thread creators to police their own topics would remove the burden of work from them. Off topic posters can post as before and would not be banned from the forum, but only the thread that they are unwelcome on. If the system were abused by the topic creators, then readers would simply move to another thread.
×
×
  • Create New...