OK: There is a website where all members freely post any code in any language that they have developed. Ordered by language and theme. Listed by download frequency. We share, we win. I mean - surely someone else has written a custom backtester in C# that can run off any data format. But here I am re-inventing the wheel. We all re-invent the wheel - our resources are limted to ourselves. That makes us all weaker. Broker and Feed APIs with our own custom systems. This is the only future I see. The software vendors can stick it where the sun don't shine.
There's plenty of open source projects. Just check sourceforge or do a google search based on what you would like to do. You're not just not looking for them well enough.
if this is something you guys are interested in, i could start and host the site, but i don't have time to moderate it. PM me and let me know if you would be interested.
I would moderate it if someone would host and webmaster it - I think an open source trading software archive would be great!
I can provide a code repository right here on ET. I would just need you guys to help me brainstorm on how it should be structured. Should the code snippets be grouped by language? Platform? Strategy?
Baron: Knowledge base structure example: http://www.tssupport.com/support/base/?action=search&pid=0&string=
Here you go. www.atsedges.com Just created this today. Was looking for the exact same thing you were talking about.
Chicago that looks nice - but we don't really need a forum. We want a code repository. Baron is right - the organisational structure is the key. I think that namespace dependency should rule the day. So a C++ utilitity that requires the .NET framework would go in .NET. A Java utility that hooks into the IB API data stream would be under IB API. A strategy code written to run off NeoTicker code would have to go into NeoTicker etc... This way a person browsing would know right away what the dependencies are (oh I can't run that beacue I don't have the TT XTAPI for example). Then all that's required is a simple description of what the code does - the poster being responsible for entering that at the time of uploading. Then the "posts" (really deposits in the bank of shared knowledge ) should self organise by download frequency, and each post should have an attached thread where the poster and users can nut out details. Open Source declarations without any restrictions must be provided by some sort of click to agree when depositing and also a there are no guarantees download at own risk when downloading.
Still a few problems with that- where does an easy language dll coded in C++ and c# go? Anyway we could start small and figure out structures step by step.