Terrain Engine - Simply efficient and detailed voxel world generator

Another thing we are wondering about is the improvements you guys are proposing on top of TE. Some of these things could be very useful to existing customers. THAT I can see paying for, but we are Source already so maybe things get a little tricky there. Would be interested in what options, if any, there may be there for existing source customers to buy into the new improvements coming from your end. Of course there could be a bit of a mess with trying to make sure versions match up as well - both Dyox and Junglee having a synced REPO or Git or something similar could really be in improvement on the update/distribution process. Source users already have Construction Kit. However, with the new proposed voxel building tools that may be coming, that could be a paid for add on for all TE owners. If the older customer base doesn't want to pay for the new tools then we still have construction kit to fall back on.

 

Exactly what I'm thinking. Right now the idea is access to a github repro for all open source options which Dyox+Junglee will be working off of. The problem is deciding how to handle old customers who do not want to pay for this support option as there will be some heavy changes on our end and merge conflicts are just going to slow down the show.

Still trying to figure it out a fair process for everyone.

 

I have some questions.

If I just want to use the asset as a map editor and for texturing and object placement (and depending on performance possibly converting to a static mesh later), how suitable is it for this purpose? I read that "editor mode" is currently disabled or something to that effect? Are there any tools to shape and alter terrain after the initial generation through the node-based process?

Also, how far along is the building construction / city generation stuff at this point?

 

Sorry for the late reply @ncho

1) Right now generating the world in real time in editor mode is not possible. We're looking into this feature as we move forward.

2) You can generate the world with different settings while running the game in real time then save the world as unity terrain or a terrainengine file. UnityTerrain will not have marching cubes / destructible terrain, but Unity Terrain will allow you to paint / customize the generation. So yes, you can have object placement + texturing randomized then you can shape it into what you want after.

3) Building Construction is the same as it is in the videos on youtube. City generation is a separate project from Terrain Engine which will not be included.

 

OK, but e.g. what is shown here as far as the editing goes:

I mean, if this is not done in the editor but ingame, then what is the intended workflow for Terran Engine for developers to begin with? I do all the editing work ingame, and it then gets saved and reflected in the editor when I stop gameplay?

Is there a video that shows a sample use case for Terrain Engine with a terrain created and edited from scratch?

 

The terrain that gets edited in this video was generated from scratch:

 

I'm just waiting for the released of the proof of concept of the galactic edition, a mock no man sky would be cool. Though I wonder how much flexibility there will be to add on top generated circulation data with the close source, Dyox don't seem to grasp this new technique at all.

 

Mind sharing some resources for top generated circulation data concepts? Didn't quite follow.

As per everyone else - Do grab your copy before prices go up this Saturday!

 

"To add on top" + "generated circulation data"
Ie being able to generate "coherent and continuous" data for circulation (path) between cells and chunk. The basic version is to use something like part of the worley noise process, ie generate data inside the cell and at the immediate neighbor hash, then use that data to create implicit connection between the cells. This data can be modulated by other noises, parameter, mask, etc ... It allow for locality of generation much like worley noise and perlin (ie no large dependency to prior state and generation in finite time).

The simplest implementation is to create points inside a cells (could be POI, objects, mask, whatever can be spawn in the cell or chunk), connect the points together and find the path to point in immediate neighbor (modulated by terrain topology inside the same cell or chunk. The rules of the path generation and point can be modulated to parameters to select various pattern and their variations, like the pattern inside cities. It would allow to make cities of arbitrary sizes (up to infinity) while being coherent and under control by the modulation of input parameter (such as the underlying terrain). It also work for any structures that have heavy connectivity in general.

It has advantage as many popular techniques prior to this one are based on growth, simulation, and have problem with control and dependency to early state, they can lead to tie breaking case and dealing with exception, they end up being harder to manage in a setting of procedural generation with infinite terrain and less flexible. They are also harder to implement and manage in the long term.

The thing is I'm working on an algorithm with infinite persistent plausible NPC, that can interact and travel, live and die across many generation, generated using similar method than typical space hash methodology like perlin (basically adding a time hash and a population hash), so that for any tile I can just query what agent is traveling through it without dependency on time or space or a huge list stored in memory of past events.

If I don't need to reinvent the wheel and focus on that part that would be a god send when I'm ready to kickstart that project after the one I'm working on. I'll pay the whatever the price is needed.

EDIT: BTW I have realized that's how no man's sky compute jump on its galactic map! but in 3D volume! It's a big box of 4096x4096x256 spatial hash with a list of 512 positions per hash (in which they may spawn or not a solar system), for each of the 256 universes.

 

Prices have increased on our website. We're bound to hit the Unity asset store late November! Hopefully in time for turkey day . Stay tuned fore more updates.

 

As many others i am interested on Terrain Engine. Though i have to say i find your preorder policy quite "strange", given the fact that there isn't a single video on your website showcasing the actual tool, its editor windows , its workflow, or any documentation.. Also the fact that the first image you encounter on your website is still a witcher 3 screenshot from gdc doesn't help. In my opinion you should do something similar to other tools (like gaia for example) where the introductory price was available for the first 1-2 weeks after release.

 

@moure Thanks for your feedback.

Think of it like a kickstarter. The project as we got it ~45 days ago, was no where near something we can easily make video and share to you. It's a powerful senior engineer friendly tool that requires a lot of tinkering to get working.

We're changing that for you and videos will be available as we get to a comfortable stage of stability with the new UI/UX that's being built.

We'll update the image soon since it wasn't effecting our pre-order sales, and working on images that would really show off the engine.

As per introductory prices, the price won't change but - we will be announcing a special introductory package for all previous owners and people who purchase in November. Thanks for the suggestion!

I do understand your concerns and look forward to addressing them in time. It's a monster of a code base that is coming together to be an amazing tool to be used by the masses soon. Unfortunately we couldn't prepare the content fast enough to sell the engine properly today, but by setting up the website in it's current state, it has helped us raise money for dedicated engineers to work alongside Dyox and soon with the Terrain Engine community in building this into an amazing tool.

 

@disturbingon3 thanks for the answer

I will keep checking out your progress. I am one of the users that are mostly looking for a replacement for the unity terrain system, don't care much about procedural endless terrain or realtime editing, but would like to take advantage of the ability to sculpt overhangs and have a good vegetation/object batching system. I am looking forward for some terrain from scratch workflow videos, cheers!

 

I agree, I have no idea what I am actually getting for the price or how the workflow is...I just hope people who are waiting for some more clarity regarding the actual usage of the asset are not going to be screwed by another massive price increase once videos of that nature are actually getting released.

 

I bought TE before the date , so looking forward to seeing the new UX/UI being a UX/UI designer for web and games for AAA and indie, I can be very critical of others UX/UI, so cant wait to give my input to make TE better. in that regards,

 

@ncho Totally understandable. It's going to take some time to take this toolset and make it extremely user friendly for people in the unity community. We're investing a lot of money and time to do that.

Discounts for early birds is definitely a risk for them diving in to a tool which is coming out soon with zero videos on it. But the support has been massive and we've been able to scale the team to have people fully dedicated to making this tool awesome.

As of today, it's taking much more time than expected to prepare everything for November 1st and we'll probably continue to discount it with no-revenue limits until it's as user-friendly as all the other tools on the asset store.

We'll have both demos and examples that can be purchase in the near future. I totally understand the lack of preparation for selling the asset, but this pre-order has given us faith to scale the team to by fully dedicated to creating something amazing for you all.

@recon0303 - Looking forward to you feedback! 1st round of it is coming from the V11 version of Dyox. It's much better than the current TE V10, but we already have alot of ideas to improve it once we get all the basics working at full speed.

 

Hi,

Missed the early bird sale earlier, so just checking that
- is the current indie price going to stay (after nov 1st)
- or will there be one more discount round
- or there's no planned discounts in near future (so should just buy the current one while can)

 

The current indie price will stsay after November 1st. The closed source version is behind schedule, but we will be releasing what we have so far on the 1st.

Discount rounds will happen in the future, but there will never be a discount for 'unlimited support' and 'unlimited revenue' deal once we close this discount.

If you plan on making a game and don't want to get capped by revenue, I highly recommend buying this deal to save $300 USD in the long run for the indie closed source version.

My guess is this deal will be available until the end of the year. There's still a lot to be put together on the closed source version to make it perfect. Due to the immense support from everyone, we've been able to hire dedicated engineers to work on it full time. Looking forward to sharing more news about this in November.

 

Eh? I got an email saying you are not releasing, anything, so which one is it??? I don't mind a delay as I rather you delay then give us a bad product, but I want to be clear here. Thanks.

 

The email I got says it will be released November 7th at its current state; so it might not be quite complete due to them being a bit behind.

 

ya that is my point, what was said here, is not the same, so I wanted to be clear. Don't matter to me if they delayed and I have no interest in a refund.. I can wait, we aren't making our voxel based game right now anyways. We are booked up with other types of games for a bit.

 

To clarify everything - We have released the Terrain Engine closed source. It's still a work in progress and we're documenting everything.

We've added a new team member working full time on the terrain engine as well now. Everything is moving much quicker and we're scheduled to have the entire tool documented with examples by the end of the month.

We are also working on an example with Nature Pack assets which can be used if you purchase the nature pack art on the asset store. We'll share more information of this soon with screenshots. The demo will be readily available online!

 

Will the closed source be on Asset Store? I got a source from Thomas, just wanna purchase another closed source to join your community support and docs.

 

Hi, we are preparing an submission to asset store recently.
we have an official website about the engine for your reference.
https://www.terrainengine.io

 

We are happy to share an real case using Terrain engine. We are looking forward to share more interesting and amazing news.

Hurtful Toad Demo
Presented by Denny Butts (slack @hurtfultoad)

• Features
  • Implement Enemies AI and UFPS on top of Terrain Engine
  • Youtube:

 

I agree with this completely. Leaving out prior customers is a poor solution. I understand you received no direct revenue from us from your purchase, but there's no way any of us are going to pay MORE into this without first being provided with meaningful documentation, tutorials, and the ability to make TE actually WORK. Take a look through the Slack thread--many of us are having frustrating and ongoing issues with the height map editor. I want to be able to make this product work before I even consider paying more for support.

To add a bit more, I'd be happy to pay for some kind of ongoing support in the future once my prior purchase is up and running. I expect TE to be updated for many years and I can see paying for support after initial use is established. Right now, though, I honestly have no idea whether or not we can make TE actually work or if we'll have to abandon it.

It's in Junglee's best interest to keep prior customers happy. No support for them means not only irritated (and vocal) TE users, but also a loss of potential future revenue. But I need to see the value of support from you before I will actually pay for that support.

 

Here's another way of looking at it and then I'll shut up. We prior purchasers all have sunk costs on a product whose proper use so far seems to be eluding all of us. There is no cost to Junglee to provide support to us! You can't look at it in that we didn't purchase it from you. That's not our fault or your fault, that's simple timing.

It costs you nothing to help us, and will absolutely generate a lot of goodwill, so the upside for everyone is positive. Not helping us also costs you nothing but has a large potential negative downside in pissing off the previous client base for no real reason. There's zero cost to you either way. Why not pick the option that has upsides for both?

 

Thanks for your suggestions and I think they are good. As the prior customers and you are well familiar with TE, your experience definitely will help us to make the engine better and better.

We do need your help and care your concern. I will raise this issue to my boss and response to you asap.

 

Just got the feedback from boss. We are happy and love to give support to the prior customers. We are still establishing the support services and the support is not prefect right now. As you know, Djox is not so easy to get in touch sometime. We will give back to you once the support is well established. We are making effort every day and night. If you like, we can update you about our progress in someway.
P.S. We do care your case and all the case similar to you.

 

It's very kind and good move of yours. The main motive is building a better TE community. So, can prior customers join https://www.terrainengine.io/dashboard/ now?

 

lol, the dashboard is empty right now. It is planed for users to login and download the source, but we are not using download. Therefore, It is empty.

Alternatively, would you like to join our slack channel?

 

Thank you for the prompt response. I appreciate the effort you're showing toward both old and new users, and I think it will work out well for everyone in the long run.

 

What did you use for your AI/pathfinding for the wolves in this setup?

 

This is the demo created by the TE users, so we are not sure he used which ways in his AI/pathfinding. Next time, when he is online, I can help to ask him about this.

I guess he is used his own AI/pathfinding and not the provided one in TE.

 

I can not buy it on that side and i read that the reason can be seen in the blog but do not find it there.
Any infos?

 

this no longer exist, I would PM, the developer, as there was issues between the developer and the site owner of that site...

 

Yeah, I'm sorry, but your absurd price point is far beyond reasonable. Considering the fact that I can buy all of this engine's functions in different Unity assets for less than $200, that is exactly what I am going to do, and what most people are going to do.