After all, let me mention that many people think that .net executes a kind of scripting / interpreted code, and thus it's much slower than native c++ or c code. This is simply not true because .net runs the same byte (machine) code as c++ executable. The only difference is that .net uses JIT that compiles IL (intermediate language) code to runtime bits of machine executables "just in time", i.e. when someone asks for this or that method.
Yes. The ATS could simply be two or three server components: one component processing a real-time quote stream, another component and deciding the rules and perhaps sending trade instructions/signals, with yet another component processing and sending messages reflecting the trade processing. Or all of this could be rolled into a single application if the user had very simple and limited demands: $soft tools are great for doing things quickly if your requirements are not too demanding....
Just a side note: Never ever use version 1.0 of any software product. Current vestion of .net is 1.1 wink wink
Yes and in many cases this produces acceptable performance. In very demanding situations this becomes an issue and the entire framework gets in the way. $soft tools and frame works can be great time savers but they come at a price and can lock you into a vendor relationship that by my standards is unreliable.
2.0 is not out yet. I like .Net. Don't get me wrong but I will use it only after I see successful products developed on top of it. Until now I saw none.
"This is simply not true because .net runs the same byte (machine) code as c++ executable"...Agreed ! I have developed and automated spread tool that was sucessafully used on CBOT during roll-over periods and generated quite a bit of money when used sensibly. This was all written in C++. But anyway, getting back to the topic. As a beginner you would probably want to start looking at C# programming for its high-level simplicity and ease of use instead of C++. Anyway, I though that 2.0 was still in its beta versions or am I behind Thanks
thanks for all the quick replies, def gives me something to go on. just for clarification, i want a program that can analyze and calculate formulas based off various data streams, then enter trades based off that. another question i have is whether or not i can use the data stream that my trading platform uses (say, for prints and market depth) as the source of data for my program, or whether i need to buy another stream from nasdaq and such. to tell u the truth, the order entry and management part i could really do myself at first, i just need a program to archive and analyze alot of information, specifically the prints and level ii, in real time. thanks again
Buy versus build is always dependent upon your needs and goals, budget, and existing resources and skill sets. In some cases avoiding vendor lockin is a risk mitigating factor ....