An earlier post outlines my interest and intent on creating an on-line game. I have been thinking steadily about this project since then, but, perhaps unsurprisingly given my past issues on earlier projects haven’t got a damn thing written (well, code/database wise, I have a document I started to record my thoughts; it is four pages and over 2K words so far). Part of the problem is, no doubt, analysis paralysis. I am confident on the large structured elements, but as soon as I start peeling back the layers to reach implementation suddenly I find myself without the ability to make decisions.
The first major issue was how to represent the interface. I had already decided that plain text wasn’t going to be the direction, text on a smart phone or tablet is just about the worst possible interaction method. My plan was to use simple graphics to represent the state of the player in the game environment, though defining ‘simple’ has not been very simple at all. I thought for a bit that I would just focus on the mechanics of dynamically creating compressed gif images (I have done a wee bit of image processing to create a web page counter) but even there I got lost in decision making. The open source libraries that are available are way too complex for what I need and, since they are packaged as libraries, it isn’t that straight forward to cut out just the bits I want to use. I studied the gif specification for a while, but just wasn’t interested in recreating, for undoubtedly at least the 10,000th time, an implementation, so passed on that. After a few more days I decided to shift focus onto the text game play as at least that could be alpha tested by people on PCs.
In contemplating my extension to the ADVENT database I had developed the idea of a somewhat conversational interaction with so-called non-player characters (NPCs). What I had envisioned initially was a sort of key word look up method in that I would parse what the user typed in, strip out ‘noise’ words (like ‘the’, ‘if’, etc.) and then do a query to find a pre-programmed sentence that would score most highly as a likely response. This, of course, would require a great deal of thought on my part to provide responses and prompting keywords and, if you haven’t caught on to that yet, I am a really lazy person, so just couldn’t get over the activation energy. So, it wasn’t that long before I gave up on that approach and started to do some Googling on natural language processing (NLP). I had looked briefly into NLP several years ago when I was interested in building a spam filter, so know it is not trivial or for the faint of heart, but when I started to learn about chatterbots I figured I had something that would go along way toward what I wanted to accomplish. However, it seemed upon further research that they would be too general for me and not provide the specificity I needed without a huge amount of customization (back to having to work! damnit!). I tried searching for some sort of database of conversations I could parse apart and use as a baseline and found this really interesting post “The Dynamic Conversation Engine Concept“. However, I wasn’t able to find much more (the document is very interesting, though) and didn’t find any contact info for the author (of course, it might be blazingly obvious; I am notorious for not being able to see things right in front of me) so that seemed like a dead end.
The conversion I started on the ADVENT program is incomplete (it doesn’t even require you to have the lamp on to go underground!) and I thought about working some more on that, but just can’t seem to get the activation energy. I have grown increasingly tired of the IT industry over the years, no doubt that has influenced my ennui, but I really like the idea and spend a lot of time thinking about it. I just can’t seem to translate that into action. Am I just a whiny procrastinator who talks a good story (presuming, naturally, you consider any of this a good story)? I used to be able to ‘do things’ programmatically. I still ‘do things’ with my hands, I build greenhouse/pools on the weekends, but there really isn’t a whole lot of thinking going on there, or rather the thinking has been front-loaded (nigh on 3 years ago now) and I am sort of on autopilot. Interestingly, I guess, when I do come up against issues that need to be resolved, it is often my wife who comes up with the solutions now (of course, I married her for more than her sexy body to begin with, so this isn’t surprising to me), so I wonder sometimes if I am no longer capable of taking action on ideas any longer.
I hope this is just a phase…
