Kerbal Space Program is a pretty quirky sandbox-style space flight simulation game, currently in paid alpha development. I’ve been following and playing the game for nearly two years now, and after the release of the 0.21 update, I set up a Q&A with most of the development team. Not an easy task when a sizeable portion of the development team is spread across the globe, so before we begin let me take a second to heartily thank them all for chipping in.
First of all, let’s talk about your overhauls of the development process. I noticed that you pretty much completely revamped the development cycle, and it seems like you’ve spent half your time in the QA and Experimental phases. Can you elaborate on what you do during those phases?
Ted Everett, QA Lead (Ted): The Testing Phases are split about 40/60 between QA and Experimental, respectively. In QA, we focus heavily on issues relating to a feature’s implementation; such as feature X causing issues with feature Y. Additionally, we devote time to ensuring that each feature gets feedback from the QA Team. One feature this was incredibly important for was the SAS Overhaul that Chad [Chad Jenkins, Technical Artist] developed for 0.21, it also required extensive testing with a wide array of vessels and setups. Finally, a key part of the QA Phase is ensuring the version is stable enough to go on to the Experimental Team.
Once in the hands of the Experimental Team, the bug reporting and fixing really gets into swing. With an ever growing Team of voluntary testers, Experimental Testing is, at the core, about performing a wide sweep of the update and catching all of the issues that are present only on certain systems or are incredibly elusive.
You’ve made a pretty big deal of this new bug-fixing process, but we don’t really get to see exactly how many bugs you quash. Can you give me a ballpark figure on the amount of bugfixes in the QA and Experimental phases? Any interesting bug stories you want to share?
Ted Everett, QA Lead (Ted): Overall, in the 0.21 Testing Phase, the Team fixed 153 bugs, 72 bugfixes in QA and 81 bugfixes in Experimentals. Compared to previous Testing Phases, this was quite a high number, owing to both a rather major rewriting and optimisation of a core feature, and to an increased number of Experimental Testers. In an average testing phase, about 70-80 bugs would be fixed overall. Bugfix numbers, however, are one of the less reliable ways to monitor how well Testing went, mainly due to the fact that a low number could imply little Testing or a high number could imply that Testing was rushed. Generally, I rate a Testing Phase by how stable the build is at the end and by the reaction from the players.
As for interesting bug stories, I have a couple for sure. Here’s a link to one called the “Earthquake Bug” that I linked to a while ago. Another, equally perplexing, one is called “Explosions Everywhere” which is pretty self-explanatory. They were both simple cases of doing things in the game that might not be taken into consideration – as most of the fun bugs are.
Now that 0.21 is out and under your belt, how do you feel about this latest development cycle?
Miguel Piña, PR (MaxMaps): As a PR guy I’m mostly an outside observer but actually seeing it develop, seeing the new testing process at work and working under Jesus [Jesus Montaño], our new Project Manager, things have been going great! It’s been awesome seeing things get slicker and slicker with each release.
Felipe Falanghe, Lead Developer (HarvesteR): I have to say, it’s incredibly satisfying to finally see the space center scene taking shape and growing out of its old placeholder form. It’s an area of the game that’s remained undeveloped for much too long, and there are a lot of new and very exciting ideas we‘ve wanted to add to it for the better part of a year now. It’s just awesome to see it start coming together now.
As for the development itself, we got to see yet again how the team’s been steadily picking up a solid pace and how we’re becoming increasingly better at working together. The next updates should be even better in that regard, and that’s a very exciting prospect.
Your transition to Steam seems to have mostly ironed out your previous issues with having your site and store down for days after each update. Any words you care to share about Steam or your own distribution process?
Rob Nelson, Game/Server Developer (N3X15): Steam has definitely been the biggest thing keeping our servers online after an update: now that most people can just download the game from Steam’s very large cloud distribution network, we don’t have to resort to moving all our downloads to Amazon AWS, which can be fairly expensive. The new patcher also helps, since it only downloads what you need and has a very small bandwidth footprint. There are still some things to iron out, like how to handle user-modified files and mod distribution, but we’re working on some ideas.
It seems that you add 2-3 new team members per update. Are you happy with your rate of growth?
Felipe Falanghe, Lead Developer (HarvesteR): As with any company, there are staff changes every now and then, but the KSP team isn’t growing nearly quite as much as that. We’ve hired a few new hands to cover for others who left, but we’re not looking to just increase our numbers. We want to make sure that everyone we add to the team is the best person for the job we can find, so we can keep the team small and nimble, more like a squad of highly trained ninjas than a big army.
What are you looking for when you’re scouting for new developers?
Felipe Falanghe, Lead Developer (HarvesteR): It’s not just about raw talent. Above all, we value people who can blend into the team and work with us as part of a unit. It doesn’t matter how talented someone is if their personality doesn’t match. We don’t call ourselves a Squad for no reason. ;)
You’re based in Mexico, but you’ve got members all around the world. How do you even begin to deal with all the difficulties inherent in your setup?
Miguel Piña, PR (MaxMaps): The secret’s organization, we have to be aware of everyone’s schedule, and we’re thankful to see some of our team organizing their timeframes around ours. In the end it comes down to communication, we’ve got a constant devchat going on through an IM service so we’re all technically sitting in the same virtual office.
Felipe Falanghe, Lead Developer (HarvesteR): Even though we work remotely, we are always in constant communication. Sometimes it actually feels like we’re just a few desks away. One of the cool things about living in the future is that we get to do things like this. Who needs flying cars when you can get to the office without leaving the house? (Still waiting on the flying cars though)
Your public relations system has changed drastically since 0.20, and your community outreach has seemingly ballooned upwards and outwards. Are you happy with your new PR guys?
Chaz Chiarello, Media Director (xPDxTV): As the community has grown public relations has become one of the more important factors in making the game well known. My mission is to promote Kerbal Space Program and outreaching toward the community is very important. Letting them in on the information by giving interviews, news articles and new version pre-release videos to help feed their desire to know more. The community, whether users or press, help drive the game and that is something we’ve taken into consideration. Kerbal Space Program is for our community and we want to share it with them. Max and Bob have been key with Public Relations by getting interviews and other coverage for our game.
Do you have any issues with any particular groups in the community?
Miguel Piña, PR (MaxMaps): Not really, I personally spend a bit of time outside of our official forums community and find all sorts of input and the like. It’s hard to point at a group and say that they’re bad, if anything I’d ask for all of them to get along, but only so much we can ask for on the internet, haha.
From what I’ve seen, Felipe has a pretty clear and concrete vision for what he wants KSP to become. Have engine limitations or community outcry changed those plans significantly?
Felipe Falanghe, Lead Developer (HarvesteR): So far, the Unity engine has proven to have been a very good choice for us, and really, none of the limitations we’ve had to work around in developing KSP are something you wouldn’t also have found in other game engines. So I don’t think there were any cases where a technical limitation stood in the way of our views for the project badly enough that we had to change the concept around it. Usually what happens is that we end up changing the way we implement something at a technical level, but that will at most result in a minor change, if any at all.
As for the community and our goals, we did have one or two cases where we realized we had to make a course correction, and there was some disappointment there, as you’d expect. However, these are one or two cases in a project which from the start was planned to hold off on big design decisions, until we could get some community feedback on them. For instance, when we first released, we didn’t even know whether we were going to keep the orbital mechanics. We completely relied on community feedback to make that call.
In fact, I can’t say there was even a single time when the community opinions didn’t have an impact on our design decisions. Yes, sometimes we have to make a hard decision about where to focus next, but we’ve never made one without taking the wants of the players into consideration. What you have to keep in mind is that as the game gets older, you start seeing a big difference in what the most veteran players want compared to what a player on his first day needs to be able to get into the game in the first place. We want to make sure the game is as welcoming and approachable as possible before we go on adding features that could increase the entry barrier, so we always try to find a balance to make sure we’re focusing on the areas that most need to be developed at each time. You have to make sure your foundations are stable before you start adding a second story to your building, but it wouldn’t be much fun if we just had a parking lot either.
What’s next for the KSP community?
Miguel Piña, PR (MaxMaps): Honestly we can’t tell. At every turn and every day our community finds a way to blow our minds with incredible builds, deep discussion, suggestions and mods that help us make KSP a better game.
How do you decide what goes into each version?
Felipe Falanghe, Lead Developer (HarvesteR): After we’re done with a release, we sit down to discuss what we think should be our main focus for the next release. We have a master plan for where we want to be at in the long run, but the actual features that get assigned to an update are something we only decide right before we get to it. This lets us stay flexible, in case we realize we have to change tack at some point or another. This flexibility is also very necessary because the farther ahead you plan, the less likely you are to be able to stick to every detail of your schemes. So we leave the specifics for each update to when we get there, and we think about broader goals for the more distant future.
Here’s a personal question of mine. Ages ago, back in 0.18 when docking was introduced, many people asked if there was going to be a way to toggle or otherwise modify the magical magnetic docking attraction, such as requiring electricity to activate it. How’s the status of such a change? Is it even possible?
Felipe Falanghe, Lead Developer (HarvesteR): It’s certainly possible, but things like this are less of a technical issue and more of a gameplay decision. Some players want as much realism as they can possibly get, while others just want to have a good time. I like how the ports attract themselves when you get close enough, because that lets us have a larger target to hit, without having to resort to magically snapping the vessels together when they come into contact. If we disabled the attraction there, we’d end up with an impossibly small target to reach, which would be very frustrating to most players. Docking is plenty hard as it is without having to strike a pinhole of a target. The magnetic docking range was one of those middle-ground solutions to a case where we had to walk the line between too simple and too difficult.
Okay, here’s the big question: What’s next for Kerbal Space Program? What can we expect to see in 0.22? 0.30?
Robert Holtzman, PR (Calisker): We’ve already announced that research and development will be in update 0.22 and we’re preparing to reveal more about R&D and update 0.22 at PAX Prime. As for future updates, you can expect to see more of Career Mode, which is going to need multiple updates to deliver a completed game mode. And update 0.30, well, Jeb Kerman swore me to secrecy and there’s no way I’m going back on my word with that Kerbal.
Alright guys, thanks a bunch for taking the time out of your schedules (and vacations) to answer this barrage of questions. Any parting words?
Miguel Piña, PR (MaxMaps): Be excellent to one another.
Felipe Falanghe, Lead Developer (HarvesteR): Remember to always make sure the tail end of the rocket isn’t travelling faster than the front end.
Chaz Chiarello, Media Director (xPDxTV): If at first you don’t succeed add more boosters and/or space tape.