A clearer picture of TANDEM 0.5 is emerging as we move forward. The reality of the thing becomes more tangible once we start to test and install application dependencies and the different modules that we plan on stitching together. While the process of stitching and feel of the interface are still undefined, we can at least rest easier knowing that pieces function to an acceptable extent on their own. A lot of our work is contingent on existing technologies which in some ways makes it easier and other ways makes it harder to work.

My team continues to hold down the fort as we divvy up tasks. I would love for them to be more technical, myself included, but our baseline of knowledge and desire to learn is making things move smoothly. We have a good process of documentation going on in Google Drive. Each member is keeping track of their own tasks. Team emails are also less confusing now that we are more selective about who is on the threads and what topics we speak about in Gchat and in email.

I think the hardest part here is avoiding rabbit holes.

It is so easy to start researching and testing one feature for TANDEM and getting lost in possibilities, questions, things that seem concerning but really have no impact on building the tool at all. An identifiable weakness of mine is surely my distractibility. My susceptibility to rabbit holes is something I need to work on avoiding.

In those rabbit holes I have found intriguing things. For instance, ocropy is another open source OCR that may be more suited and easier to implement than Tesseract. So, it may be safer to say I should police my rabbit-hold-susceptibility and knowingly enter with the ability to leave at any time. Knowing I am in a rabbit hole may be better than discovering how far I have deviated in two hours of clicking around.