Too Many Compromises

Two months ago I started down the journey of creating an Android application, my first in a long, long time. The software was aimed at an audience that seemed under-represented in both the Google Play Store and Apple's AppStore: UFO enthusiasts. In both stores there is an absolute dearth of applications and, of the ones that do exist, none of them have seen any updates in the last couple of years. What I wanted to do was to create a tool that would not only show the most recent reports, but also allow people to report things that they may have seen.

In addition to this, there were some other functions that would have given the software a little bit of an advantage over other UFO apps and websites that have been researched recently:

  • a dynamic heatmap would show where in the world (or a specified area) people claimed to have witnessed a UFO sighting
  • the ability to read the reports provided by the eye-witness
  • the ability to comment on and "score" a report
  • the ability to report a sighting and provide an accurate GPS location, photos, audio, and/or video supporting evidence
  • a library of different sighting types
  • a library of different vehicle types
  • a library containing theories of where these visitors come from
  • the ability to send the report to one of the more popular UFO sighting databases
  • plus a few other nice features

Using my weekends and evenings, the application was written and ready for deployment in about five weeks and it works pretty decent on my test devices which include a Sharp phone from 2013 running Gingerbread. The API back end is fully coded and the database contains almost 80,000 records of eye-witness accounts -- including geographic coordinates -- which had been pulled from the NUFORC web reports. However, as I got further into building the application, the less enthusiastic I felt about it.

These are some of the issues that crossed my mind:

  1. The data seemed highly suspect, with the most common sighting location being a place called 100 Mile House in Canada. Any place that is not Nevada with a ridiculously high proportion of sightings is going to be suspect, especially given that the 1,200+ reports seemed to be written by the same person.
  2. Allowing people to comment on and score the testimony of others -- even with the aim of "democratising the reporting process"1 -- would quickly devolve into people calling everything "bullshit" and adding zero value to the topic. Too much negativity is not a good thing.
  3. The people that would use this application likely strongly believe they saw something … and the application was actually designed in such a way as to show people that 99% of UFO reports are complete fabrications.
  4. I would be trying to monetize the application via advertisements, with an in-app purchase to shut them off.

The further I got into development of the system, the more wrong it felt to continue. There were simply too many compromises to make:

  1. the app was written for Android rather than iOS, because that's what most people use
  2. in-app advertising (and its subsequent tracking) would be in place for the vast majority of people, as Android apps generally do not earn a great deal of revenue unless they're a game of some sort
  3. the 80,000 records collected from NUFORC were of dubious quality
  4. people who report sightings generally believe what they think they saw, and I would be encouraging people share the story with me so that my database would become larger so that the growing size could be used as part of the reson for more people to use the application

But, perhaps most damning of all, is reason number 5: I don't believe a single one of the reports.

Yeah, I was happy to learn how to write an application in Kotlin, have it compile and run on actual hardware, and see it go from a concept to a working system. But if I don't have any faith in the things that the people using the system have to say, is it right to carry on with the project? The honest answer is "no".

So despite the efforts, the application has been scrapped. All of the code is archived in GitHub and will likely sit there for a long, long while before it's either forgotten or deleted. The database I may make available as a MySQL dump, as it's a right pain to collate this data from various sources online, but only if I receive written permission from NUFORC to do so.

I am still very much interested in building some small applications as a means to test the feasibility of earning a living through independent application development and this project ejection has certainly allowed me to confirm how I do not want to go about it. This does bring me back to square one, though, which means something else will need to be devised, planned, and built to scratch the itch that leads to self-employment; ideally without compromises.

  1. As it stands, all UFO reports go through no more than 5 gatekeepers around the globe, depending on which group you're reporting the sighting to.