Timeline
Week 6's Tasks, Own Work.
I've been humbled by this week. It was easy to plan ambitiously when I wasn't taking learning into account, however, not knowing what's ahead and what needs to be done has made some tasks big undertakings. I explore this more in my low-fi prototyping sections below. For now, in terms of my week's tasks, having a web app with three different sections is my ideal minimum prototype. I'll be focusing all my attention on that, with some extra time allocated for the few design research tasks I have left.
For the overall timeline I've done some rearranging. This week used to be the 'Test and Refine' stage, however, I am still in the 'Design' stage. It feels a lot more realistic to design now and over the break, and then test and refine on the first week back. Previous to this change, the first week back was an 'Immerse and Align' stage to ensure that my work was still on track. I believe this stage would fit better in Week 9, where I've had a few weeks to test and refine and now I'm looking ahead to assignment submissions. As well as this I've added an abstract 'Research' task to this week to remind myself. As the Report check in has been moved to the first week back, over the break I want to get an outline done, ideally fitting the writing I have so far into a rudimentary draft.
I make all these changes to fit to my current progress so that my methodology fits to development. Although planning ambitiously led to this, I think it's been the best motivator to give the sense of an 'upcoming deadline'. I did assume that prototyping would be easier than this, and I still believe getting that first prototype will be easy, I just need to get the hang of a new language and a coding mindset. The name of this blog alludes to how big of a step this is for my process, I just need to figure out how to take that small step for man. The only consequence of this setback is that I have nothing to show my peers prototype-wise before we head into the break. However, I aim to have something better to show in our first week back. Next time I may aim to create two timeline structures, one that is ambitious and one that is realistic to my time constraints. That way when I make changes to the ambitious one I still have a frame of reference for being on track. At the moment this frame of reference is the progress of my peers, which I am on par with (if not ahead).
React JS Learnings
I had worked with Javascript before, but not extensively. I assumed it wouldn't be too hard, and started by following 'React JS Full Course 2023' to build my first web app. This app was a website that allowed you to search for movies and see their descriptors, and features: dynamic PC/phone layout, movie data from an API, classic React commands, and a CSS file. Following this video was hard for me as my attention span isn't great for videos over 30 minutes. I'm proud of myself for completing it.
Currently I'm in the process of attempting to recreate my wireframe in React while learning along the way. I'm currently stuck with someone's bottom navigation bar implementation that shows up, but doesn't work. Reflecting on this, a big challenge I'm facing is attempting to use code other people have created while not fully understanding it. This makes my process take more time as I'm using guesswork to develop. Following from now I want to strip my app back and develop from the ground up, therefore understanding the commands I'm using along the way.
AI Integrations
I initially coded in React JS as Inworld AI specified integrations with it. Since then, I still think React is a good language to develop in based on it's efficient management of pages. I attempted to integrate my Inworld companion into my React app with little success. At that time, I had made the assumption that their Quick Start guide was misinformed, and there was hidden knowledge that I didn't have that I needed to 'quick start'. However, just before finishing this blog I had a chat to a PhD student from the ECL and he insisted that there was code Inworld supplied alongside the guide. I found the code, and I'll be working to test and integrate it over my lecture break. Because of this experience, I feel more inclined to reach out to the ECL with future development questions over the break as the chat I had showed incredible relevance to my project dev.
After my first challenge with Inworld I swapped to trialing Dialogflow. I went through a small part of setup: Small Talk. Through this I answered seven categories of common chatbot questions with specific dialog I wanted my bot to follow - you can see this below in Figure 1. Through this I was able to develop the tone I wanted. In the future I want to fully finish this chatbot prototype, but at the stage it's at now, I'm able to use it as a UX/UI dev tool. This is great for my motivation, as it gives me something to work with. It turns out that Dialogflow's integration is a lot more simple than Inworld's, potentially due to the fact that there's little room for implementation customisation. This statement is an assumption due to what I've seen.
Figure 1. Small Talk sections and questions. DialogFlow, 2023.
References
DialogFlow. (2023). Small Talk sections and questions [Screenshot]. https://cloud.google.com/dialogflow