So I stumbled across something several months back, totally by accident. In a small project, there was a page that would fire AJAX search requests to several different websites all from one search box. After getting the street view service working, testing with search terms like addresses and cities, I began testing the other service with whatever word came into my head: butter, biscuit, bitcoin, etc. To my astonishment, google street view’s api returned an image for ‘bitcoin’ and the idea for StreetViewHaiku was born. Haikus and silly images to associate with some/all of the terms in it.
So having this idea kicking around in my head and notes for some months, it has many more bells and whistles I want to add one day, but for now it will serve as an acceptable guinea pig capstone for my time in Flatiron’s Learn program. Being the final assessment at the end of over a hundred lessons each about Rails and Angular, these requirements are technical, not practical. So if a feature seems a bit ridiculous, it’s because I needed some feature to demonstrate some technical capability, not necessarily because it’s part of what I imagined.
Tidbit about me: I like to do things piece-by-familiar-piece. Check my work. Know what’s worked and what hasn’t.
Tidbit about Rails+Angular: don’t try to minify code twice. It’s just a recipe for disaster. Also, when there are two options suggested, and one adds yet another gem to depend on and one is changing your configuration by one line, try the one-liner first. It just might work. Sure, it will slow things down a bit when your demo app’s userbase breaks a million, but you will have hired engineers to do the work FOR you by that point.
Launching on Heroku is one of the things that I wanted to be sure would go smoothly. I wanted to see a ‘hello world’ running before I committed too much time into it to scrap either the app or having it on heroku altogether. Fortunately, everything worked out on the initial launch and all of my features/demo-widgets will just be add-on’s to the basics.
Having the app on heroku gives us the added benefit of having some ENV variables already in production that we can pull into development/test or more typically vice versa. One of the biggest downsides though is that there are issues outside of my control. The Glyphicons are giving 404’s even though the files are located in the project in 2 places and referenced (what seems to be) correctly.
The other things is that I’m realizing now how silly an idea for an app this really is. So I’m going to press on, because I must have something to show at my appointment next week. However, in an employment setting, this could go one of two ways: in large, unstoppable-force style company you would press on because to push back against what will happen will only bring pain for all; or in a small, dynamic company, you could say “this is probably not going to generate enough revenue to be worth it” and they would say “you know what, you’re probably right. Move on.”
As my project, and my bootcamp, near completion, I’m realizing how scatter-brained these descriptions of my projects must seem. I might as well just record my mutterings throughout the process and edit them for silence and swearing.
There’s a live version of the app on Heroku. (I *knew* I’d be glad I spent 3 days getting local PostgreSQL working.)