How Prototyping Saves My Sanity

bugatti-1612389_1280I am a firm believer in “good enough” mentality. When creating something, it helps to decide what would be good enough for the purpose you have in mind. Then you intentionally aim for that target. I would say that is the crux of any “lean” approach to writing software or starting a company. You do just enough to reach your objective and no more. That way you can avoid waste. Reasonable people can agree that avoiding waste is generally a good thing.

Still, the phrase “good enough” can be an excuse to avoid difficult work. It might also be used to rationalize being stingy, holding back what would benefit others. When striving to exceed expectations, an abundance mentality is more appropriate than squeaking by with something that is “good enough.”

To be effective, the phrase “good enough” has to be followed by the question “for what?”

If your purpose is to sell something, it has to meet a need that people will pay for. If your purpose is to gain notoriety or to build a legacy so that you are remembered for generations, you are setting the bar quite high. Whatever you create had better be outstanding.

As a solo-preneur—that is, someone working alone on bootstrapping a business around game design and development—my bar for good enough has to be relatively low. I am continuously testing ideas for viability. The ideas need to be good enough to gain even a little attention and to see what might work in a grander scale. I do not count on selling much in the early stages.

In my mind, ideas evolve quickly, and before long a vision begins to crystallize. Imagine with me that I have attracted thousands of followers. People are happy to use my amazing software and pay for it without hesitation. I have grow my company to a half-dozen people, and we are busy creating ever more software people love.

Hollywood has conditioned me to suspend my disbelief, so believing in my fantasies is easy. The hard parts are ignored, risks are dismissed. Would it be prudent to go all-in based on the movie of success in my mind? No, that would be crazy, or at least rather foolish.

Others would think so, too. They might whisper behind my back. You can almost hear the critique.

  • “Is that guy nuts?”
  • “What was he thinking?”
  • “He thought this drivel would draw an audience?”

On the other hand, all I need to do to keep my sanity is prove that the idea can work. Engaging in research and development toward in the pursuit of an idea is certainly reasonable. If it works, I’ll have something to show that others can understand. If I cannot find a way, or if it does not work as well as I imagined, I can use that evidence to adjust my idea or walk away from it.

With any non-trivial project, such as putting together an online game, I like to work in stages. The first stage is intended to answer a few questions.

  1. Can my idea work?
  2. What kind of effort will it require?
    • What skills?
    • What creative elements?
  3. Will the idea be interesting enough for others?
    • To try?
    • To pay for?

This first stage is a great time to build a prototype. Our ideas can be full of magic that needs to be turned into moving parts based on physical reality. A good prototype does the trick.

An Analogy

Think about a car manufacturer who wants to develop its first electric car. Does the company sell the first car it produces? No. Does it even have an assembly line for the first car? No. Although the car company knows how to produce cars, an electric car is fundamentally different than those that burns fuel. That requires some new ideas.

A team in a research lab might start by assembling by hand a wheel, an axle, an electric motor, and a battery. Someone might write software to control the motor. Although far from being a complete car, this is a basis for learning how an electric car might work.

Once those pieces are working together, the team might add a brake, updating the wheel and controller to accommodate. The team might try cutting-edge technology as well as something more traditional. In fact, they would consider the technology choices for each component, whether to stick with what has worked, to try something ground-breaking, or to invent their own technology specifically for this new purpose.

They probably have things wired to test and measure their attempts. Do the breaks get too hot? Can they overpower a runaway motor? Can a bug in the controller software lead to a safety issue?

The point is to learn, to push the boundaries, and to prove what can work. This is the essence of prototyping. See if your ideas can work, get a sense of the effort involved, and give yourself a way out before spending too much on full development and the trappings of a production launch.

Decisions about paint colors and interior fabrics can probably wait. Adding a rocket booster is probably also out of scope, as tempting as it might be to explore.

A True Story

Back in December, I wrote and launched my first release of StoryTime, which is a game platform for writing and playing interactive, text-based adventures. The first release was good enough in about two weeks.

By now you know to ask, “Good enough for what?”

Glad you asked. My target was to launch a “reader” application for playing story-games, with one complete story loaded and ready to play. The launch was merely a matter of hosting the system where anyone with an Internet connection could try it.

I did that.

Then I needed a way to register authors, so I added a database to the back-end and hooked up an identity provider to handle sign up and authentication. That work included a feature for players to opt-in and become authors.

Once it was working, I released it.

The release was nothing fancy. Real authors would probably want to provide a lot more information about themselves, like the write-up that you find on the back of a book cover. I only wanted to show that I could save any information about the author. What I had was good enough as a proof.

Then I added the Writing Desk, where authors go to write their story-games. It’s slightly more complicated than using a word processor because you have to hook up the scenes in a certain way. However, there’s no coding involved for the author. Just type and save.

So I released that.

The Writing Desk also lacked features that would make for a carefree user experience. One problem was the way the story was organized. The layout made it possible to get to everything but not necessarily intuitive. Also, the editors lacked feedback cues, and without auto-save, it would be easy to lose work.

Still, it was good enough to show how all of the pieces of a story-game could be written and connected.

Then I added a way to publish story-games. That promotes a game into the library where anyone can find it and try it out. Although the UI had a couple of refresh issues where updates were not being displayed immediately, nothing was broken fundamentally.

So I pushed that out, too.

The publishing step involves categorizing the story-game by assigning a rating (for age-appropriateness) and one or more genre. While capturing the information was import to show how stories might be grouped and searched, new features are needed to make use of the information. That didn’t stop me from releasing.

And there it stands, a working example of a simple gaming platform. Is it good enough? Well now, that depends, as you know. Good enough for what?

If I believed the movie in my head, the fantasy that I was creating something thousands would love and pay to play, my only option at this point would be to break down and cry. StoryTime is nowhere near good enough for commercial purposes.

Luckily, I have a mindset that allows me to appreciate what I have done and use that to understand what I could do next.

What I have done is to build a prototype of the gaming system I imagined.

  • The prototype is not colorful, at least not in an artful sense. At least I used Bootstrap for styling, so it isn’t horrible. Unless you don’t like Bootstrap.
  • It is not elegant. As I mentioned, it’s pretty easy to lose work, which is about the worst thing that can happen to a writer.
  • The library is still nearly empty. It’s got the one story-game that I wrote back in December. I have been too distracted to write another. Also, droves of writers have not decided to invest irrationally in a game platform without players, an ecosystem, or an obvious payout. Go figure.
  • Mostly players don’t want to become writers. In fact, players want graphics. A text-only game is a tough sell these days.

A game designer ask if StoryTime could be used as the story line in Unity, a popular game platform. That’s a good idea. So perhaps what I have done is assemble a new mechanism that could be built into a larger system. Like figuring out a drive train based on electricity.

As a stand-alone gaming platform, StoryTime could use so many more features. What about illustrations, at least for the cover art? How about collaboration? Shouldn’t there be a review process to make sure what’s being published meets community standards?

Say authors actually used StoryTime to create alternate-universe games. How would games be sold? What’s the business model? Some companies have made this work. Others have closed their doors.

In short, the gap to realize my dream for StoryTime is huge. Finishing will involve a lot more work. It will also take more talent than I have. To see it through, I will need that six-person team. And I still don’t know if would be worth it.

Meanwhile, something else happened that would be enough to drive a normal person bonkers.

Who Moved My Technology Cheese?

The last release of StoryTime was in April. Since then, I have been busy with other concerns. As I returned to my little coding project last week (now July), I found what has become a familiar scene. The software still runs, but everything it relies on has moved. Allow me to explain.

When I started this incarnation (let’s call it the 3rd go) of StoryTime, I made some choices for the underlying software technology. First, I decided to go 100% JavaScript. StoryTime uses ReactJS and its ecosystem of modules for single-page-app support. On the back end, StoryTime offers a service with a REST API defined using Open API 3.0 (a.k.a. Swagger). The service uses Express on NodeJS 8. That connects to a PostgreSQL database where the stories and data are kept. As any React developer might guess, the app uses React Router for navigation and Redux to manage UI state. I also chose Bootstrap for styling, as I mentioned before.

Everything is written using es6+. That’s a relatively new form of JavaScript that anyone serious has been using for a while, with Babel transpiling down to something that older Web browsers can understand. Of course, it’s all built by Webpack, with as many unit tests as I could remember to write (not enough).

The whole thing is hosted on Heroku, which had just introduced pipelines. So I check my code into Git and pushed to GitHub, where Heroku notices the update, pulls it down and builds it on my staging server in the cloud. After that, I try my changes in staging, which is as close to production as you can get. In fact, when things check out, I simple move that build to production using a little button on the pipeline. Easy.

Also, I installed PaperTrail in order to see the last few days of logs. After all, I want to know when people are using the game.

Even if you don’t write software for a living, you can see that there are a lot of parts. Many of these parts are moving, too, in that they are supported by communities of active developers who keep changing things.

So I am used to having to upgrade components rather frequently. But stepping away for a bit changed my perspective. Here are a few of the shocking revelations.

  • JavaScript is losing favor to TypeScript.

TypeScript is a superset of JavaScript that introduces type safety, a much-needed measure of control, to JavaScript. Now you can ensure that an object conforms to the shape that is expected. All the flexibility of a scripting language with the safety of something that gets compiled, like C, C# or Java.

  • Redux sucks.

Redux is the “recommended” approach to managing client-side state. A Redux store is a good place to keep facts that you want to show in the UI. For instance, the fact that an edited scene needs to be saved might be recorded in a Redux store.

The truth is that the Redux approach is hard. I’m not just whining about it. It adds overhead and comes with a mental tax. I waited as long as I could before adding it into StoryTime because I knew that it would slow me down. Now every time I return to add something else to the store, I have to remember the gymnastics to make it happen.

  • Bootstrap looks horrible.

I mean it’s okay, but not really. Certainly not with my limited skill at styling. Also, I tried using reactstrap with its React-ready components for applying Bootstrap styles, but the documentation was insufficient for someone in a hurry. I had to keep looking at their code to know what was ready and how to use it. Bah.

  • Node is on to version 10.

You can still use 8, although it’s now on long-term support (LTS). Might as well start thinking about upgrading and what that will mean.

  • auth0 has a pricing model that will crush me beyond 6,000 (or so) users.

I knew this going in, but figured I would adjust if I ever got close to that level of adoption.

Any one of these gaps might not be worth a full-scale rewrite. Yet combined with the realization that I have been building a prototype and that no one has even tried to write a story-game yet (aside from my dad. Thanks, Dad!), I feel completely at ease walking away from this PROTOTYPE and starting over with a better idea.

Once again, prototyping has saved my sanity. I get to chalk it up to learning and decide what to try next.

 

 

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s