This is a mirror of the original BitMEX trollbox archive that used to be online here.
BitMEX disabled their archive after the DDOS attack so I have decided to make my mirror publicly available.

Well now, a few days after I made this mirror available to the public, the original archive from BitMEX is online again.
But since it is still limited (neutered to the last few hundred messages) I will keep my FULL mirror alive.

You have access to more than 50 million trollbox posts. This is more than 7 gigabyte of data hosted on a $5 server so please be patient or consider a donation

<<  >>  English  中文  Русский  한국어  日本語  Español  Français 
[2016-02-07 15:14:19] BitMEX_Sam : Sorry to hear that - you have to watch leveraged positions like a hawk
[2016-02-07 15:14:21] Juba : had a great entry on that breakout
[2016-02-07 15:15:43] Juba : should not be able to do what I just did.. anyways thx for responding
[2016-02-07 15:16:08] BitMEX_Sam : I disagree, this is exactly how the system should work - if your orders execute, you need to work up an offsetting position to get out
[2016-02-07 15:16:32] BitMEX_Sam : Canceling partially executed orders does not remove you from the position.
[2016-02-07 15:17:03] BitMEX_Sam : Sorry you had to learn it the hard way here but this is simply how markets work.
[2016-02-07 15:18:25] Juba : or simply how your platform works :)
[2016-02-07 15:19:25] BitMEX_Sam : Hah, I'm sorry, but I'm not sure what you expect
[2016-02-07 15:19:35] BitMEX_Sam : If you are short, and want to go long, you must close the short before going long.
[2016-02-07 15:20:26] Juba : I tough I did :P to me it looked like it. anyways, hard lesson learned haha
[2016-02-07 15:20:50] Juba : salty lesson to learn right at the break out
[2016-02-07 15:21:11] BitMEX_Sam : Understood - thing to watch is the chart (which shows your current position), or, of course, the Positions pane or the "Your Position" readout on the order controls
[2016-02-07 15:21:20] BitMEX_Sam : Best of luck in the future
[2016-02-07 15:22:20] Juba : hehe thx Sam.
[2016-02-07 15:30:30] sleger : BitMEX_Sam: maybe remove the "Long" and "Short" text from the buttons and leave just Buy / Sell
[2016-02-07 15:31:01] BitMEX_Sam : Have thought about that - we originally added "Short" to the sell button to make it clear that you can sell even when you don't own anything to sell
[2016-02-07 15:31:38] BitMEX_Sam : I think a better solution will be to prioritize the display of your current position
[2016-02-07 15:32:08] BitMEX_Sam : I have a design in our books of a simplified position widget that really makes it obvious where you stand, and we mean to expand it out visually to help you see where you are at a simple glance
[2016-02-07 15:32:18] Juba : BitMEX_Sam: " I think a better solution will be to prioritize the display of your current position" I agree with that :)
[2016-02-07 15:32:43] sleger : but its right under the buy sell buttons already
[2016-02-07 15:34:04] rapidtrades : it says it right there tho
[2016-02-07 15:34:09] rapidtrades : your position
[2016-02-07 15:34:13] rapidtrades : X contract
[2016-02-07 15:34:27] rapidtrades : like wtf plev
[2016-02-07 15:34:48] rapidtrades : *b
[2016-02-07 15:40:11] Tulip_stefan : is there any way to get the charts to show without depositing bitcoins first?
[2016-02-07 15:40:53] rapidtrades : charts show even if ur not logged in
[2016-02-07 15:41:18] rapidtrades : if u need the order controls, try testnet
[2016-02-07 15:41:29] Tulip_stefan : There is a deposit panel in place of where the charts normally are :P
[2016-02-07 15:43:21] BitMEX_Wally : Switch to advanced view
[2016-02-07 15:43:38] BitMEX_Wally : Click on your username in the top right
[2016-02-07 15:44:19] Tulip_stefan : ahh that's be
[2016-02-07 15:44:39] Tulip_stefan : better
[2016-02-07 15:50:20] aethlios : BitMEX_Wally: Hello! Are you planning a 48H contract? When this will happen? any estimate?
[2016-02-07 15:50:58] BitMEX_Wally : Hopefully this quarter
[2016-02-07 16:00:38] Tulip_stefan : Is there anywhere i can get high-res tradeblock index data?
[2016-02-07 16:01:32] Tulip_stefan : i feel their index has a few flaws compared to my own solution, but i can't quantify that without an history of their index.
[2016-02-07 16:02:59] BitMEX_Wally : Tulip_stefan: Send an email to support@bitmex.com and I'll see if we have the data you need
[2016-02-07 16:04:01] Tulip_stefan : I'll do that right away. Tha
[2016-02-07 16:08:53] Tulip_stefan : sent.
[2016-02-07 16:33:49] BitMEX_Wally : Replied :)
[2016-02-07 17:04:39] Tulip_stefan : It's funny how people still use dropbox public folder to share files. I do it all the time, too. Makes me wonder why they ditched that feature for new registers..
[2016-02-07 17:05:08] BitMEX_Sam : Looking forward to the new Keybase FS when it hits public availability
[2016-02-07 17:05:44] Tulip_stefan : yeah i have been a member since a few months. My business partner was in early.
[2016-02-07 17:05:54] Tulip_stefan : it's pretty clumsy to use on windows though...
[2016-02-07 17:06:35] BitMEX_Sam : I have been for a while as well but it appears I wasn't lucky enough to have the feature enabled
[2016-02-07 17:06:55] Tulip_stefan : i haven't looked yer
[2016-02-07 17:07:00] Tulip_stefan : yet*
[2016-02-07 17:08:11] BitMEX_Sam : It's a nice idea. Even if there is, potentially, some vulnerability if someone is able to gain access to user's twitter account or some other social share, in which case they'd be able to intercept a message if you sent it to that account,
[2016-02-07 17:08:16] BitMEX_Sam : It is still far better than the usual practice
[2016-02-07 17:08:28] BitMEX_Sam : I built https://securesha.re/ about three years ago as a half measure for that sort of thing
[2016-02-07 17:08:36] BitMEX_Sam : I still use it now and then to send files to accountants et al
[2016-02-07 17:10:43] Tulip_stefan : that's pretty neat
[2016-02-07 17:43:30] zanza : http://qntra.net/2016/02/bitcoin-mining-difficulty-crosses-144-million-with-20-06-rise/
[2016-02-07 17:55:08] REKT : Liquidated long on `XBT24H`: Sell 3001 @ 377.26
[2016-02-07 19:06:30] REKT : Liquidated short on `ETH7D`: Buy 307 @ 0.00795
[2016-02-07 19:27:10] REKT : Liquidated long on `XBTH16`: Sell 1333 @ 396.04
[2016-02-07 20:09:00] Tulip_stefan : I'm not sure if i understand XBT correctly. If i buy a single contract, i'll receive whatever the tradeblock 30m-index is on the settlement date, minus the settlement fee (0.05%) on the settlement date? And the mark price and fair value are only used to calculate margin calls? If so, is there any other reason than trade fees that the XBT24H price is higher than the tradeblock index plus the settlement fee?
[2016-02-07 20:09:54] Tulip_stefan : I'm a computer scientist, not a trader :P
[2016-02-07 20:12:34] habibi : lets ask our expert, the sleger
[2016-02-07 20:21:38] BitMEX_Wally : Tulip_stefan: https://blog.bitmex.com/bitcoin-futures-basis-trading-lesson-1/
[2016-02-07 20:21:54] BitMEX_Wally : Basically currency futures trade at a premium due to the time value of money
[2016-02-07 20:23:32] Tulip_stefan : I was under the impression that XBT24h contracts had no interest?
[2016-02-07 20:24:20] BitMEX_Wally : Why?
[2016-02-07 20:25:10] Tulip_stefan : I don't see it mentioned on the fees page.
[2016-02-07 20:25:39] BitMEX_Wally : We don't charge interest. But if the future is trading at a premium to spot, this is effectively interest
[2016-02-07 20:26:09] lockhedge : there are other factors that are more important for a premium/discount: inventory/risk management (market maker will quote higher bids if they are short) and XBX as a bad reference price
[2016-02-07 20:26:31] BitMEX_Wally : Plus it is a quanto contract so you have a long USD component
[2016-02-07 20:26:36] BitMEX_Wally : long gamma
[2016-02-07 20:26:45] Tulip_stefan : yeah that makes sense.
[2016-02-07 20:34:51] lockhedge : now we got a discount/negative premium/interest rate...
[2016-02-07 20:35:36] Tulip_stefan : yeah i noticed.
[2016-02-07 20:53:05] lockhedge : Tulip_stefan: that's the index my bot is following, it's open source https://robotfinance.org/bix/ read that you are also developing your own reference index, would be great to learn more about your project
[2016-02-07 20:54:29] Tulip_stefan : I'm not developing my own reference index. I already have a trading bot which has been doing well for some time, internally it uses something i can modify to function as an index which should be better than tradeblock.
[2016-02-07 20:54:37] Tulip_stefan : (hopefully)
[2016-02-07 20:55:34] Tulip_stefan : This is my website: http://tuliphft.com/
[2016-02-07 20:56:57] Tulip_stefan : profit figures are slightly outdated. i think there we have 78 or 79 BTC profit now rather than 35 when we made the page.
[2016-02-07 20:57:06] Tulip_stefan : actually that's more than 'slightly' outdated...
[2016-02-07 20:58:26] sleger : Tulip_stefan: how many licence did you sell ?
[2016-02-07 20:58:37] rapidtrades : u think the current gap with china will close once their new year ends?
[2016-02-07 20:59:40] rapidtrades : even okc usd is $3 higher...seems like a liquidity issue
[2016-02-07 21:00:43] Tulip_stefan : sleger: not much, but i'm more focused on improving the bot than marketing, i don't really care how many licensed i sell, i profit a lot from the bot already.
[2016-02-07 21:01:49] habibi : so basically u saying u are selling money :)
[2016-02-07 21:03:17] Tulip_stefan : If the customers can't get their investment back my product has failed.
[2016-02-07 21:03:22] sleger : i dont really understand the logic for selling it, assuming all is true
[2016-02-07 21:04:01] Tulip_stefan : i don't expect that you do. I'm not selling for the money. I'm selling for the motivation and feedback my customers give.
[2016-02-07 21:04:37] rapidtrades : logic: he needs the mmoney now....or he doesn't think strategy is scalable....or he doesn't believe in it longterm
[2016-02-07 21:04:37] habibi : no offence but firstt impression is "its a scam, watchout, too many red flags"
[2016-02-07 21:04:43] sleger : the website is not super explicit, do you sell an executable or source code ?
[2016-02-07 21:04:50] rapidtrades : the usual reasons why ppl go into money management i assume
[2016-02-07 21:04:52] Tulip_stefan : rapidtrades: i had no clue that okc was higher than the US markets. I don't follow the markets.
[2016-02-07 21:05:01] Tulip_stefan : executable.
[2016-02-07 21:05:35] habibi : Tulip_stefan: please provide some screenshots. anything coz page looks like 5 minutes scam
[2016-02-07 21:05:54] rapidtrades : what's the bot about?
[2016-02-07 21:05:58] Tulip_stefan : Call it a scam if you want. I'm well aware of the state of trading bots, because i call all of them scams as well :P
[2016-02-07 21:06:20] rapidtrades : Tulip_stefan: that's just what a scammer would say :)
[2016-02-07 21:06:56] habibi : "I'm selling for the motivation and feedback my customers give." so i will give as much feedback and motivation as u will need, is that a fair pay ?
[2016-02-07 21:08:19] Tulip_stefan : Just because i value the feedback doesn't mean it's a good idea to give it away for free. If will hurt my profits if i do. It will hurt my profits if i sell as well, but to a much smaller extent.
[2016-02-07 21:08:36] sleger : why not sell the source code as well ?
[2016-02-07 21:08:49] Tulip_stefan : habibi: it's a command line application. No point in providing screenshots.
[2016-02-07 21:09:00] rapidtrades : yeah no source code is a deal breaker for most traders
[2016-02-07 21:09:06] j8 : i think there's a legit place for selling trading bot infrastructure + interface.. selling strategies is a bit scammy though
[2016-02-07 21:09:17] rapidtrades : can't run smth I don't know what it does
[2016-02-07 21:09:19] Tulip_stefan : how much would you pay for the source code :P
[2016-02-07 21:09:50] Tulip_stefan : 10 BTC? not enough. 100 BTC? maybe, but who's going to pay that?
[2016-02-07 21:09:54] sleger : provide lots of feedback of course
[2016-02-07 21:10:24] rapidtrades : i would pay the same as for a bot with no source....it's a requirement for me, u don't get a bonus
[2016-02-07 21:10:46] micmix : Tulip_stefan: what's the trading profile of your bot? taker/maker, # of trades per day etc.?
[2016-02-07 21:11:12] Tulip_stefan : rapidtrades: then you're not our target audience :P
[2016-02-07 21:11:49] j8 : no one will pay for open source either unless you can make them somehow
[2016-02-07 21:11:51] rapidtrades : while we're on this subject....I need some arb bots coded...where should I go and how much should I expect to pay
[2016-02-07 21:11:53] Tulip_stefan : i trade around once or twice my portfolio per day. I'd rather not give any specifics over the strategy
[2016-02-07 21:12:38] rapidtrades : Tulip_stefan: that sounds awful
[2016-02-07 21:12:41] sleger : micmix: could you do dev for me? (java)
[2016-02-07 21:13:00] habibi : Tulip_stefan: so the app just follows ur trades? or its bot...?
[2016-02-07 21:13:00] rapidtrades : so it's not making markets...nor is it catching bigger moves on longer-term charts
[2016-02-07 21:13:13] rapidtrades : seems like the worst timeframe to choose for bitcoin
[2016-02-07 21:13:22] micmix : sorry sleger, have zero spare time :-(
[2016-02-07 21:13:52] sleger : Tulip_stefan: can it double down, ie say you're long 1btc, it loses so next day it's long 2, and if it loses then next day long more etc... ?
[2016-02-07 21:14:06] Tulip_stefan : Sorry, i only do C++ or python, unless you want your bot coded in assembly or rust.
[2016-02-07 21:14:55] sleger : micmix: would only be some fix gateway in java, probably couple of hours for you ;
[2016-02-07 21:15:09] Tulip_stefan : sleger: my bot always goes all-in in either direction. I don't win every day, but averaged over a month it's usually profitable.
[2016-02-07 21:15:47] Tulip_stefan : doubling down is bullshit. I'm a computer scientist, not a trader. I only make moves that make sense from a statistical standpoint.
[2016-02-07 21:15:49] sleger : oh ok so daily output is : long flat or short. ok
[2016-02-07 21:16:49] sleger : Tulip_stefan: well people who do what you want to do dont sell their algo, they manage capital and take % of returns
[2016-02-07 21:17:08] micmix : sleger: I didn't do any Java in the last 5+ years, will take more than a few hours unfortunately :-(
[2016-02-07 21:17:27] sleger : micmix: oh sorry i thought you coded everything in java
[2016-02-07 21:17:42] Tulip_stefan : That's a good point, and we have thought about that. But if we do that, we'll have a lot of liabilities we want to avoid.
[2016-02-07 21:17:52] rapidtrades : sleger: it's a bit hard to collect money for bitcoin trading tho :) has anyone done it yet?
[2016-02-07 21:18:12] micmix : no, C# and javascript @sleger
[2016-02-07 21:18:14] lockhedge : sleger: do you need a FIX api wrapper for OKC?
[2016-02-07 21:18:22] sleger : lockhedge: yes
[2016-02-07 21:18:59] lockhedge : sleger: i was searching for python, only found java... https://github.com/sutra/okcoin-client
[2016-02-07 21:19:00] sleger : i have some java code I took from their repo, manage to get market data, construct order book, submit an order but not cxl for instance
[2016-02-07 21:19:56] sleger : lockhedge: i saw that one, but not much more than on 0kc github
[2016-02-07 21:20:25] micmix : tscha is a java guy, our system is C#/F# and javascript
[2016-02-07 21:20:46] sleger : micmix: i thought JS was similar enough to java :p
[2016-02-07 21:21:05] Tulip_stefan : java is more akin to C# than js...
[2016-02-07 21:21:17] rapidtrades : is there a lot of work for a python programer?
[2016-02-07 21:21:36] micmix : just the first 4 letters of the name are the same, everything else is different
[2016-02-07 21:21:41] rapidtrades : if one focuses on trading bots/apps whatever
[2016-02-07 21:21:41] sleger : micmix: well TBH, I am looking for 0kc fix gateway running as standalone on linux (so everything but c#) and communicating with my algo via socket, so could be almost any language
[2016-02-07 21:22:12] sleger : rapidtrades: python is the easiest
[2016-02-07 21:23:39] Tulip_stefan : rapidtrades: i wouldn't know. I only deal with trading in my spare time. I guess that python/R skill are in demand for algo trading firms on research positions and C-family stuff for the actual implementation.
[2016-02-07 21:24:04] REKT : Liquidated short on `XBT24H`: Buy 250 @ 375.26
[2016-02-07 21:24:41] Tulip_stefan : I actually hate python but i still prototype a lot of stuff in it because the ecosystem is so good.
[2016-02-07 21:25:17] rapidtrades : idk if the time investment to learn pyhton is worth it...prolly is
[2016-02-07 21:25:34] rapidtrades : most of the stuff I need coded is fairly simple
[2016-02-07 21:25:34] micmix : sleger: sorry, I could've made something in C# in 4-6 hours but not in java
[2016-02-07 21:25:36] sleger : rapidtrades: learn python VS doing what ?
[2016-02-07 21:25:43] Tulip_stefan : what do you want to use it for, and what would you use otherwise?
[2016-02-07 21:25:49] rapidtrades : vs pay smone @sleger
[2016-02-07 21:26:11] j8 : yup it's great for prototyping, not used a lot for professional development. it's probably all you need for bitcoin trading
[2016-02-07 21:26:26] sleger : rapidtrades: algo to trade on one exchange or many?
[2016-02-07 21:26:28] Strobelov : Our traders and devs use Python for stuff like backtesting. Nothing performance sensitive.
[2016-02-07 21:27:21] Tulip_stefan : rapidtrades: my first trading bot was written in python, with 2 weeks of prior experience in python. It took 2 days to write and turned 0.1 LTC into 1 BTC in slightly less than 2 weeks. You don't lear
[2016-02-07 21:27:23] rapidtrades : sleger: for bitcoin i need an arb bot so across sites...but I would turn it on/off tho so nothing too complex re:execution
[2016-02-07 21:27:38] Tulip_stefan : you don't learn python and use it for something, you learn it wile coding :P
[2016-02-07 21:27:59] tscha : Tulip_stefan: i guess that applies to every language
[2016-02-07 21:28:41] Tulip_stefan : Strobelov: backtesting not performance sensitive.. haha. My backtesting engine is written in C++ because backtesting in python would take days.
[2016-02-07 21:28:47] sleger : rapidtrades: having something done well you would probably have to pay ~4weeks of dev I would say
[2016-02-07 21:29:30] sleger : Tulip_stefan: why is that, python can be really fast if coded well, almost all cpu intensive functions are coded in c++ anyways in python
[2016-02-07 21:30:22] Tulip_stefan : python can be really fast if you can use numpy. The problem is, you probably can't unless you're solving a textbook math problem.
[2016-02-07 21:31:36] Tulip_stefan : since i'm a wizzard with SIMD and multi threading, the difference between cpython and C/C++ is about 100-1000x, and C/C++ has the advantage of being much more memory efficient.
[2016-02-07 21:32:09] Tulip_stefan : I can't even load a quarter of my dataset in python before i run out of memory.
[2016-02-07 21:32:18] tscha : Am I the only one not backtesting? I don't see much sense in it for Bitcoin markets since they are so illiquid and thin, so if you trade, you will have a noticable effect on books and price
[2016-02-07 21:32:23] tscha : might work for micro-volumes though
[2016-02-07 21:32:38] sleger : he is trading daily
[2016-02-07 21:33:02] sleger : but yes you're the only one not backtesting
[2016-02-07 21:33:07] Strobelov : Well, when I mentioned backtesting, I wasn't referring to Bitcoin.
[2016-02-07 21:33:23] Tulip_stefan : I'm mainly backtesting for research and 'these settings seem to work fine, lets create 4 accounts with approximately these settings and see what they do'
[2016-02-07 21:33:51] rapidtrades : so for backtesting in python...CPU matters? not GPU?
[2016-02-07 21:34:33] Tulip_stefan : That depends on what you're backtesting. If you're using machine learning, you might be able to use the GPU, but i haven't tried that yet.
[2016-02-07 21:34:35] Strobelov : Can you even do GPU programming in Python?
[2016-02-07 21:34:43] laisee : standard python code is optimxed for CPU, not GPU
[2016-02-07 21:35:02] rapidtrades : guess I could rent out a cloud for peanuts to do my testing?
[2016-02-07 21:35:06] Tulip_stefan : There are libraries that let you write opencl shaders for python.
[2016-02-07 21:35:12] sleger : yes you can use gpu in python
[2016-02-07 21:35:13] Tulip_stefan : opencl kernels*
[2016-02-07 21:35:22] micmix : tscha: backtesting in bitcoin is more useful for taker algos but you can still do it for maker
[2016-02-07 21:35:35] Strobelov : ah, cool
[2016-02-07 21:36:46] tscha : micmix: you use backtesting when evaluating strategies for your makers?
[2016-02-07 21:38:28] micmix : tscha: yes, and it's very useful. of course you cannot fully simulate the effects of your maker on other bots
[2016-02-07 21:38:53] tscha : interesting
[2016-02-07 21:39:07] rapidtrades : so f interesting
[2016-02-07 21:39:18] rapidtrades : PASS THE GUN PLZ
[2016-02-07 21:39:37] micmix : still backtesting and simulation gets you close enough
[2016-02-07 21:39:57] Tulip_stefan : I used to trust them blindly, but nowdays i use the results of my backtests as a starting point and run multiple accounts to refine my parameters further.
[2016-02-07 21:40:49] tscha : micmix: but with backtesting, you mean backtesting on historical orderbook data right?
[2016-02-07 21:41:02] micmix : yes, on full orderbook history
[2016-02-07 21:41:32] Tulip_stefan : i only test on the tick data but having orderbook data as well would be nice. That depends on the kind of information your bot needs.
[2016-02-07 21:42:12] tscha : Tulip_stefan: without book data, you don't know how much you could have bought at what price
[2016-02-07 21:42:31] BitMEX_Sam : Hard to get that data
[2016-02-07 21:42:37] BitMEX_Sam : At least in Bitcoin
[2016-02-07 21:42:46] tscha : tradeblock provides it
[2016-02-07 21:42:55] rapidtrades : yeah tradeblock
[2016-02-07 21:42:58] micmix : I run cryptoiq.io on the side, have full orderbook history
[2016-02-07 21:43:14] micmix : my small startup :-)
[2016-02-07 21:43:25] lockhedge : i don't do any backtesting, at least as maker you always change the market (depending on your size) so backtesting does not really work
[2016-02-07 21:43:59] Tulip_stefan : How are your guys' profit figures?
[2016-02-07 21:44:07] rapidtrades : ++++++++++++++++++++++++
[2016-02-07 21:44:32] Tulip_stefan : for obvious reasons, i'm the only one stupid enough to say them out loud. :P
[2016-02-07 21:45:10] micmix : if anyone need the data, send an email to info@cryptoiq.io. I will give you top 50 asks/bids every 5 seconds for free, most exchanges available (not BitMEX yet)
[2016-02-07 21:45:16] belcher : i used to have profits but then i lost them in a tragic boating accident
[2016-02-07 21:45:43] Tulip_stefan : interesting. Tell me more.
[2016-02-07 21:45:46] micmix : * one year of orderbook history
[2016-02-07 21:45:51] tscha : micmix: cool thanks. is 5 sec the way you collected it or just the "free version"? :)
[2016-02-07 21:46:43] Strobelov : a lot can happen in 5 seconds...or 5 microseconds...though I suppose that's less of a problem with Bitcoin :)
[2016-02-07 21:46:47] micmix : I collect full history, multiple updates per second
[2016-02-07 21:47:06] rapidtrades : micmix: where do u store that
[2016-02-07 21:47:39] sleger : hard drive ....
[2016-02-07 21:47:46] tscha : nice... cause tradeblock only has 1 second snapshots... which is sufficient 99% of the time, but the 1% where the market moves quickly is the most important time unfortunately
[2016-02-07 21:47:57] lockhedge : micmix: do you still collect data from BIX via REST? i would need a backup for my historical data :)
[2016-02-07 21:48:08] micmix : basically full stream if websocket/FIX is available or polling every 1-2 seconds if REST
[2016-02-07 21:48:47] lockhedge : micmix: are you using kdb?
[2016-02-07 21:49:21] micmix : lockhedge: sorry, not BIX. have tradeblock XBX though :-)
[2016-02-07 21:51:02] lockhedge : micmix: what db are you using to store all this data?
[2016-02-07 21:51:24] BitMEX_Sam : mongodb right?
[2016-02-07 21:51:50] Tulip_stefan : I was interested in that as well. My business partner once tried to store orderbook data but i think he underestimated the vast scale of such data....
[2016-02-07 21:51:55] BitMEX_Sam : :smile: :gun:
[2016-02-07 21:52:25] BitMEX_Sam : KDB's tables compress well on disk
[2016-02-07 21:53:04] micmix : mongodb was too slow :-( storing data in custom binary format optimized for speed as binary blobs to cloud storage. then process, clean the data and convert to json
[2016-02-07 21:53:27] Tulip_stefan : i use cvs's for anything that requires interop with other tools and convert them into memory mapped format for fast access from c++.
[2016-02-07 21:53:52] BitMEX_Sam : BTW was completely kidding about mongo, probably the worst possible tool for this job
[2016-02-07 21:54:01] micmix : json blobs in cloud storage as well, so the API just gives a signed URL. super fast
[2016-02-07 21:55:01] micmix : yeah, mongo was the first thing I tried. would be much easier to use mongo
[2016-02-07 21:56:53] BitMEX_Sam : Doesn't make sense to imo, tick data is hugely regular - and mongo doesn't care much about fast querying, efficient disk storage or durability
[2016-02-07 21:57:17] Strobelov : ugh, mongo. It's been somewhat usable since v3.
[2016-02-07 21:57:56] BitMEX_Sam : I'd give that a very qualified 'somewhat'
[2016-02-07 21:57:58] BitMEX_Sam : :)
[2016-02-07 21:58:27] Strobelov : I use it for storing audit data for our controls. Nothing like market data—that would be absurd.
[2016-02-07 21:58:41] micmix : https://cryptoiq.io/api/marketdata/orderbooktop/{exchange}/{market}/{date}/{hour} if you want to give it a try. top 20 every 10 sec with no authorization
[2016-02-07 21:59:10] micmix : https://cryptoiq.io/api/marketdata/orderbooktop/bitfinex/btcusd/2015-01-20/22
[2016-02-07 21:59:20] micmix : for example
[2016-02-07 22:01:21] micmix : mongo can be used to store market data if storing big chunks like one hour but it's far from optimal of course
[2016-02-07 22:03:35] rapidtrades : wait a min...finex deposits are also restricted due to new year right
[2016-02-07 22:06:49] j8 : rapidtrades: yeah no USD deposits until the 15th
[2016-02-07 22:08:04] rapidtrades : maybe that's why its at 372
[2016-02-07 22:08:08] j8 : mhmm
[2016-02-07 22:10:32] sleger : rapidtrades: but its same for 0kcoin usd so doesnt explain
[2016-02-07 22:11:05] habibi : lately some big guy was closing his position, just look at usd swap and compare it with that 380+ drop to 370
[2016-02-07 22:12:39] lockhedge : interesting https://bfxdata.com/sentiment/longshort
[2016-02-07 22:13:09] rapidtrades : sleger: it does explain...have u tried shorting on okc usd? 0.1% interest
[2016-02-07 22:13:47] sleger : nothing to do with deposits being delayed
[2016-02-07 22:14:17] rapidtrades : it's china...they're not trading right now
[2016-02-07 22:14:46] rapidtrades : remember Christmas when they dumped the market on us but we didnt move?
[2016-02-07 22:14:54] rapidtrades : now it's the reverse
[2016-02-07 22:21:43] lockhedge : sleger: if you can't deposit fresh fiat money into your exchange account and want to go long, you have to borrow funds, when interest rates become too high nobody is going long anymore...
[2016-02-07 22:25:50] tscha : lockhedge: the sentiment dashboard is pretty useless unfortunately
[2016-02-07 22:26:03] Tulip_stefan : swap rates are not higher than usual?
[2016-02-07 22:26:37] lockhedge : tscha: it's useless if you try to predict future market movements
[2016-02-07 22:28:12] tscha : no it's useless because it's missleading
[2016-02-07 22:29:06] sleger : lockhedge: yes but it is true for both finex and 0kc usd, but there is 5-6$ difference, why ?
[2016-02-07 22:29:21] lockhedge : Tulip_stefan: depends on what's your reference for "usual" - does supply/demand of swaps are always the same or do they correlate with volume/volatility?
[2016-02-07 22:30:43] Tulip_stefan : i was just looking at the swap rate charts from bfxdata. By eyeballing the historical swap rate it doesn't appear to be higher than average.
[2016-02-07 22:31:09] Tulip_stefan : i don't do business on bitfinex, so i wouldn't know what their usual characteristics are.
[2016-02-07 22:31:50] lockhedge : sleger: there seems to be some arb working between okc cny and okc usd, i guess. but arb between finex and okc is frozen due to bank holidays.
[2016-02-07 22:33:29] tscha : this bfx dashboard should should show opened swaps in the specified period, which implies people went long/short actively according to the ratios presented during this period. It's wrong though, cause it also includes renewed swaps, so even if people just renew their expired swaps (or even decrease their position by opening a smaller long than they had before), it shows up in there with the full new position.
[2016-02-07 22:34:09] sleger : yes but you can look at cum open swaps
[2016-02-07 22:34:20] lockhedge : tscha: ok, thanks for the explaination
[2016-02-07 22:34:23] sleger : you dont have to look at the wrong chart
[2016-02-07 22:37:09] tscha : i am saying that there ARE wrong charts in there, which shouldn't be there in the first place
[2016-02-07 22:37:16] tscha : they don't improve by not looking at them
[2016-02-07 22:37:39] tscha : and most people don't know about it
[2016-02-07 22:38:38] tscha : and with "chart" i actually mean those four 180°-pie-charts
[2016-02-07 22:39:06] sleger : 180 or 360 ?
[2016-02-07 22:39:14] sleger : ...
[2016-02-07 22:39:57] tscha : π
[2016-02-07 22:40:13] Tulip_stefan : a pi-chart
[2016-02-07 22:40:21] tscha : Tulip_stefan: wow... mind-blown
[2016-02-07 22:40:43] tscha : comment of the day and good opportunity to say goodnight
[2016-02-07 22:40:47] tscha : cya tomorrow
[2016-02-07 22:41:07] tscha : pi-chart.. nice one
[2016-02-07 22:45:40] Tulip_stefan : Not sure if that was sarcastic or not. It seemed really obvious to me.
[2016-02-07 23:24:39] Tulip_stefan : bah i have NAN's somewhere in my plotting code...
[2016-02-08 00:19:09] REKT : Liquidated long on `ETH7D`: Sell 95 @ 0.00769
[2016-02-08 00:42:19] zanza : XBU volume is 0 :)
[2016-02-08 00:49:30] REKT : Liquidated short on `ETH7D`: Buy 500 @ 0.00831
[2016-02-08 00:58:41] habibi : hm does ui lags only for me after sitting here for a longer time?
[2016-02-08 01:02:19] BitMEX_Sam : Should be much better recently than before, I spent the better part of a week optimizing the app
[2016-02-08 01:02:28] BitMEX_Sam : I keep it open all day though too - I'll run a profile right now
[2016-02-08 01:02:54] BitMEX_Sam : More likely than not, the chart is leaking memory
[2016-02-08 01:03:35] habibi : i recently started to have strange lags on bitmex with ui and with tables as well, more i put buy order at diffrent price and it went to book on diffrent but seems like its something bad on my side
[2016-02-08 01:04:05] BitMEX_Sam : Do you use the Basic or Advanced UI?
[2016-02-08 01:04:08] habibi : basic
[2016-02-08 01:04:41] BitMEX_Sam : Should be fine... I'm reading completely normal memory usage and I've had the app open for 12hrs
[2016-02-08 01:05:16] habibi : yep, sometimes recent trades tabbles goes from eth and stays on 24h fe
[2016-02-08 01:05:36] BitMEX_Sam : Ah that is something I've seen before
[2016-02-08 01:05:48] BitMEX_Sam : Testnet has a fix for it... there are ways to switch contracts in such a way that will confuse the app
[2016-02-08 01:05:57] habibi : but the thing with completely diffrent price order is strange :p
[2016-02-08 01:06:17] BitMEX_Sam : Yeah that is bad if it is on our end - can you explain what you saw?
[2016-02-08 01:06:33] JMG99 : where is the liquidity on eth7d, look at him who shorted at 0.00840 and covered at 0.00879.
[2016-02-08 01:07:07] habibi : JMG99: that was mine exacly, wanted to put buy order at 0.0082 and it went to the 879 exacly
[2016-02-08 01:07:16] habibi : BitMEX_Sam: ^^
[2016-02-08 01:07:49] habibi : but still it shows normal data on order history which tells its sth on my end
[2016-02-08 01:08:34] BitMEX_Sam : So you're saying you put a buy in at 0.00820 but you saw it in the book as 0.00879?
[2016-02-08 01:08:40] BitMEX_Sam : Or just that there was low liquidity at the time?
[2016-02-08 01:09:14] habibi : i had at box 00820 and it instantly hit sell book
[2016-02-08 01:09:40] j8 : ^^ it's happened to me before. order submits at a different price than is in the order controls
[2016-02-08 01:09:54] habibi : j8: and i lost 0.2 btc due to that :p
[2016-02-08 01:10:10] BitMEX_Sam : So you submitted at 0.00820 but it executed at 0.00879?
[2016-02-08 01:10:15] habibi : right
[2016-02-08 01:10:43] BitMEX_Sam : Hmm. Let me do some investigation, that of course shouldn't be possible
[2016-02-08 01:11:03] BitMEX_Sam : Was this today?
[2016-02-08 01:11:28] sleger : i think that happens if you enter an order switch tab and back and then press the buy/sell button. If you stay on the same tab I have never seen that
[2016-02-08 01:11:31] j8 : it's like the change in the UI doesn't get passed on to the app in certain circumstances, i haven't seen it today but occasionally previously
[2016-02-08 01:11:32] habibi : 2016-02-08 01:54:24, so not more than 20 minute
[2016-02-08 01:11:34] habibi : ag
[2016-02-08 01:11:45] j8 : usually it makes me question my sanity
[2016-02-08 01:12:17] habibi : j8: thats what i was thinking :D but the problem was i was sure about th
[2016-02-08 01:13:04] habibi : sleger yea that could
[2016-02-08 01:13:14] habibi : but box was with 00820
[2016-02-08 01:17:04] sleger : oh ok, if you switch tab, box price would change. All my orders were always submitted to the price shown in the box though
[2016-02-08 01:19:19] habibi : sleger: quantity is changing but prices remains the old one
[2016-02-08 01:19:21] habibi : on my case
[2016-02-08 01:19:35] habibi : ah nvm quantity is the same as well
[2016-02-08 01:19:43] sleger : price changes, not qty
[2016-02-08 01:19:49] sleger : dont drink and trade !
[2016-02-08 01:19:59] habibi : i mean old price
[2016-02-08 01:20:29] habibi : so price dont move from eth to 24h in box
[2016-02-08 01:20:54] sleger : does for me
[2016-02-08 01:21:26] habibi : so when u move from 24h to eth u have 377 price on eth box?
[2016-02-08 01:22:32] sleger : it switches between 0.08 and 377 yes
[2016-02-08 01:22:58] sleger : 0.00829 and 377.23
[2016-02-08 01:24:11] habibi : yep, thats my point, probally bad explained
[2016-02-08 01:24:37] BitMEX_Sam : That's intentional, the price moves to the lowest ask when you switch contracts
[2016-02-08 01:26:18] habibi : yea thats ok, missunderstanding
[2016-02-08 01:33:39] sleger : BitMEX_Sam: interested in doing side dev : 0kc fix gateway ?
[2016-02-08 01:34:01] BitMEX_Sam : sleger: We have a pretty big release upcoming that's going to take up all my time, but feel free to send an email
[2016-02-08 01:34:03] BitMEX_Sam : sam@bitmex
[2016-02-08 02:36:44] rapidtrades : fellas?
[2016-02-08 02:43:09] j8 : whats up fella
[2016-02-08 04:26:33] rapidtrades : omg can't believe that finally caught on
[2016-02-08 04:26:46] rapidtrades : this could be my shot for e-fame
[2016-02-08 04:27:02] rapidtrades : bringing the fella back
[2016-02-08 04:30:57] j8 : you're already famous here
[2016-02-08 04:31:44] rapidtrades : tnx fella :)
[2016-02-08 05:08:50] laisee : good fellas?
[2016-02-08 05:28:39] zanza : did rapidtrades invent that word?
[2016-02-08 05:37:55] laisee : a literary superstar, no less
[2016-02-08 06:10:57] rapidtrades : zanza: I didn't invent but I'm bringing it back...hence the e-fame
[2016-02-08 06:11:08] rapidtrades : I'm a trend starter
[2016-02-08 06:18:34] rapidtrades : is REKT bot out?
[2016-02-08 06:18:39] rapidtrades : was that a liq
[2016-02-08 06:18:43] rapidtrades : 30k
[2016-02-08 06:23:03] habibi : nah just from negative premium to positive with same crazy trader
[2016-02-08 06:23:11] habibi : let's called it "new guy"
[2016-02-08 06:23:37] rapidtrades : sledger can be pretty crazy
[2016-02-08 09:25:26] tscha : morning fellas
[2016-02-08 09:39:59] laisee : hey, fella
[2016-02-08 11:30:59] tscha : rapid asleep?
[2016-02-08 11:31:14] tscha : oh I forgot, middle of the night in the US
[2016-02-08 11:44:25] slavo : doom soon ^^
[2016-02-08 11:45:57] rapidtrades : tscha: morning fella
[2016-02-08 11:46:44] tscha : rapidtrades: morning! so eastcoast?
[2016-02-08 11:46:59] tscha : :)
[2016-02-08 11:47:20] chromaticcr : close 'em positionz
[2016-02-08 11:49:57] slavo : stop lying its afternoon that's all
[2016-02-08 11:49:59] slavo : :D
[2016-02-08 11:50:09] slavo : haaaaa traders
[2016-02-08 11:50:44] chromaticcr : For traders, anytime that money shines is morning
[2016-02-08 11:51:27] slavo : if bitcoin rally i'll be rekt so please don't fomo
[2016-02-08 11:52:43] slavo : have like 100btc in shorts spread accross exchanges; i think biggest position ever
[2016-02-08 11:52:58] slavo : smells like rekt then :D
[2016-02-08 11:55:01] slavo : sub 300 anyone? or am i alone dreaming?
[2016-02-08 11:55:08] slavo : i'm so drunk tho
[2016-02-08 11:55:13] slavo : proof: i'm talking on bitmex chat
[2016-02-08 11:55:14] slavo : :D
[2016-02-08 11:55:35] chromaticcr : i hope for that, so I could roll more longs in
[2016-02-08 11:55:52] chromaticcr : not sure about sub 300 thou
[2016-02-08 11:57:14] rapidtrades : another boring settlement
[2016-02-08 11:58:35] chromaticcr : +1 to rapid....
[2016-02-08 12:00:00] slavo : i think no rally before march; i'll hodl shorts until late feb
[2016-02-08 12:00:06] chromaticcr : no more bets!
[2016-02-08 12:00:29] slavo : as noob says: mark my words lolol
[2016-02-08 12:01:26] slavo : eth is crazy; i think bobsurplus is pwning wallstreet atm
[2016-02-08 12:01:31] slavo : and mcrsft
[2016-02-08 12:07:12] slavo : why is there so few MM over here?
[2016-02-08 12:28:30] chromaticcr : lol wtf bro eth
[2016-02-08 12:32:58] slavo : yeah i dumped waaaay to early
[2016-02-08 12:33:11] slavo : whatever... profit is profit
[2016-02-08 12:45:29] rapidtrades : just made $90 on a mistake...feels good
[2016-02-08 12:45:51] rapidtrades : took slightly higher risk by mistake...closed excess for profit now
[2016-02-08 12:47:22] tscha : rapidtrades: probably wasn't a mistake but your trading intuition
[2016-02-08 12:47:35] tscha : subconsciously
[2016-02-08 12:48:01] tscha : happens to many pro traders, they have no explanation sometimes
[2016-02-08 12:51:17] rapidtrades : you give me too much credit fella :)
[2016-02-08 12:52:06] tscha : you are too humble fella
[2016-02-08 13:01:21] slavo : taking too much risk is always a mistake :D
[2016-02-08 13:02:53] slavo : i like bitmex but that interface is messy + market is illiquid what a pity
[2016-02-08 13:03:06] slavo : might be good to run small bots
[2016-02-08 13:05:21] BitMEX_Sam : slavo: We have an interface redesign in beta at https://testnet.bitmex.com/app/
[2016-02-08 13:05:29] BitMEX_Sam : If you're at all interested would be happy to hear your thoughts
[2016-02-08 13:05:57] slavo : checking it
[2016-02-08 13:05:58] slavo : ty
[2016-02-08 13:06:08] slavo : i like your work don't hear me wrong
[2016-02-08 13:06:21] slavo : but some liquidity injection would help a lot
[2016-02-08 13:06:33] slavo : any idea why there's so little volume?
[2016-02-08 13:06:38] BitMEX_Sam : Agreed - it improves day by day as more traders come online, we're also working on new MM agreements
[2016-02-08 13:06:56] BitMEX_Sam : To be honest we're usually in the top 3-4 non-CN exchanges, the volume is there, it's just mostly concentrated on 24H
[2016-02-08 13:08:01] slavo : yeah i'm a position trader so it's annoying :)
[2016-02-08 13:08:47] BitMEX_Sam : Yeah, understood. All I can say is that it will get better - we're working hard to get more traders online every day
[2016-02-08 13:14:53] slavo : if you want feedback... those trading view charts are lame; i have to connect to trading view or open cryptowat.ch to chart
[2016-02-08 13:15:11] slavo : maybe you could do something like okcoin did
[2016-02-08 13:15:20] slavo : haha okcoin
[2016-02-08 13:18:15] BitMEX_Sam : What's the issue with the charts?
[2016-02-08 13:20:35] slavo : timeframes
[2016-02-08 13:21:24] BitMEX_Sam : As in...?
[2016-02-08 13:21:39] BitMEX_Sam : You know it's adjustable on the top right, you can zoom out, or you can fullscreen the chart?
[2016-02-08 13:26:43] slavo : yeah but i only have 1h 6hr etc
[2016-02-08 13:28:10] slavo : i use W D 4hr 1hr usually; or did i miss something? + trading view data is lagging
[2016-02-08 13:29:24] BitMEX_Sam : There's two different things going on here,
[2016-02-08 13:29:40] slavo : having cryptowat.Ch chart would be better for me. no lag and every tieframe on bitmex directly for fast checking
[2016-02-08 13:29:44] BitMEX_Sam : One is that you choose an interval on the top left, and it doesn't really make sense to choose anything more coarse-grained than 30m (this is the candle size)
[2016-02-08 13:30:05] BitMEX_Sam : The bottom has pre-programmed timeframes you can click, or you can just zoom it out
[2016-02-08 13:30:18] BitMEX_Sam : The chart here on the site will have less latency than cryptowatch or any other external provider
[2016-02-08 13:32:37] tscha : new MM agreements? I don't think the books are so thin, it's more that we have not much takers
[2016-02-08 13:36:03] tscha : are you thinking of a compensation depending on % of time quoting in a certain range?
[2016-02-08 13:37:10] BitMEX_Arthur : tscha: Well we already offer rebates
[2016-02-08 13:37:40] rapidtrades : BitMEX_Arthur: lower fees
[2016-02-08 13:37:42] tscha : BitMEX_Arthur: Yes, I think that compensation is fair.
[2016-02-08 13:37:53] BitMEX_Arthur : It's the Lunar New Year Bitcoin won't be fun until next monday
[2016-02-08 13:38:15] BitMEX_Arthur : China is shut and you can't deposit USD on OKCoin or Bitfinex
[2016-02-08 13:38:21] tscha : next monday is going to be interesting though, a lot of money hitting exchanges
[2016-02-08 13:38:48] rapidtrades : prices can still move...its only the plebs that can't move money around
[2016-02-08 13:39:22] BitMEX_Arthur : Yeah but when all the traders are on holiday in China
[2016-02-08 13:39:30] BitMEX_Arthur : you take away 70% of the market
[2016-02-08 13:39:58] BitMEX_Arthur : And non Chinese can't even access the most liquid USD exchanges with new money
[2016-02-08 13:40:28] rapidtrades : BitMEX_Arthur: is volume down 70% tho
[2016-02-08 13:41:24] rapidtrades : i'm a trader and local holidays don't mean much....bitcoin is traded online, not like forex where u need the banks to quote the market
[2016-02-08 13:41:58] BitMEX_Arthur : Yeah but you still need price action
[2016-02-08 13:43:22] rapidtrades : it's like the Chinese are partying all day :)
[2016-02-08 13:43:40] REKT : Liquidated short on `ETH7D`: Buy 455 @ 0.00830
[2016-02-08 13:44:08] rapidtrades : gap okc/finex down to $1 now...suggesting liquidity is picking up maybe?
[2016-02-08 13:45:43] tscha : just to clarify, by "thinking about new MM agreements" you mean getting new market makers on board, and not changing the compensation scheme, right?
[2016-02-08 13:47:00] BitMEX_Arthur : We are always open to suggestions, but we will be focusing energy on bringing in more market makers, one thing we are working on is a XBT48H which will allow MM to roll their positions
[2016-02-08 13:47:09] BitMEX_Arthur : That should help liquidity quite a bit
[2016-02-08 13:49:27] tscha : BitMEX_Arthur: yeah, sounds good
[2016-02-08 14:07:06] 4697 : wow someone just ate a 15k eth sell wall on polo
[2016-02-08 14:38:42] REKT : Liquidated long on `XBT7D`: Sell 503 @ 375.10
[2016-02-08 14:39:42] REKT : Liquidated long on `XBT7D`: Sell 4000 @ 375.02
[2016-02-08 18:11:10] tscha : Volume since settlement 40k contracts... hopefully it's not the whole week like that
[2016-02-08 18:40:00] BitMEX_Sam : New feature just hit - you can now cancel a level of order by clicking on the orderbook
[2016-02-08 18:40:05] BitMEX_Sam : Hover over a highlighted row to see the cancel button.
[2016-02-08 18:41:01] BitMEX_Sam : This is part of the UI work being done for the redesign, but it was self-contained enough to easily promote here
[2016-02-08 18:48:18] Tulip_stefan : my plot tool crashed due to a floating point underflow exception.... i didn't even knew such a thing existed.
[2016-02-08 18:50:42] tscha : Tulip_stefan: you sure it wasn't this error: https://www.youtube.com/watch?v=bLHL75H_VEM&feature=share ?
[2016-02-08 18:51:33] Tulip_stefan : yeah i'm sure. The culprit was the expression std::exp(-1.250001797e+011);
[2016-02-08 18:52:57] BitMEX_Sam : Yeah, that'd be a really tiny number
[2016-02-08 18:53:01] BitMEX_Sam : You don't see underflows too often though
[2016-02-08 19:19:50] REKT : Liquidated long on `XBT24H`: Sell 1250 @ 373.38
[2016-02-08 19:22:10] REKT : Liquidated long on `XBT24H`: Sell 189 @ 372.43
[2016-02-08 19:22:10] REKT : Liquidated long on `XBT24H`: Sell 1688 @ 372.53
[2016-02-08 19:22:40] REKT : Liquidated long on `XBT7D`: Sell 1 @ 374.22 :punch: :whale:
[2016-02-08 19:22:40] REKT : Liquidated long on `XBT7D`: Sell 40 @ 374.72
[2016-02-08 19:22:40] REKT : HODL HODL HODL!
[2016-02-08 19:27:45] Juba : lol
[2016-02-08 19:32:51] Tulip_stefan : that whale unicode character is a nice touch.
[2016-02-08 19:35:28] laisee : its a very, very small whale ... ;-)
[2016-02-08 20:08:31] tscha : steam about accept bitcoins via bitpay according to some strings found on a server https://np.reddit.com/r/Steam/comments/44sgnm/bitcoins_in_steam_coming_soon/
[2016-02-08 20:08:38] tscha : about to*
[2016-02-08 20:27:04] sleger : tscha: you're doing your algo in java right ?
[2016-02-08 20:27:14] tscha : sleger: yes
[2016-02-08 20:27:22] sleger : have you connected to 0kc fix ?
[2016-02-08 20:27:58] tscha : no, I am using websocket for the moment, and only .com, but I will switch to fix when I have time... few weeks probably
[2016-02-08 20:28:41] tscha : they (okc) have a java impl for fix
[2016-02-08 20:28:44] sleger : ok, I have started to do the java fix connection
[2016-02-08 20:29:03] sleger : yes but its very poor code sample, however I managed to do everything but cant cxl an order
[2016-02-08 20:29:28] sleger : so i was willing to pay someone to do the job
[2016-02-08 20:29:38] tscha : oh, so it's like the rest they provided.. wrote websocket myself too cause their implementation sucked
[2016-02-08 20:29:52] tscha : i don't have time unfortunately
[2016-02-08 20:30:02] sleger : its a starting point, but their example does not place an order or anything useful really
[2016-02-08 21:12:33] REKT : Liquidated short on `XBT24H`: Buy 50 @ 373.54
[2016-02-08 23:37:39] lockhedge : sleger: did you take a look at OKTradingRequest.java
[2016-02-08 23:37:50] sleger : yes
[2016-02-08 23:38:16] sleger : but whatever combination of clordid, origclordi i put it returns an error order not found
[2016-02-08 23:38:38] lockhedge : so createOrderCancelRequest() is the problem?
[2016-02-08 23:38:55] lockhedge : OrderCancelRequest.set(new OrigClOrdID("1231234"))
[2016-02-08 23:39:39] lockhedge : <code>public static Message createOrderCancelRequest() { quickfix.fix44.OrderCancelRequest OrderCancelRequest = new quickfix.fix44.OrderCancelRequest(); OrderCancelRequest.set(new ClOrdID("2")); OrderCancelRequest.set(new OrigClOrdID("1231234"));//订单编号 OrderCancelRequest.set(new Side('1')); //1买入;2卖出 OrderCancelRequest.set(new Symbol("BTC/CNY")); //BTC/CNY or LTC/CNY OrderCancelRequest.set(new TransactTime(new Date())); return OrderCancelRequest; }</code>
[2016-02-08 23:40:09] sleger : yes, i can submit an order but cant cancel it
[2016-02-08 23:41:11] lockhedge : can you check the order status to get the correct ids to cancel?
[2016-02-08 23:42:07] sleger : you should be able to cancel without exchange order id
[2016-02-08 23:42:45] lockhedge : that seems not possible, you need OrigClOrdID
[2016-02-08 23:43:27] sleger : i thought this was created by me since it says "Cl" for Client in it
[2016-02-08 23:46:57] lockhedge : yes it shouldn't be the exchange set ID that's OrderID
[2016-02-08 23:47:32] lockhedge : OrigClOrdID is a new unique id just for the cancel request (set by you)
[2016-02-08 23:48:14] lockhedge : so OrigClOrdID is a new one, ClOrdID is the one the you want to cancel, both are required
[2016-02-08 23:48:31] lockhedge : you don't need OrderID to cancel
[2016-02-08 23:48:44] sleger : "ClOrdID" should be equal to the "ClOrdID" I created when sending the order ?
[2016-02-08 23:49:07] lockhedge : that's how understand it...
[2016-02-08 23:49:57] lockhedge : and OrigClOrdID has to be a different unique one
[2016-02-08 23:50:37] lockhedge : sorry.... other way around. OrigClOrdID = Original client identifier of the order being cancelled
[2016-02-08 23:51:06] lockhedge : ClOrdID = Unique identifier of this (cancel) request
[2016-02-08 23:51:48] lockhedge : so OrigClOrdID should be equal to your old ClOrdID
[2016-02-08 23:52:03] lockhedge : and you have to set a new unique ClOrdID for the cancel request
[2016-02-08 23:53:32] sleger : thats what I did
[2016-02-08 23:54:05] sleger : for create I did : newOrderSingleRequest.set(new ClOrdID("01234567890")); and for cancel : OrderCancelRequest.set(new ClOrdID("012345678901")); OrderCancelRequest.set(new OrigClOrdID("01234567890"));
[2016-02-09 00:02:32] lockhedge : strange, it should work that way. did you try to check the order status and receive the IDs by the exchange before you cancel, to check if you really have the right IDs? are these IDs strings or integer?
[2016-02-09 00:10:51] sleger : if I request order status, I see my order (same price for instance) but no ID that I sent
[2016-02-09 00:12:47] sleger : maybe its like with WS you cant cancel using your own id and have to use exchange id...