• Hello Guest,

    Welcome to the Survive the Nights community forums. Please feel free to share your ideas with the team and discuss current game features. Please do not report bugs here, use our bug tracker for that.

    The team at a2z(Interactive);

Post Stress Test 2.0 and Early Alpha

Status
Not open for further replies.
Thread starter Admin #1

Jayty

Developer
Developer
Joined
Jul 26, 2014
Messages
282
Likes
1,129
Points
93
Location
Hull england
Hello everyone!

Just a brief check in from me to let you know where we are and what we’re doing after Stress test 2.0. It probably won’t end up being that brief but I’ll keep it as short as possible while keeping it as informative as I can. I’d like to apologise for the lack of changelog or detailed updates since the test and I will explain the reasons below.

On behalf of the whole team I’d like to thank everyone involved in the last stress tests. We gathered so much crucial information that just wouldn’t be achievable without you all! Unfortunately, I didn’t get to play during the stress tests all that much other than jumping on occasionally to test stability but I know JB was there with you guys a lot and keeping you all informed.

Overall, the stress tests did everything we needed them to do. The stability was what we hoped for considering there were so many new features that hadn’t been tested on a wide scale. Of course there were issues that we hadn’t anticipated and toward the end we discovered an issue which caused the servers to crash.

Moving past the stress tests we had internal meetings to discuss what is working well, what isn’t and what we feel could go be going better. A week was naturally dedicated to utility, project cleaning and internal structure organisation. Some of the bigger limitations that we face is the fact that the team is small and widespread in terms of JB and Vinny being in NY and myself and UK team being here as well as the need for better tools to aid development.

We decided overall that a mix up of people’s job roles was needed. Ultimately, we decided that it was more efficient to give Andre the Lead Designer position since he’s here in the UK with the other designers and since he’s been heading up the environment design for some time now. We all felt that a community manager was greatly needed to help bridge the gap between us here internally and all of you who’s ideas we put into the game. Naturally we all agreed that @_JB is the perfect man for the job which he already pretty much does anyway. He will now be able dedicate just a little more time to the community. JB will still be working on world design with the addition of Jos joining him. 3 weeks in and I can happily say that it’s working great. With the addition of the utilities described below, our whole design workflow has increased dramatically. Overall though, it doesn’t really matter what position people hold here at a2z. We’re a small team and all opinions and thoughts are considered. We all have the same ultimate goal of delivering an epic game to you all.

I’ll summarize below some of the main things we’ve worked on since the stress tests and also explain our plans moving forward.


UTILITY
As mentioned above, we decided to spend some time working on internal tools which would help speed up and make development quicker and easier for everyone involved in the project. We already had a whole bunch of tools that we have made through the development of STN. Most of them were fragmented and individual in their design. It was a time waster and pain to find what you’re looking for and use it efficiently. Getting all these features into a single utility window with simple navigation and controls that made sense to everyone working with them was a huge priority. I’ll list some of the bigger issues we had and how we’ve managed to get around them.

The problem - One of the biggest issues developers face when using Unity is the lack of ability for multiple people to edit scene files at the same time. This results in multiple edits of the same files and a whole bunch of overwriting each other by accident. The stress test area was manageable although it was fair to say that everyone was pretty fed up with it. This was something that we felt needed time dedicating to since there would now be another person in addition to Andre and Joe working in the scenes on a daily basis.

The Solution - I decided to develop a tool which allows the “world” to be edited in bite sized pieces (1km x 1km pieces to be exact). This new tool as seen in the picture below allows people to assign themselves chunks, load them efficiently, work on them and then save and submit them to source control as normal.



The problem - This issue was kinda two things rolled into one but they were resolved using the same utility so we’ll call it one for the sake of this write up. The lack of ability to place props and general design in a really efficient and quick manner was a big problem. Also Unity doesn’t support nested prefabs which we desperately required since the implementation of more complicated interactable objects like doors with locks. Nesting prefabs is very crucial due to the need to be able to edit each object on mass individually from their “master” object. Without the ability to do this, it meant that we had to nest some objects inside prefabs. This meant that the link to the original master item was gone. Small things like this cause huge headaches when you find something needs editing in multiple places. It leads to inconsistencies and lots of time trying to put things right. Since we’re a small team, we try to focus on spending our time in the most efficient way possible and spending hours going over things that will undoubtably need changing again just isn’t scalable.

The Solution - The first thing here was fixing the whole nesting dilemma. I did this by working around some of Unity’s built in features while still being able to utilise their awesome prefab system. The short of it is that we can now have nested, individual prefabs which can retain their links and be grouped together with ease. The second part of the tool is the ability to quickly search for, preview and “paint” on these new items onto the terrain with random rotations and auto snapping to the curvature of the surface being placed on. With the addition of a few other little mistake preventing features, this tool takes the complexity out of level design and just allows the designer to focus on designing an awesome world without having to worry too much about anything else.



The problem - Moving around a 64sq km island with the standard methods within unity can prove to be a real pain. It’s amazing to think that minutes of every day could be wasted from jumping from spot to spot or moving around as you’re designing to gain different perspectives.

The solution - A tool for saving and moving the scene camera to predetermined spots was something we knew we needed for some time and now that it’s implemented it makes things way more streamlined and enjoyable to work with.


In addition to the things listed above, we have implemented a whole selection of options tailored to our exact needs for changing and altering things on mass, from adding and removing prefixes, to fixing pivots on models. Again, all things that save headaches and time when designing.


DESIGN PROCESS IMPROVEMENTS
One thing we felt could be much better was the draw distance that the player sees. Although it wasn’t necessarily bad during the stress tests, we were hitting some upper limits in terms of what we were able to do. Andre and myself decided to dedicate some time to really optimising the way items load and are seen in the distance beyond the capabilities of what is possible with standard Unity features. We soon discovered that with some changes to our previous methods, we could not only increase the range the player sees but also optimise by a considerable amount!

I won’t go into too much detail about how it’s done but I’ll let the pictures beneath show the changes from Stress test 2.0 and what you’ll see the next time you log in. You can also see improvements to the way the roads are laid out to give a more realistic feel.



With addition of the new item placement tools and better loading/management of items, you can see below how design can be placed more densely allowing for greater detail on a whole. This design probably took 1/3 of the time is would have taken pre stress test 2.0



WE HAVE LIFE!
One of the most exciting things we’re now finally adding is AI. This has all been in the works for a long time from animating, network connecting, to behaviour and load testing. There simply wasn’t any point in adding it before now due to the need for testing backbone and core functionality for database persistence and all the other stuff you’ve tested up until this point. We’re happy to say that the next time the game is playable, you’re going to be getting chased by zombies, blowing them up (just like in the original concept videos), as well as hunting highly skittish deer and most likely chasing around chickens trying to gather eggs for your bases chicken coops!



Summary
We’re all feeling very confident about the progress over the past weeks and can’t help but be excited for the release of the game. Now that our core back bone functionality is tested and implemented, it really opens up our options in terms of what user submitted features we can implement and the speed in which we can do it. I think the coming months will be an exciting for STN.

Thank you everyone for your support and commitment to a2z and STN. Keep submitting epic ideas and we’ll crack on implementing them. The next time you play the game we’ll be in Alpha and the servers will be staying up!
 
Status
Not open for further replies.
Top