I was originally going to title this post ‘Took a wrong turn…’ but I really, it was less of a mistake, and more of a learning experience. Vague, feel-good banter aside, I should probably elaborate. My current task with SunPy is to write and interface that allows our library to politely talk to the HELIO web-service. My mentor helped me into contact with an individual (Marco) who has already created such an interface, but it was written in another language and with different end goals in mind. After much brow-beating and one rather long Google hangout later, we were able to get the code working on my computer. I was exploring the possibility of interfacing with the Java library that Marco had written. After spending upwards of a week and a half exploring Python modules to interface with jars, browsing Marco’s code, and playing with tests, a severe potential problem emerged: a single point of failure. Though well written, and though it would simply implementation of an interface, relying on a single Java archive, developed by a single individual outside of the SunPy group, presented a worrisome possibility that should anything happen, there would be little SunPy could do other than scrap the HELIO module and start over from scratch.
It was an attractive option, but the other option shouldn’t be too bad. My new course path will be to connect with HELIO via a RESTful or WSDL interface, which in conjunction to the currently written Java code of a similar nature, I’m sure that I can crank out this module without too much issue. I’ll just finish with a special thanks to my mentor David for helping out and Marco for providing so much help with understanding his code. Now, I’ve got to run to class.
My Summer of Code 2013 