Earlier this year, we exhibited at Game Developer Conference (GDC) in San Francisco. It was our first dedicated conference focused on game developers—especially those building on Unreal Engine 4—and we had a ton of fun exhibiting and meeting studios and developers. We also reinforced our thesis that crash management for game studios presented some unique challenges our offering was well positioned to improve.
Fast forward a few months later, and we had our first game studios using our platform to revamp crash management process, increase user retention rates, and improve the efficiency of their development team. One of those studios,, the creators of >SQUIDS FROM SPACE, was kind enough to share their experiences.
Quick Pit Stop
Let’s first review our primary offering for game studios. We provide the easiest-to-integrate crash management solution with the most powerful debug capabilities for games running on Unreal Engine 4. I’ll spare you the details—you can check out more in our docs—but in short that means studios can submit crash data to Backtrace with nothing more than an .ini file update. Once integrated, developers now have a centralized server that captures, stores, and processes the Unreal crash archive. This allows us to do all sorts of interesting things, like deduplication, extracting key values for investigation, and incorporating this data into the triage process. In an earlier blog post, we described this offering in detail.
We were thrilled when Fun Bits signed up for our free trial shortly after our trip to GDC 2017. Their game, SQUIDS FROM SPACE, is a multiplayer action-strategy game where players unite to collect resources, fortify bases, upgrade weapons and capture the enemy’s Commander. The game is developed on Unreal Engine 4 for the PC, and is currently available as Early Access on Steam.
Greasing the Wheels for Early Access
Earlier this year as Squids in Space neared the Early Access deadline, Fun Bits determined they needed to address some challenges with crash detection and monitoring. Crashes, while infrequent, were also unpredictable, and having access to the necessary information to investigate or triage was not always available. Ian, one of the gameplay programmers, acknowledged that they “…didn’t want to be flying blind with regard to how or why users or servers we crashing.”
For Fun Bits, the primary factors in selecting a crash management solution include:
- Cost of integration
- Cost of service
- End-user privacy
- Reliability of reporting end-user crashes
They ruled out other offerings—including developing their own—and selected Backtrace. As a small team, Fun Bits understood that integrating an external solution would free up their critical bandwidth to focus on what they do best: developing great games. According to Cameron, Lead Programmer, “The cost of using Backtrace has been greatly outweighed by the personnel cost of creating and maintaining that kind of service internally.”
Reducing Crash Collection Headaches: That’s Epic!
Backtrace provided Fun Bits a revamped process for detecting and managing crashes. Their process evolved beyond talking across the room, sending emails, and moving files across shared folders into centralizing crash data into a single source of truth. Backtrace allows Fun Bits to collect critical crash info, including hardware data such as graphics card, to better understand crash patterns and root cause. They now know with certainty how many servers are crashing—previously a tricky proposition due to their hosting environment—and triage based on quantifiable crash severity.
To further improve automation, Fun Bits integrated the crash data collected by Backtrace into Slack notifications. Now, crash notifications appear directly into their Slack channel, which instantly alerts their team to start investigating. Backtrace acts as a force multiplier for the Fun Bits development team by allowing them to act faster with more reliable information.
It is nice to see crash dump automation like Backtrace come to market in a way that helps reduce a lot of headaches, and also helps automates processes for smaller teams who do not have the luxury of a full QA Staff on board to manage, track, and report crashes on a daily basis. – Chris Millar, CEO of Fun Bits
I Can’t Believe It’s Not Harder
Right?! It’s almost… unreal… that you can easily elevate your crash detection and analysis with a simple configuration change.
Have your own UE4 war stories to share? We’d love to hear them. Send a note firstname.lastname@example.org you too can be featured in a blog post.