David Aldridge

  • There is a good bit of old Bungie code in D1 and D2, there is some Halo (and to lesser extent Myth and Marathon) DNA threading through a 1/3rd of the client codebase. Suspects there is very little pre-Destiny code that has been totally untouched by Destiny's work. More info and logs here.
    • Bungie invented Behavior Trees when Damian Isla wrote them for Halo 2, and they still use an extended version of that pattern for combatant AI in Destiny. Extensions include heavier scripting integration, improved debugging, a much richer navigation system, and more improvements. (Reddit)
    • Very technical info about tech debt and build system here.
  • Would be a super fun* engineering challenge to get Destiny on the Switch. Main challenge is that Destiny pushes hardware pretty hard. With the game's current scope, Switch would present major challenges on memory bandwidth, CPU, and storage. It's not just about resolution and LODs. They find it really challenging to throttle CPU per-platform without disrupting the spectacular symphony of Destiny. A bug in the system that helped ship D1 on 360 and PS3 was pretty rough on gameplay and AI. (Reddit)
    • A longer read about challenges of publishing Destiny from a technical standpoint can be found here.
    • David is excited about improved load times on next gen consoles, about crossplay and other social features, and improved workflows and iteration/build times internally so that all the talent at Bungie can get closer to a model of directly pouring their inspiration into the game, reducing the time spent wrestling the game into submission (Reddit)
  • Wouldn't say Switch is impossible. But, it would be very costly and it would increase their ongoing maintenance costs, such as new content with Switch-only performance bugs. They're ultimately always trying to allocate limited resources to the most impactful set of investments. (Reddit)
    • * folks at Bungie have learned to be afraid when David uses the word "fun"
  • On Stadia: It was challenging, but fulfilling. They had to shift their internal priorities pretty radically to make it happen. It took 7 months from start to ship but ultimately it was really exciting. (Reddit)
  • Vaulting content is a bit of a mixture of technical limitations and creative direction of the game, with goal being making the best Destiny experience for players. The tradeoffs are definitely tough. (Reddit)
    • Chris:

Certainly some of the basic [reasons] are that as the game continues to grow the amount of testing resources needed grow as well. You make one change and get regression in totally different parts of the game. We all know that at some point we'll miss the bug that lets Telesto become sentient but we're trying to delay that as long as possible!

Along those same lines if we want to make a core change to some tech that say breaks all activities then we have to go back and rebuild every activity in the game. If we let the number of activities grow forever then our ability to make those kinds of bone breaking changes goes away.

You can't just "code the game better" it's not the code that takes up the most space and we already compress assets. We have a lot of smart people working on Destiny and we've looked at a lot of options and from both a technical standpoint and design standpoint we believe the content vault is the best solution. (Reddit)

Chris.

  • Engine talk:
    • To make something the size of Destiny engine, there is a lot of specialized knowledge required for everything from performance, optimization, rendering, networking, etc. (Reddit)
    • Every engine has strengths and weaknesses. Chris would not classify the Destiny engine as archaic, and he wouldn't want to build destiny in another engine. There are some processes that are slow, and they're making improvements to address the tech debt. They updated the engine from D1 to D2 and they continue to update the engine during D2 development. They'll never make a whole new engine and move Destiny to it, but any future games that Bungie makes wouldn't necessarily need to use the Destiny tech if it doesn't make sense for that game. (Reddit)
    • Chris wouldn't want to make Destiny on a 3rd party engine, and there will be people who either like to control their own tech or need something specific from their tech which necessitates writing their own engine. (Reddit)
    • Beyond Light brings significant changes to the engine, and of course they'll continue making improvements in many areas of their tools and tech. It depends on how you define at what point it becomes a "new engine". (Reddit)
  • No plans on releasing a Linux client. The support cost wouldn't make sense. (Reddit)

DMG.

  • Probably won't be talking too much about Transmog until after Beyond Light. They don't currently have an ETA on when it will be available, the team is in planning stages. They're invested in creating a good experience. (Twitter)