Three Big Days in the Trolley Tracker Project

From Thursday evening until Sunday night the trolley tracker (www.YeahThatTrolley.com) worked for both trolleys with no down time. This is the first time we’ve unlocked the 100% uptime achievement. It seems like as good a time as any to write down some of the big highlights from this year and half completely volunteer project.

July 18th, 2014 — It’s over before it starts

I sent the below email to Greenlink (the folks who run the buses and trolleys in Greenville) because there was an article in Greenville News saying Greenlink just made a large investment in a software that would track the buses and trolleys.

Mark,

I volunteer with a newly formed Code for Greenville group. The group’s goal is to help solve civic problems through code. We had our first meeting last month and one of the projects many folks were interested in was developing a way to track the trolley in realtime.

We’re meeting again tonight to discuss the project further and I just happen to read the article in the Greenville News about the work you guys are already doing on that front. Do you have a couple of minutes to chat? I’d love to learn more about the project and share what you’re up to with the group.

Also, if the trolley tracker is not the right project, I’d love to learn a little bit more about projects that Code for Greenville might be able to help with.

We continued to work on this project for two main reasons:

1. Greenlink responded and took the time to meet and was open to letting us work on the project.
2. Ryan McAllen made a very convincing argument. To paraphrase, let’s make Greenlink choose. If we build a better product and we’re going to offer it for free why wouldn’t Greenlink use our product. At the very least riders can choose the app they want to use.

As of today, the Code for Greenville trolley tracker is the only app available for folks to use and that’s mainly because we were willing start and confident we could get it to work.

April 23rd, 2015 — Android Adam Arrives

We made some good progress on the API and iOS app side of things, but all our location data was dummy data. We never had a working solution for, you know, tracking the trolley.

Screen Cap of from the Yeah, That Trolley Imagine That

Thankfully, Adam Hodges commented on this ImagineThat call for help.

I have some code I could contribute to this project… I wrote an app that transmitted signal/location data as a background service for my thesis. You are using Android devices for the trolley tracker?

We needed help on several fronts and we asked a bunch of different folks for help. Adam raising his hand and showing up to meetings to help was a real key to getting this project kickstarted.

July 26th, 2105 — Mike, Nice Work

Probably the first part of the project to ever be completed was a prototype of the API needed to share location data to the web and mobile apps. It was built by Dan Fields and done mostly in Clojure. It was working great, but needed some admin type screens built and some general polish to be ready for production. The problem, Dan’s wife was about to have a child and he knew he was going to be short on time to contribute. Mice Nice offered to help. After a couple attempts at working on the Clojure project Mike sent this email a few days before a Code for Greenville meeting.

FYI — Since Visual Studio 2015 was announced, I thought I’d see what it would be like to develop an API using that. I think it came out OK. I haven’t approached Dan about this yet. But since I’m a latecomer, it’s quite possible that my test project will never be used. I’ve got quite a bit more information already loaded into my API. (The public server isn’t deployed yet, but a local software house is eager to host it on their spare server)

When Mike arrived at the meeting he had a test that was damn near fully formed. There was much rejoicing from everyone, including Dan, and we moved forward with an ASP API.

With that, we had all the major pieces working for our little project to get finished.

There is a good amount more to write about this project and there is a large number of vital folks involved that weren’t mentioned here. These three moments just stood out as major milestones for a long project with many important milestones.

TL:DR

1. We weren’t afraid to start the project and had confidence we could build something useful.
2. We asked for help over and over again and eventually the right people raised their hands and got on the bus.
3. The right tech stack for us was the tech stack people were willing to build. We discussed a hundred different options, but the winning stack was the stack people actually built the damn thing in.