NexusFi: Find Your Edge


Home Menu

 





A new (open source?) trading platform


Discussion in Platforms and Indicators

Updated
      Top Posters
    1. looks_one MXASJ with 13 posts (12 thanks)
    2. looks_two Big Mike with 12 posts (5 thanks)
    3. looks_3 tradelink with 10 posts (10 thanks)
    4. looks_4 NetTecture with 10 posts (13 thanks)
      Best Posters
    1. looks_one NetTecture with 1.3 thanks per post
    2. looks_two wh with 1.3 thanks per post
    3. looks_3 tradelink with 1 thanks per post
    4. looks_4 MXASJ with 0.9 thanks per post
    1. trending_up 78,546 views
    2. thumb_up 90 thanks given
    3. group 38 followers
    1. forum 142 posts
    2. attach_file 5 attachments




 
Search this Thread

A new (open source?) trading platform

  #51 (permalink)
 MXASJ 
Asia
 
Experience: Beginner
Platform: NinjaTrader, TOS
Posts: 796 since Jun 2009
Thanks Given: 109
Thanks Received: 800

Might be splitting hairs (or asking a silly question) here but is there an advantage to having your exchange(or broker)-facing engine running a very light core written in C++ on a Linux box? Everything else can be dot Net, and perhaps a sockets connection is used between the FE and BE? Is sockets even the prefered way to get a Linux app to talk to a Windows app?

The reason I bring that up is I used a commercially-available platform recently that did just that. The C++ core was so lean it didn't event have a concept of a bar, only a tick plus all the other raw price and order data. If you required a bar you would have to build it either on the FE or the BE.

I can see the workflow advantage of a full dot Net environment, but is one at a disadvantage with that technology facing the exchange?

As always your opinions are most welcome...

Reply With Quote
Thanked by:

Can you help answer these questions
from other members on NexusFi?
Exit Strategy
NinjaTrader
REcommedations for programming help
Sierra Chart
NT7 Indicator Script Troubleshooting - Camarilla Pivots
NinjaTrader
MC PL editor upgrade
MultiCharts
ZombieSqueeze
Platforms and Indicators
 
Best Threads (Most Thanked)
in the last 7 days on NexusFi
Just another trading journal: PA, Wyckoff & Trends
34 thanks
Tao te Trade: way of the WLD
24 thanks
GFIs1 1 DAX trade per day journal
16 thanks
Vinny E-Mini & Algobox Review TRADE ROOM
13 thanks
My NQ Trading Journal
12 thanks
  #52 (permalink)
 NetTecture 
Szczecin
 
Experience: Intermediate
Platform: Ninja, writing own now
Posts: 211 since Mar 2010


MXASJ View Post
Might be splitting hairs (or asking a silly question) here but is there an advantage to having your exchange(or broker)-facing engine running a very light core written in C++ on a Linux box? Everything else can be dot Net, and perhaps a sockets connection is used between the FE and BE? Is sockets even the prefered way to get a Linux app to talk to a Windows app?

The reason I bring that up is I used a commercially-available platform recently that did just that. The C++ core was so lean it didn't event have a concept of a bar, only a tick plus all the other raw price and order data. If you required a bar you would have to build it either on the FE or the BE.

I can see the workflow advantage of a full dot Net environment, but is one at a disadvantage with that technology facing the exchange?

As always your opinions are most welcome...

Yes and no. As I said - you loose a core, and you have to block task switching because it does bring a slower scheduler in. OTOH.... you gain speed. Do you? The moment you hit the network you are mostly in another task anyway with interrupts (due to the network driver). I am not sure that at the end you really safe something. It may make sense for HFT, but then you need a VERY lightweight OS below it - one that is quite primitive. Is there a real time linux with guaranteed network driver latency?

Reply With Quote
  #53 (permalink)
 
Big Mike's Avatar
 Big Mike 
Manta, Ecuador
Site Administrator
Developer
Swing Trader
 
Experience: Advanced
Platform: Custom solution
Broker: IBKR
Trading: Stocks & Futures
Frequency: Every few days
Duration: Weeks
Posts: 50,440 since Jun 2009
Thanks Given: 33,212
Thanks Received: 101,599


Guys -- dare I say it, but why not work together?

A lot of you have expressed that you have worked on, are working on, or plan to work on your own platform. Assuming the majority of you are doing this to become better traders, and not to sell the platform for revenue stream, then why not join forces together?

Mike

We're here to help: just ask the community or contact our Help Desk

Quick Links: Change your Username or Register as a Vendor
Searching for trading reviews? Review this list
Lifetime Elite Membership: Sign-up for only $149 USD
Exclusive money saving offers from our Site Sponsors: Browse Offers
Report problems with the site: Using the NexusFi changelog thread
Follow me on Twitter Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #54 (permalink)
 MXASJ 
Asia
 
Experience: Beginner
Platform: NinjaTrader, TOS
Posts: 796 since Jun 2009
Thanks Given: 109
Thanks Received: 800


Big Mike View Post
Guys -- dare I say it, but why not work together?

A lot of you have expressed that you have worked on, are working on, or plan to work on your own platform. Assuming the majority of you are doing this to become better traders, and not to sell the platform for revenue stream, then why not join forces together?

Mike

I think that would present a challenge on a few fronts, beginning with something as basic as Use Case.

I, for example, don't normally use charts for anything other than post-trade visualization i.e. I have no need for real-time charting in anything I build myself. The requirements for static data such as instrument names, expiration dates, tick sizes and values, exchange names, trade times, etc will vary from user to user. And then there is an API. Nothing is really free, and John, Paul, George, and Ringo might have access to TTFIX, Rithmic, RTS, and PATs respectively but that will present a challange in a distributed development environment. And then of course my requirement for MSSQL in the BE will be met with the open source crowd wanting mySQL...

All that before a single line of code is written .

NetTecture was kind enough to post some of his base code, and some of the other stuff out there is also a good place for ideas.

Not in any way shooting down the idea of collaboration, but I think it is more of a challenge for the trade platform than say, a middle-office trade reconciliation/risk management platform. And personally I'm an absolute crap coder at present... but working on it.

My 2 pesos...

Reply With Quote
  #55 (permalink)
 NetTecture 
Szczecin
 
Experience: Intermediate
Platform: Ninja, writing own now
Posts: 211 since Mar 2010


MXASJ View Post
I think that would present a challenge on a few fronts, beginning with something as basic as Use Case.

I, for example, don't normally use charts for anything other than post-trade visualization i.e. I have no need for real-time charting in anything I build myself. The requirements for static data such as instrument names, expiration dates, tick sizes and values, exchange names, trade times, etc will vary from user to user. And then there is an API. Nothing is really free, and John, Paul, George, and Ringo might have access to TTFIX, Rithmic, RTS, and PATs respectively but that will present a challange in a distributed development environment. And then of course my requirement for MSSQL in the BE will be met with the open source crowd wanting mySQL...

All that before a single line of code is written .

NetTecture was kind enough to post some of his base code, and some of the other stuff out there is also a good place for ideas.

Not in any way shooting down the idea of collaboration, but I think it is more of a challenge for the trade platform than say, a middle-office trade reconciliation/risk management platform. And personally I'm an absolute crap coder at present... but working on it.

My 2 pesos...

Ok, here is my new years goal. I am currently in the process of getting my code base more organized. Sadly this is blocked by me currently not trading in person (I have someone trading some of my strats) but being ina data warehouse proejct for 50 hours per day After that I am more a WOW addict than anything. As in: I also need some rest. That will take possibly till mid of the year (that said, it pays handsomely). Anyhow.

* I have a place on codeplex where I want to put some stuff. Tradex.codeplex.com
* I have my own TFS in my company for the builds and source control. THat said, there is a bridget that would allow partial publication to TFS. I never managed to acutally set that up.
* So, today is my last fuull working day this year. Tomorrow I will spend traveling and doing shoppnig. After that it is 2 weeks with time.
* And in this I want to get some of my code at least up to Codeplex. This will be core libraries, data structures, and possibly connectivity code as far as I have it. All my current services interfaces and possibly the general service runners (my "server" is some generic code that then loads services that do the work - keeps the individual services nicely packed).

Licensing will be very lenient (BSD style - take, use, modify, use comercially, just credit origina). Feel free to jump on it.

Services, front end etc. will be excluded form the free upload, at least "in general". Front end also needs commercial components (user interface).

Reply With Quote
  #56 (permalink)
 traderwerks   is a Vendor
 
Posts: 692 since Jun 2009
Thanks Given: 436
Thanks Received: 465


MXASJ View Post
Might be splitting hairs (or asking a silly question) here but is there an advantage to having your exchange(or broker)-facing engine running a very light core written in C++ on a Linux box? Everything else can be dot Net, and perhaps a sockets connection is used between the FE and BE? Is sockets even the prefered way to get a Linux app to talk to a Windows app?

The reason I bring that up is I used a commercially-available platform recently that did just that. The C++ core was so lean it didn't event have a concept of a bar, only a tick plus all the other raw price and order data. If you required a bar you would have to build it either on the FE or the BE.

I can see the workflow advantage of a full dot Net environment, but is one at a disadvantage with that technology facing the exchange?

As always your opinions are most welcome...

I did something a few years ago. The core was a small CEP engine that ran on Linux, and the UI / Control user interfaces was in WxWidgets connected with zeromq so the UI and execution were separate.

Reply With Quote
Thanked by:
  #57 (permalink)
tradelink
Miami Beach FL USA
 
Posts: 21 since Dec 2010
Thanks Given: 4
Thanks Received: 16

Hey sam, I see you were having trouble using tradelink.

Did you join the tradelink users list? We'll help you out with whatever problems you're having.

You can do this by googling tradelink, clicking on the tradelink project. Then on the right side of the page you will see the link for 'TradeLink users'

Please let us know what trouble you encountered.


sam028 View Post
Good luck for tradelink, I haven't been able to run most of its pieces.

I agree with Aslan's post, and he's right when he says the archiecture is the key (but I think C++ will be a better choice than C#, and forget about Java ).


Reply With Quote
  #58 (permalink)
 NetTecture 
Szczecin
 
Experience: Intermediate
Platform: Ninja, writing own now
Posts: 211 since Mar 2010

Just to give an update redaring Tradex (my framework).

* I installed the sync tools. WIll see I get them working sunday. This will then start moving source into codeplex. Live link - all checkins automatically get moved, including work items.
* I make good progress on the NxCore connector. This is a hell of a decent data feed api Also got permission to publish the NxCore connector code- but it will need access to the local API before compiling, sorry. i dont distribute other peoples code and dll's.
* And finally I just got the first smaaaaaalll class covered 100% by unit tests. Another one on the way - the really nice thing is that I find obscure bugs with this. Nice. I start moving the core asemblies under full unit testing now because basicall I finalyl got code coverage reporting working in TFS and - they are so central I really dont want any bugs in them.

I would expect some updates to happen january.

Reply With Quote
  #59 (permalink)
 
Big Mike's Avatar
 Big Mike 
Manta, Ecuador
Site Administrator
Developer
Swing Trader
 
Experience: Advanced
Platform: Custom solution
Broker: IBKR
Trading: Stocks & Futures
Frequency: Every few days
Duration: Weeks
Posts: 50,440 since Jun 2009
Thanks Given: 33,212
Thanks Received: 101,599


NetTecture View Post
Just to give an update redaring Tradex (my framework).

Congratulations, a nice end of year accomplishment

Mike

We're here to help: just ask the community or contact our Help Desk

Quick Links: Change your Username or Register as a Vendor
Searching for trading reviews? Review this list
Lifetime Elite Membership: Sign-up for only $149 USD
Exclusive money saving offers from our Site Sponsors: Browse Offers
Report problems with the site: Using the NexusFi changelog thread
Follow me on Twitter Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #60 (permalink)
tradelink
Miami Beach FL USA
 
Posts: 21 since Dec 2010
Thanks Given: 4
Thanks Received: 16


Hey guys,

I would encourage you all rather to write your own project from scratch to join the TradeLink project.

It was created 3 years ago when I left one of the largest commodity funds in the world and went to trade on my own. There was not much out there open source wise so I created a platform and open sourced it. That was three years ago, now tradelink supports 15 brokers and data feeds including all the major proprietary trading firms and many retail ones.

It works with equities, options, fx, futures on both live and historical tick data. You can backtest level 1 and level 2 strategies at over 300,000 ticks a second, which is faster than many commercial platforms costing hundreds of thousand of dollars a year. Live trading operates at 40,000 ticks a second. We have over 12,000 downloads, 200 users on the mailing list and 50 commercial users representing half a billion in assets.

We also have a seperate company providing commercial support to hedge funds, prop trading firms and larger users who need extra assistance. Corporate supports helps to accelerate the growth of the open source project, because the more big companies who use it the more powerful the open source platform becomes for everybody. Again we've been at this for three years. Regardless of how big it gets, We're 100% commited to open source.

What I'd love to see happen is for you guys to use and contribute to tradelink and make it better. Together we can build something so much more impressive and powerful than having a bunch of fragmented individual platforms, which are great for personal use but to really make open source work you need a powerful community.

If this sounds interesting to you, please join up on the mailing list.

google tradelink
click on the 'tradelink project', hosted by google code (google's version of source forge)
here is where you can download the installer, view instructions/videos, read tutorials, download the source code
to join the group, click on 'tradelink users group'. here we talk about basic technical support issues in tradelink as well as extending the platform.

looking forward to seeing some of you there,


Big Mike View Post
I think the right people haven't seen the thread yet. I can't remember everyones names is part of the problem -- but a lot of people have told me over the last year they've got code already started, not just an idea but actual work done.

So hopefully they'll find this thread and we'll see if there is any interest in working on a group project. I guess it will depend on their objective.

Mike


Reply With Quote
Thanked by:




Last Updated on March 10, 2024


© 2024 NexusFi™, s.a., All Rights Reserved.
Av Ricardo J. Alfaro, Century Tower, Panama City, Panama, Ph: +507 833-9432 (Panama and Intl), +1 888-312-3001 (USA and Canada)
All information is for educational use only and is not investment advice. There is a substantial risk of loss in trading commodity futures, stocks, options and foreign exchange products. Past performance is not indicative of future results.
About Us - Contact Us - Site Rules, Acceptable Use, and Terms and Conditions - Privacy Policy - Downloads - Top
no new posts