While last week’s edition ended up a bit on the short side, this one needs sections 
Steam (delayed)
In week 16, we handed AirlineSim in for review by Steam. A mandatory step before any game is allowed to go live on their store. Sadly, we failed this review the first time around, mostly due to somewhat arbitrary rules around payment models: AirlineSim is free to start playing but you run into a paywall once your credits run out, which apparently isn’t something Steam allows. There needs to be “some basic free gameplay that will be available perpetually”. We are currently working out what this could look like, both internally and with Steam. Either way, this means the Steam release will likely be pushed back a bit.
Meanwhile, we continued preparing this release as if it was happening as scheduled, meaning our freelancer was putting on the finishing touches on the Steam UI components and I took care of some…
…statistics code for Steam (and beyond)
For basic account activities (like sign-up, payments, logins etc.) and for many Prosperous Universe in-game events we’ve had an events tracking database for ages. It basically works like this: Whenever an “interesting” event happens, we store this event with a timestamp and some optional contextual data in a large database. Then we can use this data to generate high-level reports on things like conversion rates, user activity etc. As an example, a report (or a series thereof) could look like this: “For all users that signed up in week X, how many are still active after day 3, how much money did they spend and what percentage got to the point of opening a station? How are these values different for organic vs. Steam sign-ups?”.
I guess you can see how this can be useful. One of the side-effects of the Steam release is that we need to be a lot more conscious of how people find their way into the game. We expect the Steam crowd to be a bit less “ambitious” than people finding AirlineSim through active search outside of a store, meaning a higher percentage might fail to get started which in turn increases the chance of negative reviews. But to implement features or changes in that direction, we first need to figure out how and where people get stuck.
So last week, I wired up a whole bunch of events for AirlineSim so we can see how many people, for example, create or reset an airline, open a station, create a flight plan, lease an aircraft and so on, and so forth. Since this touched a lot of places around the codebase, including very old ones, I constantly had to resist the urge to rEfaCtoR EveRyThInG! And I succeeded…mostly. So now we’ll collect data on this for a while post release and then analyse it to coordinate next steps.
A (not so) brief return to maps
With all that very technical non-game stuff I had to work on last week, I really needed a brief change of scenery. So I wanted to “sit down real quick” to fix a few quirks in our new maps that I didn’t like. Here’s an example of what I mean:
While technically correct, this likely isn’t the representation anyone would like to see for a path from LAX to SYD. Instead, it should look like our old maps, with the path nicely centered on the map:
As it’s so often the case, this lead down a rabbit hole of other quirks I discovered, all having to do with the anti-meridian (aka date line). So once I was done and after a spending a few hours more than originally planned on this, my test panel looked something like this:
There certainly are more edge-cases (literally) that I haven’t uncovered yet. Once the new maps go live all across the game, please don’t hesitate to let me know which ones you’ve found!