State of Production #10
Hey everyone, my name is Tim Perreault and I’m the Executive Producer at New World Interactive. This past month has seen a lot of exciting things as we released our biggest update yet, Operation: Nightfall, and moved our team into the Clean Sweep update 1.7.1. Clean Sweep has been five solid weeks of time dedicated to making general improvements to the game’s health through bug fixing, code optimization, and an engine upgrade. While the patch notes you’ll see when it’s released will give you a lot of the details, I wanted to speak with a few team members to give you some behind the scenes insight into some of the work we’re doing and why.
[h2]Key Points of State of Production #10[/h2]
[h2]1.7 Release and Hotfixes[/h2]
With Operation: Nightfall being released on June 16, we’ve been busy trying to fix some of the issues that resulted from the engine upgrade we undertook. The highest priority issues remaining are a proper fix for the suicide bomber crash (we have temporarily removed ragdoll physics on suicide bomber corpses to mitigate this issue while we continue working) and fixing the remaining cases of floating magazines after resupplying. We’ve been working with Epic to resolve these issues and hope to have them all resolved soon. We will continue to put out Hotfixes as soon as fixes for issues are available.
[h2]Update 1.7.1 Clean Sweep[/h2]
Over the last five weeks, all of our developers have been hard at work dealing with community-driven requests, cleanup and polish from our backlog, code optimization, and finalizing our engine upgrade plan as a part of what we’re calling Clean Sweep.
[h3]Community Driven Requests[/h3]
While there are many small features and tweaks the community has been looking for coming in this release, one important one I’d like to discuss is drivable vehicles in co-op Checkpoint mode counter-attacks. I’ve asked our Lead Game Designer, Michael “Mikee” Tsarouhas and Level Design Director, Jeroen van Werkhoven to talk a bit about why vehicles were removed, why they’re coming back, and what that will look like:
Why did we remove vehicles from co-op? Jeroen: The initial problem we had with vehicles was performance. We had a lot of issues with how they behaved when there was high ping. Additionally, the AI ran into problems driving into walls and not being able to consistently arrive at objectives which made for a poor player experience and hurt the Fun Factor™ which is really important for us.
Mikee: In addition to what Jeroen said, we felt removing vehicles in co-op would improve performance and optimization of the game. Removing the insertion vehicles in the intro sequence did help with this; however, removing the driveable vehicles that the enemy bot AI drive and the players can steal to drive themselves didn’t make that big of a difference. Not as much as we’d hoped. So we’ll be keeping the insertion vehicles out to keep those performance gains, but putting the drivable ones back so players can enjoy using them.
How are vehicles coming back to co-op? J: We’re bringing back the vehicles back where they were in the previous iteration of coop vehicles. The big difference is we’re looking at a new system that will allow the level designers more control over the directions where they’ll come from, where they’ll stop at an objective and overall they will be more reliable and know where to drive. The old system was more of a gamble. Sometimes they knew where to go but they didn’t know how to get there. With the new system we can actually add new locations into the level itself where the vehicles can go. The new system will give level designers more options on where vehicles will go when driven by AI which will make the counterattacks feel more exciting. Removing vehicles from counter-attacks removed a lot of the challenge for the players. It was a lot more difficult when a vehicle was attacking you while you were on the objective. It was more like a cinematic experience.
M: Not to mention it was just more interesting, having vehicles attacking.
J: Yeah, definitely more interesting.
Final Thoughts M: I think this is actually a really good example of how our team needs to be sure it keeps its ear to the ground and that we have a good feedback loop with the community. The reason we’re doing it is because the community asked for it and because it made sense with our production schedule and the health of the game. We removed something, we thought it would have a positive effect with performance and it didn’t as much, the community said they wanted it back, so we’re bringing it back hopefully better than before.
J: I’m with Mikee, we need to listen to our community. With the vehicles, we learned a lesson about how much more investigation and testing we should do before making decisions. I’m really excited that vehicles are coming back, I think it’s an important part of the co-op experience.
M: Even though it might not seem like a natural and clean progression, and you can go back and say “You shouldn’t have removed them in the first place,” well frankly this is a part of what we’ve been doing with Sandstorm where we’re not afraid of being experimental and trying different things with our game. If we need to revert something we’ve done and try making it better, we’re not afraid to do it. It’s a really good example of iterative game development and more evidence that we’re not done with Insurgency: Sandstorm.
[h3]Backlog Cleanup and Polish[/h3]
Since Sandstorm’s release there have been plenty of ideas that ended up on the cutting room floor for one reason or another. Clean Sweep was an opportunity to act on a few more minor things we'd always meant to do or would have liked to have, not to mention address a couple of community requests. While we’ve taken on lots of bugs and polish during the last few weeks, Mikee and I spoke with Ben “Mr. Brightside” Turtell about one special feature that he worked on, a transparent magazine for the G36K:
Where did the transparent magazine idea come from? Ben: It’s a thing that’s been bugging me since we first implemented it. I haven’t seen anywhere that a G36 doesn’t have a transparent mag and I wanted it to be true, as close as possible, to how it should look. It’s also really cool to just see your ammo going down and have that visual cue of your ammo going down.
Mikee: It’s both aesthetically cool and gameplay cool. You can just look at your mag and see how much ammo is available. So you’re saying you’ve been wanting to do this since release, because the G36K was one of the guns we had at release.
B: Yeah. I don’t think when we released I had the time to look into this option because there were a lot more guns waiting to go in and a lot of other things that needed to be done. Clean Sweep really gave me the time to think about things I’ve wanted to update and this was definitely one of the top ones. I don’t know if we can mention the upcoming guns (Producer note: Not yet!) but it’s going to be really useful for at least one of the future guns we have coming. Even for some of our current guns, we could potentially use this feature on.
What were some of the challenges with creating a transparent magazine? B: Oh god [Laughs in Weapon Animator], probably the rig for the spring and the follower and all the bullets in there. With the extended mag there are 35 bullets in there, each with their own bones each with individual bullet meshes weight painted to the bones. The whole thing is animated with the spring moving and the bullets moving when they get to the top. I had never really rigged a spring or weight painted a spring so that was a challenge. Also just the logic for the whole system and getting it to work. I had a lot of help from the team working out those details. When I first implemented it we couldn’t get the info of how much ammunition was in your magazine when you reload because that all gets updated when the magazine is fully inserted, so we had to get an extra function from code so we could finalize it.
[h3]Optimization[/h3]
One of our big goals with 1.7.1 Clean Sweep was to dedicate some time to review older systems to see how they could be optimized, similar to the efforts spent late last year to optimize our art. We’re aiming to increase the overall performance of the game, especially on lower-end systems. The general approach we took was to run profiling tools on specific gameplay recordings and look for areas where we noticed the heaviest resource requirements or hitches. These issues were investigated, documented, and assigned to other developers to fix. The original investigator reviewed the same recording to confirm performance improvements were successful.
[h3]Engine Upgrade[/h3]
As part of our ongoing improvements we’ve finished upgrading to Unreal Engine 4.25. There were a number of great features that we’re already taking advantage of including some of the improved Unreal Insights which helped drive our optimization work. While this engine upgrade wasn’t a trivial effort, it did result in a net gain of features and has opened up possibilities for the future of Sandstorm as well.
Upcoming Releases
With Clean Sweep mostly completed and getting finalized for a release in early August, the team is starting to transition into working on the 1.8 release. As much as I’d like to Tell you about what’s coming in 1.8, it will have to wait for SoPs in the coming months.
That’s it for this month’s SoP, until next time, keep your stick on the ice.
[h2]Key Points of State of Production #10[/h2]
- 1.7 Release and Hotfixes
- Update 1.7.1 Clean Sweep
- Community Driven Requests
- Backlog Cleanup and Polish
- Optimization
- Engine Upgrade to 4.25
- Community Driven Requests
- Upcoming Releases
[h2]1.7 Release and Hotfixes[/h2]
With Operation: Nightfall being released on June 16, we’ve been busy trying to fix some of the issues that resulted from the engine upgrade we undertook. The highest priority issues remaining are a proper fix for the suicide bomber crash (we have temporarily removed ragdoll physics on suicide bomber corpses to mitigate this issue while we continue working) and fixing the remaining cases of floating magazines after resupplying. We’ve been working with Epic to resolve these issues and hope to have them all resolved soon. We will continue to put out Hotfixes as soon as fixes for issues are available.
[h2]Update 1.7.1 Clean Sweep[/h2]
Over the last five weeks, all of our developers have been hard at work dealing with community-driven requests, cleanup and polish from our backlog, code optimization, and finalizing our engine upgrade plan as a part of what we’re calling Clean Sweep.
[h3]Community Driven Requests[/h3]
While there are many small features and tweaks the community has been looking for coming in this release, one important one I’d like to discuss is drivable vehicles in co-op Checkpoint mode counter-attacks. I’ve asked our Lead Game Designer, Michael “Mikee” Tsarouhas and Level Design Director, Jeroen van Werkhoven to talk a bit about why vehicles were removed, why they’re coming back, and what that will look like:
Why did we remove vehicles from co-op? Jeroen: The initial problem we had with vehicles was performance. We had a lot of issues with how they behaved when there was high ping. Additionally, the AI ran into problems driving into walls and not being able to consistently arrive at objectives which made for a poor player experience and hurt the Fun Factor™ which is really important for us.
Mikee: In addition to what Jeroen said, we felt removing vehicles in co-op would improve performance and optimization of the game. Removing the insertion vehicles in the intro sequence did help with this; however, removing the driveable vehicles that the enemy bot AI drive and the players can steal to drive themselves didn’t make that big of a difference. Not as much as we’d hoped. So we’ll be keeping the insertion vehicles out to keep those performance gains, but putting the drivable ones back so players can enjoy using them.
How are vehicles coming back to co-op? J: We’re bringing back the vehicles back where they were in the previous iteration of coop vehicles. The big difference is we’re looking at a new system that will allow the level designers more control over the directions where they’ll come from, where they’ll stop at an objective and overall they will be more reliable and know where to drive. The old system was more of a gamble. Sometimes they knew where to go but they didn’t know how to get there. With the new system we can actually add new locations into the level itself where the vehicles can go. The new system will give level designers more options on where vehicles will go when driven by AI which will make the counterattacks feel more exciting. Removing vehicles from counter-attacks removed a lot of the challenge for the players. It was a lot more difficult when a vehicle was attacking you while you were on the objective. It was more like a cinematic experience.
M: Not to mention it was just more interesting, having vehicles attacking.
J: Yeah, definitely more interesting.
Final Thoughts M: I think this is actually a really good example of how our team needs to be sure it keeps its ear to the ground and that we have a good feedback loop with the community. The reason we’re doing it is because the community asked for it and because it made sense with our production schedule and the health of the game. We removed something, we thought it would have a positive effect with performance and it didn’t as much, the community said they wanted it back, so we’re bringing it back hopefully better than before.
J: I’m with Mikee, we need to listen to our community. With the vehicles, we learned a lesson about how much more investigation and testing we should do before making decisions. I’m really excited that vehicles are coming back, I think it’s an important part of the co-op experience.
M: Even though it might not seem like a natural and clean progression, and you can go back and say “You shouldn’t have removed them in the first place,” well frankly this is a part of what we’ve been doing with Sandstorm where we’re not afraid of being experimental and trying different things with our game. If we need to revert something we’ve done and try making it better, we’re not afraid to do it. It’s a really good example of iterative game development and more evidence that we’re not done with Insurgency: Sandstorm.
[h3]Backlog Cleanup and Polish[/h3]
Since Sandstorm’s release there have been plenty of ideas that ended up on the cutting room floor for one reason or another. Clean Sweep was an opportunity to act on a few more minor things we'd always meant to do or would have liked to have, not to mention address a couple of community requests. While we’ve taken on lots of bugs and polish during the last few weeks, Mikee and I spoke with Ben “Mr. Brightside” Turtell about one special feature that he worked on, a transparent magazine for the G36K:
Where did the transparent magazine idea come from? Ben: It’s a thing that’s been bugging me since we first implemented it. I haven’t seen anywhere that a G36 doesn’t have a transparent mag and I wanted it to be true, as close as possible, to how it should look. It’s also really cool to just see your ammo going down and have that visual cue of your ammo going down.
Mikee: It’s both aesthetically cool and gameplay cool. You can just look at your mag and see how much ammo is available. So you’re saying you’ve been wanting to do this since release, because the G36K was one of the guns we had at release.
B: Yeah. I don’t think when we released I had the time to look into this option because there were a lot more guns waiting to go in and a lot of other things that needed to be done. Clean Sweep really gave me the time to think about things I’ve wanted to update and this was definitely one of the top ones. I don’t know if we can mention the upcoming guns (Producer note: Not yet!) but it’s going to be really useful for at least one of the future guns we have coming. Even for some of our current guns, we could potentially use this feature on.
What were some of the challenges with creating a transparent magazine? B: Oh god [Laughs in Weapon Animator], probably the rig for the spring and the follower and all the bullets in there. With the extended mag there are 35 bullets in there, each with their own bones each with individual bullet meshes weight painted to the bones. The whole thing is animated with the spring moving and the bullets moving when they get to the top. I had never really rigged a spring or weight painted a spring so that was a challenge. Also just the logic for the whole system and getting it to work. I had a lot of help from the team working out those details. When I first implemented it we couldn’t get the info of how much ammunition was in your magazine when you reload because that all gets updated when the magazine is fully inserted, so we had to get an extra function from code so we could finalize it.
[h3]Optimization[/h3]
One of our big goals with 1.7.1 Clean Sweep was to dedicate some time to review older systems to see how they could be optimized, similar to the efforts spent late last year to optimize our art. We’re aiming to increase the overall performance of the game, especially on lower-end systems. The general approach we took was to run profiling tools on specific gameplay recordings and look for areas where we noticed the heaviest resource requirements or hitches. These issues were investigated, documented, and assigned to other developers to fix. The original investigator reviewed the same recording to confirm performance improvements were successful.
[h3]Engine Upgrade[/h3]
As part of our ongoing improvements we’ve finished upgrading to Unreal Engine 4.25. There were a number of great features that we’re already taking advantage of including some of the improved Unreal Insights which helped drive our optimization work. While this engine upgrade wasn’t a trivial effort, it did result in a net gain of features and has opened up possibilities for the future of Sandstorm as well.
Upcoming Releases
With Clean Sweep mostly completed and getting finalized for a release in early August, the team is starting to transition into working on the 1.8 release. As much as I’d like to Tell you about what’s coming in 1.8, it will have to wait for SoPs in the coming months.
That’s it for this month’s SoP, until next time, keep your stick on the ice.
Before
In Progress