The path to the alpha version

Successful campaign on KickStarter (KS) didnt stop the madness. I was still going through large number of interviews and more importantly I started to get a lot of questions from indie developers, especially about our campaign. I tried to help as much as I could, even though it took a lot of my time. Fortunately the interest receded after some time, so we could fully re-focus on the project.

We had to start working on alpha now. Originally we wanted to finish before the end of the year, but in the end we finished later. A bit.

O fourse we had a roadmap – a graph with all the important project milestones put in a timeline. Basically it was an extended KS roadmap, it was close to the original tech demo roadmap, but due to time press a lot of features were moved to alpha.

Fig 0 – Kickstarter roadmap

Graphics-wise we had to focus mainly on individual characters and tweak the environment. Characters need to be built in an universal fashion, so they can be “dressed up” with as little effort as possible. Lot of textures and models had to be spruced up as well. With increasing magnitude of models processed, problems started to brew unnoticed (or rather ignored) – we should have started to build an internal asset database, catalogue each prefab (model + texture), adjust layers, add tags, etc., so the prefabs work well with each other. Alas this has not happened and lot of assets were used as they were purchased, making a lot of mess.

Our programmers had a major challenge ahead as well – moving from Unity 4 to Unity 5 engine, which allowed for a lot of new features, but also required a lot of work. This upgrade brought significant improvement of graphics, our world looked more realistic (as much as fantasy world can be realistic) and better overall thanks to new shaders. Another feature that needed programmers attention was an internal game design tool (tool we used for dialogs, quests, creating items, etc.).

Alpha development was influenced mainly by enviroment graphics that was ready at hand. Issue was that as Daniel found new asset items he was doint various completely random locations that didnt have virtually any connection to each other, so we had to adjust the game design. Originally we wanted to have several quests to show a city location, exterior location, a few dungeons and battlegrounds. Expectation was that we will start – quite logically – with game opening. The outcome was different, the alpha version contained two locations – village of Manto, where the player gets quite late in the story and the woodlands near the village.

Fig 1 – Original project roadmap

Unfortunately we made the same mistake as with KS demo. A lot of stuff was hand tailored for alpha instead of building universal components usable throughout whole game development. So again we ran into a development dead-end. The only positive thing on this was that our lead develper gained a lot of valuable experience. In the long run I see this as the first nail in the coffin.

We needed to stop the development completely and rethink basically everything, as it started to become clear that we cannot continue in such a chaotic way. We needed to do a complete asset revision and redo most of the scenes from the scratch. We did not, we did continue to patch things up on ad hoc basis.

Alpha itself was received rather positively, even though the audience was rather small, only about 50 backers tried it out. We had complaints as well, mainly the scale, whcih was limited. Inventory was pre-defined and could not be altered much, there was one quest and two locations. In general the alpha was of a larger tech demo size than full fledged alpha (as alpha should have at least the core features up and running, which our did not). At least the combat system was received enthusiasticaly, which was great for us, as that was the pillar of our alpha.

What was ciritcized heavily were dialogues, this was hard for me personally, as I wrote all of them. The only sensible way was to swallow my pride and find someone who will write dialogues for us, which we did without delay. During the alpha version preparations our translator – Honza Říha – left the team and didnt want to give us explanation.

In hindsight I can see that the reason were disagreements between him and Dan, which were growing by day. Our greatest beef was a contract, even though we all pushed for signing a contract clarifying roles, responsibilities, financial questions, etc., Daniel was strictly against. He claimed that he has successfuly finished a lot of projects without a contract and usually when there was one, things ended up in failure. Well…

The aplha was out for the public, but we knew that we could hardly consider it more than a larger tech-demo. As a matter of fact we should have known at that point that we are walking down the bad road.