Think of how many films are about making films or how many novels essentially boil down to being about how difficult it is to write a novel, and then compare that to how few examples of this there are in games. This might be partly because the code is generally invisible within games. We are familiar with the role of the camera in film and obviously know how difficult writing can be, but we only really think about the coding of a game when it goes wrong and we start complaining about bugs and crashes.
In while True: learn (), which is surely one of the most unwieldy titles this side of a JRPG, you must use the basic coding practices of machine learning to solve a succession of increasingly complex sorting puzzles and earn a living.
Obviously, given its setting, graphics in while True are rudimentary, but have a charm that appealed to my memories from the 1980s of typing in programs in BASIC on my ancient Amstrad CPC464. The story is conveyed in a nicely cartoony fashion while the interface is mostly clear and easy to navigate, even if I struggled to really get to grips with the thought processes required to get good at the game itself.
On one level, while True acts as a tutorial in how the logic of machine learning works and is pretty successful at this, I just had overcome my brain being too old and set in its ways as more options were opened up and multiple levels of instruction were required. While the relatively simple nature of the puzzles lures you into a false sense of security, the restrictions on the type and number of processes you can use to solve each swiftly makes things challenging.
There is a nicely silly story driving the game along. You’re a struggling coder, commissioned to provide solutions to numerous problems by devising effective neural networks. These can range from organising delivery supply lines to sorting examination results, but the core solution always involves using blocks of code to enable the computer to sort through the inputs accurately and quickly.
Early on this only involves a couple of steps, but as things develop, multiple options become available and finding the best approach to each problem adds a layer of complexity beyond simply completing the level. Once you have devised your network, you can test it for efficiency before rolling it out. Running the program requires paying for a server, so you must balance making enough from the job with the tools at your disposal. As you earn more you can buy parts for your computer that will help to speed up calculations or increase efficiency, but these are expensive.
You are not alone in your coding endeavours, however. The overall story of while True involves you trying to program a cat interpretation system, since it appears that your feline friend is far better at machine learning than you are. Anybody who has had a cat deign to share their house with them will not be surprised that beneath the aloof exterior lies a brain that is ruthlessly efficient at maximising output from minimal effort. It is a long and complicated journey to develop this cat language translator, however, so you must take on many other jobs to cover the costs.
As well as taking on commissions, you can invest in start-ups which involves devising specific programs to solve their concerns. The more efficient and elegant your solution the more widely it will be taken up which will result in greater profit. This aspect is interesting, but I didn’t manage to make any real profit in my attempts.