2015/12/16

Banner Saga 2 Launches

This spring and summer, we had the good fortune of launching the Banner Saga 2 across all our supported platforms.

For Banner Saga 1 it took us 2 years to launch console, and 9 months to launch mobile. This time around we have the time window shrunk to 3 months for console and 5 months for mobile.

We are working on a few more updates for Banner Saga 2, including bringing our localizations to console, and continuing with 3 more translations for Spanish, Portuguese, and Russian, which are underway.

We've been cleaning up and improving our engine and asset pipeline. One of our ongoing technical tasks is to upgrade our story writing capabilities. For Banner Saga 1 and 2, we've relied on Inkle's InkleWriter Beta. This is a fine tool, but is no longer supported and does not scale well to our game. We've studied Inkle's new Ink language, and their new tool, Inky. These are really neat, but to retrofit into our game would be a bit too difficult. We want our new solution to be backwards compatible with our old conversations, and vice-versa. So, inspired by Inky, we've created our own tool, which implements a subset of the Inky features, and in a way that can be 1:1 mapped onto our existing conversation technology.

As part of this effort, I've had the pleasure of working from the ground up with Test-Driven-Development again. The Banner Saga was originally developed in a test-first way, but this was interrupted after about 18 months into development. Adobe AIR released an update that broke their testing framework. The primary issue was that we simply had too many tests and their framework fell apart at scale. We spent some time trying to get it working again, but ultimately had to abandon our tests. They certainly served us well during the initial 18 months of development.

Now, as part of pre-production, I've taken the time to create my own custom testing framework, a framework that I can control and scale as needed. All of the new 'Story' system is being written test-first and it sure does feel wonderful to program that way.

A few weeks ago, we had an issue with our Amazon EC2 servers running Factions. One of the servers, the one that hosts the RabbitMQ message server, had a hardware failure. The server was still running, but Amazon informed me that it could go down at any moment, and furthermore would be shut down automatically in 10 days, regardless of what happened in the interim. This may seem like a fairly simple problem, but since I haven't administered any maintenance to the Factions servers is years, I had to go through the whole process and remember how to do it, and how to configure the RabbitMQ server once it was restarted. The server was also running on some obsolete Amazon features, so I needed to figure out how to migrate into the new paradigms (specifically migrating into Virtual Private Cloud). Believe me, I took the opportunity to carefully document my steps for next time. It was fun, and made me miss doing server development!

I'd like to back up a bit and talk about Survival Mode. Shortly after PC launch, we were hanging around the office, waiting for console launch, and waiting for our holidays to start. We decided to do something productive with our time, and brainstormed various updates we could do to the game. We settled on Survival Mode, a 40-battle challenge mode that focuses on character progression and tactics, with permadeath and lots of items. This was essentially a 2-week game jam on our end, followed up with several weeks of testing and polish. We ran an open beta on Steam for Survival Mode, and many of you participated. This open beta was a great experience for us, and we really felt invigorated by our close collaboration with our community. This experience is one of the things that made us realize that we need to invest more in our community, the result of which is that Khatie has now joined us as Community Manager on the Stoic team.

Survival Mode was a good technical exercise, as it forced me to formalize the structure of these 'sub-sagas' that can be entered into from the main game. We added item drops and permadeath in battle, which caused us to upgrade the reporting on the battle 'Resolution Screen'. This screen (the wide red banner), now reliably displays deaths (injuries in regular play), items discovered, and achievements earned.

Once we all reconvened after holidays, we recognized the fact that we now have a bunch of out-of-game stuff to share with everyone. These include the Warbands miniatures board game, the double LP Vinyl albums of the soundtrack, the digital soundtracks, and the Gift of Hadrborg, James Fadeley's great Banner Saga novel.

Aside from this, we often have a desire to communicate news and event information with our players. As a result, we have created a 'news ticker' GUI that can sit at the top of the Start Screen. This news ticker allows us to push news updates to clients without the pain and expense of a full patch. We had done something similar for Factions, however this new one looks much nicer and can contain custom images and CSS text formatting. We'll be rolling this out soon, starting with PC.

We're also working on an update to TBS1. For one thing, we would like to include the News Ticker there. As another point, we recognize that the Bellower fight is still a major problem for many players. Of the players that reach the Bellower fight, only 47% make it through to the end of the game. This is a pretty terrible completion rate, and we are looking at ways of tuning that battle to mitigate. We are also looking at tuning the supply availability in Chapter 6. This last chapter has always been unintentionally devoid of supplies. It's not supposed to be easy, but you should have enough supplies and renown to at least have the option of keeping your Morale up. We hope to roll this out in a few weeks as well.

Last week, I attended Steam Dev Days with Zeb West, our new Producer. We spent 2 days in Seattle for this event, and found it extremely informative on many topics. A major topic was of course VR. There were VR demos and tech talks galore. It is pretty amazing what is being done in this field, not least of which is Steam's opening up of the Lighthouse technology as OpenVR. What this means is that you can use the Lighthouse tech to DIY all sorts of projects, or even commercial products ranging from pet trackers to in-home high-quality motion capture rigs. The positional accuracy of these Lighthouses is about 300 microns, so the potential uses are enormous. Banner Saga VR anybody?

Thanks for reading - please share your feedback in our forums, here:

https://stoicstudio.com/forum/forum.php