Today is PERL testing for this Streambase C++ open source option with Fastflow being revisited for CUDA GPU in my HFT
Today is the day I learn the basics of Perl to get my trading engine framework tested with the rest of the components of this high frequency trading platform. I know most people ask why I am not choosing another language like Python or Ruby? It is simple. When you look at this Triceps open source package, the author (Sergey Babkin) has used Perl as the scripting language so that is what I use. It seems he has worked in the space of Sybase CEP and Streambase enough to know how to do it smartly. Also, he was a VP at Deutsche Bank so whom I am to argue how they would do it. He even gave an neat strategy idea for currency trading with this thing as an example in his development guide. For those that don’t know, this is a very solid solution to handle my REAL TIME data analytics where there really is no other C++ option.
Although Perl may be nice, I do see some challenges working with the other C++ HFT components like the FIX engine and acquiring the data. So it would might be necessary to implement these within C++. The author of Triceps did hint at caveats with this option so I am fully aware of it. Also, thanks to Matlab Code and now that R the bottleneck has been removed, I will investigate the Fastflow library as well for clustering or CUDA GPU options. It may get exciting!!
http://babkin-cep.blogspot.com/FACEBOOK ACCOUNT and TWITTER. Don't worry as I don't post stupid cat videos or what I eat!