Vapourware Sui Generis + Exanima Early Access

[Gumsmith]

Guh, finish the later levels before you start putting new locations into the old ones.

 

[Technomancer]

9/1/23

We've got everything up and running again and we're back to testing and adjusting things affected by AI and pathfinding changes, which can both be a bit unpredictable and difficult to test. We're only really looking for obvious problems right now, an early test version would get us your feedback and a better opportunity to notice things that may not be quite working as we intended. We're also making some updates to our content development pipeline with the new pathfinding and going through our backlog of issues both on the development side and player bug reports. From our testing so far the basic pathing data and player movement seem to work really well, which is great as it's now fully automatic and we don't really have explicit control over it anymore. One notable issue we finally managed to solve is a problem with leg physics which caused jittery and unstable movements, loss of balance and in extreme cases, since the last update, a leg to become permanently "broken".

Not much else to say at this time, we should be able to get started on some public test builds soon, but we can't be sure just yet that there isn't something time consuming that we think needs to be improved first. Once a test build is out we plan to update it with further improvements and some new features like basic environmental / fall damage to characters.

Have a great week!
-the BM team

 

[Technomancer]

16/1/23

It's been a busy week with testing, fixing and improving things and we got through almost everything on our radar and a few things we discovered along the way. We've got a couple more small additions to make and we want to do a quick write up explaining the changes, what they should allow and how you can help us test and improve things further. If everything goes well we should manage to get a test build ready within a week.

As we mentioned before, pathfinding is quite an open ended problem and not just a matter of finding the most direct path. At the core of the new pathfinding is a much more detailed and accurate 3D model which supports a lot of things, but another important aspect of it that we've been focusing on now is the AI logic for how to use this data. There's really a lot of potential tricky cases the AI can face and different ways to tackle them. We've only been testing a subset of difficult cases we're aware of and tuning things accordingly, but so far we've been impressed with the AI's ability to choose intelligent paths and tackle difficult terrain, especially considering that everything is generated automatically on the fly and there's no human designed element to it. This already contributes to making the AI feel a lot smarter which is important to us, but there's many related things that can probably be improved further so we're looking forward to hearing your thoughts and feedback.

There's going to be a fair amount of back end changes in this test build, but be aware that it won't include brand new content as we want that to come together and in a generally polished package so you can experience it properly.

Have a great week!
-the BM team

 

[Technomancer]

23/1/23

We've got a release build up and running, but we've been reluctant to pull the trigger as there's a few things we're still working on improving and we'd like to avoid being flooded with feedback about things we're already aware of. At this point we're going to write some release notes and probably update the test branch later today. Over the next few days we'll be patching in a few things that aren't quite where we'd want them to be yet. This is mostly AI behaviour and movement stuff which shouldn't be critical, but we do think needs some final touches.

The release notes will cover what's in the test build in detail and what kind of feedback we're looking for, so there isn't much more to say here. There will be an announcement on discord linking to them, which will include a password for the test branch. This is mostly so that people check out the release notes and don't inadvertently receive the update without knowing what it is. As with most first releases after a lot of back end changes it will probably have some issues, this is very much for early testing purposes and not a proper update.

Have a great week!
-the BM team

 

[Technomancer]

Quick update: I've been struggling with various weird AI behaviours we found when testing the release build, which is really difficult and time consuming stuff to analyse and work on. I also wasn't feeling well the last couple of days which didn't help. I've definitely solved some issues, but I'm still aware of at least one thing I need to fix and it's just really not productive to release a test version with known issues I'm already working on resolving. Will keep you updated.

 

[Technomancer]

Exanima 0.8.5 beta test​

OVERVIEW

We've made extensive changes to major game systems to support some new and upcoming features and content, we've also made several important engine improvements and fixes. This test release is to make sure everything works as intended, get some early feedback and tune the new systems for the upcoming major update release. We will continue to update the test branch with some additions, further improvements and in response to feedback over the next weeks. Note that as a test build there will be some unresolved issues and incomplete features, it is strictly for testing and early feedback, it is not a preview for the full update and does not include the new content.


NEW PATHFINDING SYSTEM

The most significant change in this test is a major overhaul to a more advanced and complete pathfinding solution. This doesn't only affect how AI navigate the environment, but also player movement.

The new pathfinding system is based on 3D maps with multiple layers and various more detailed information, rather than a simple 2D map dictating where you can and cannot go. These maps are fully generated on the fly and support dynamic updates. This means you can now walk on any surface you're able to access without artificial restrictions, and the exact shape and orientation of objects matters. For example, you can walk on top of a table if you can get on it, or over it if you flip it and so on. You can create dynamic paths by say putting a plank over a gap and AI can now understand this too. AI should also understand specific elevation and the importance and danger of drops of different heights etc. These multiple layers mean we can now introduce more verticality and complexity to level design.

Some further improvements have been made to player movement in an effort to make obstacles feel less "sticky" and movement more responsive and smoother.

Every part of pathfinding and movement is new; this is quite an open ended problem that concerns many aspects of AI behaviour, it's not about finding the shortest path between two points, but making intelligent choices, tactical decisions and is a low level component of many basic interactions. There's a lot to evaluate and we may need new solutions to new problems. We are aware of a few things already and will be making some improvements over the coming days, but we're generally looking for feedback on anything that you think seems odd or needs improvement.



NEW AI BEHAVIOUR FRAMEWORK

The AI behaviour system was completely overhauled, not really to change existing behaviour, but to support more interactions, specific behaviours and overcome some limitations and issues with the old system.

Although we tried to make some general improvements along the way, we deliberately tried to replicate existing behaviour, even some things that are perhaps a bit odd and we intend to review. This is in an effort to separate intentional changes from implementation errors, as the AI functions on a delicate emotion system that has already undergone a lot of tuning. What we're hoping to see is AI behaving essentially as it did before, so if you see it diverge significantly from what you expect from previous versions, this is the sort of thing we'd like to see reported. Once we've established the baseline we can look at introducing some actual changes and additions and move the discussion to what can be improved.

Golem behaviour has received some important fixes, but there were a number of issues and some improvements to make, so let us know what problems you encounter with golems.



EXPERIMENTAL COMBAT MOVEMENT CHANGES

After the recent big changes to combat movement we felt movement was just a little too snappy, particularly when changing directions, in a way that didn't seem very realistic and was perhaps a little detrimental to the flow of combat. We've made some experimental changes to make movement somewhat smoother. This however does not represent a final version of movement, as some general changes to physics that are in the works mean we will have to tune movement again.

We have also fixed a very long standing issue with leg mechanics that could cause many issues from just jittery leg motions to sudden somersaults, other unlikely acrobatics and, since the most recent update, legs to enter a permanent broken state. These issues should hopefully no longer occur and negotiating very uneven ground should generally work better.



Brief changelog for Versions 0.8.5 Beta:

• New multi-layered 3D pathfinding system
  
  
• New AI behaviour framework
  
  
• New thaumaturgy UI supporting multiple domains and Potential mechanics
  
  
• Significant graphics performance improvements
  
  
• Physics performance improvements
  
  
• New "ultra" shading setting with specular global illumination
  
  
• Global illumination quality improvements
  
  
• Fixed some unbalanced lighting on Steam Deck
  
  
• Fixed latency issues with vsync enabled
  
  
• Experimental combat movement changes
  
  
• Fixed some graphical issues with new AMD drivers
  
  
• Improved cursor selection for small objects
  
  
• Improved cursor selection for inventory items
  
  
• Fixed a bug causing Confusion to have unintended permanent effects
  
  
• Fixed dialogue closing after every player response in many cases
  
  
• Many important bug fixes

Please use our Discord server (https://discord.com/invite/Z5aV2bF) to report any issues and give feedback, you will find a thread dedicated to test builds called "Testing - Experimental" in the #Exanima channel. To access this test version you must use the Experimental Steam branch by first entering the beta access key: BMXATestAU01


Thanks!

Bare Mettle

 

[Technomancer]

Nothing exciting content wise on this patch, its basically just testing tech (so its pretty buggy).

Some cool UI stuff that hints on things to come made it in, however.

UI wheel for all six schools on thaumragy in left corner (Exanima will only have 2 until SG develops more):

 

[Technomancer]

Madoc also posted this. Apparently this is how busted end game psionic may look like (though likely with way less stuff unlocked), "when you have achieved ultimate power". I believe the cloudy thing in the center is absorbed potential you get from slaying other casters, Highlander style.

 

[Dadd]

Has he ever talked about how much magic there will be in this game and Sui Generis, and how powerful they could be? Some of the mind control spells I've seen look interesting and strong.

 

[Technomancer]

Has he ever talked about how much magic there will be in this game and Sui Generis, and how powerful they could be? Some of the mind control magic I've seen looks interesting and strong.

Yeah but its far away so only rarely. I know there will be no generic fantasy stuff like fireballs. And even though with Energy you can cause fire to start, it happens because you control fundamental forces of nature like electricity. Or magnetism for example. Or rays of concentrated light aka lasers with Light school. Magic is supposed to be extremely strong, both for the player and NPC thaumaturges (this is in part why they are hated and feared in game world). I believe the plan is to make it so that player char is really only gifted in one school which is their primary one and they can only dabble in others (Unknown main character from Exanima is a rare prodigy who can fully master two schools). I did read about hybrids and jack of trades types possible too, where you sacrifice the power of specialization on one school and can do all magic but never can reach the highest levels of power specialists can, finding strength in versatility instead. What else..ah, there will be specific combination spells unlocked by a perk that meld two different thaum schools into a powerful hybrid effect.

 

[Dadd]

Has he ever talked about how much magic there will be in this game and Sui Generis, and how powerful they could be? Some of the mind control magic I've seen looks interesting and strong.

Yeah but its far away so only rarely. I know there will be no generic fantasy stuff like fireballs. And even though with Energy you can cause fire to start, it happens because you control fundamental forces of nature like electricity. Or magnetism for example. Or rays of concentrated light aka lasers with Light school. Magic is supposed to be extremely strong, both for the player and NPC thaumaturges (this is in part why they are hated and feared in game world). I believe the plan is to make it so that player char is really only gifted in one school which is their primary one and they can only dabble in others (Unknown main character from Exanima is a rare prodigy who can fully master two schools). I did read about hybrids and jack of trades types possible too, where you sacrifice the power of specialization on one school and can do all magic but never can reach the highest levels of power specialists can, finding strength in versatility instead. What else..ah, there will be specific combination spells unlocked by a perk that meld two different thaum schools into a powerful hybrid effect.

Thanks. Has he talked about how skill-intensive spellcasting will be? It appears spellcasting is much simpler than melee combat in terms of controls and how it's affected by the environment, from the video uploaded two years ago.

 

[Mr. Magniloquent]

I have played around with Thaumaturgy in the dungeon mode. It tends to not work on anything I want it to work on. Effectiveness appears to be linked with how accurately you keep the cursor over the target mind (actual mind, not just the creatures body) and how long you hold the casting command.

The demons are immune to mind magic. I think golems too. I have had little luck with skeletons, particularly the more advanced ones. When an effect does work, it lasts just a few seconds. Sir is supposed to be weak to confusion, but it has never worked for me.

Thus far far I mostly just use the "Detect Mind" skill because it shows me if something is behind a wall or door. Conceal mind doesn't seem to work against skeletons either, which it should.

The magical concepts seem excellent, but they don't yet work. That or I am fundamentally misunderstanding them.

 

[Technomancer]

The demons are immune to mind magic. I think golems too.

They are not completely immune but it is largely ineffective due to their alien minds.

Sir is supposed to be weak to confusion

Not really. You are basing that from information from a scroll but it was not a veiled game hint but part of world building. The only thing you can surmise from it, is that caster who wrote was far more powerful Mind school practitioner than your character.

 

[Gumsmith]

It's always been funny to me that they decided to add mind tricks first when the game is set inside of a dungeon filled with mute corpses and uncontrollable monsters.

 

[Technomancer]

It's always been funny to me that they decided to add mind tricks first when the game is set inside of a dungeon filled with mute corpses and uncontrollable monsters.

It does seem weird from player's perspective but Madoc does stuff in order that makes sense to him from programming side. Since SG's prime feature is complex AI naturally he decided to tackle a version of it in Exanimas controlled environment first. So developing mind magic in parallel made perfect sense because it is all AI. Now when focus shifted on finalizing physics damage model and reimplementing fall damage back to the game, force thaumaturgy is in the spotlight. Other thaum schools are way more simple in development terms because its mostly about design and VFX without any complex programming backend to support them.

 

[Gumsmith]

Has he ever talked about how much magic there will be in this game and Sui Generis, and how powerful they could be? Some of the mind control magic I've seen looks interesting and strong.

Yeah but its far away so only rarely. I know there will be no generic fantasy stuff like fireballs. And even though with Energy you can cause fire to start, it happens because you control fundamental forces of nature like electricity. Or magnetism for example. Or rays of concentrated light aka lasers with Light school. Magic is supposed to be extremely strong, both for the player and NPC thaumaturges (this is in part why they are hated and feared in game world). I believe the plan is to make it so that player char is really only gifted in one school which is their primary one and they can only dabble in others (Unknown main character from Exanima is a rare prodigy who can fully master two schools). I did read about hybrids and jack of trades types possible too, where you sacrifice the power of specialization on one school and can do all magic but never can reach the highest levels of power specialists can, finding strength in versatility instead. What else..ah, there will be specific combination spells unlocked by a perk that meld two different thaum schools into a powerful hybrid effect.

Thanks. Has he talked about how skill-intensive spellcasting will be? It appears spellcasting is much simpler than melee combat in terms of controls and how it's affected by the environment, from the video uploaded two years ago.

One example of Body thaumaturgy Madoc once mentioned is stopping a guy's heart by holding your reticle over his chest for several seconds. Mind thaumaturgy is also really fiddly since you need to click on a dude's head to cast it so it looks like a lot of spells will be tricky to use in the thick of combat, especially against groups with ranged weapons.

It's always been funny to me that they decided to add mind tricks first when the game is set inside of a dungeon filled with mute corpses and uncontrollable monsters.

It does seem weird from player's perspective but Madoc does stuff in order that makes sense to him from programming side. Since SG's prime feature is complex AI naturally he decided to tackle a version of it in Exanimas controlled environment first. So developing mind magic in parallel made perfect sense because it is all AI. Now when focus shifted on finalizing physics damage model and reimplementing fall damage back to the game, force thaumaturgy is in the spotlight. Other thaum schools are way more simple in development terms because its mostly about design and VFX without any complex programming backend to support them.

Yeah that does make sense. Has he mentioned implementing fall damage or are you inferring from the work on Force thaum?

 

[Technomancer]

Mind thaumaturgy is also really fiddly since you need to click on a dude's head to cast it so it looks like a lot of spells will be tricky to use in the thick of combat, especially against groups with ranged weapons.

Yeah, you don't really want magic to completely supplant the physical combat.

Has he mentioned implementing fall damage or are you inferring from the work on Force thaum?

He did mention it, supposed to be the big deal of Force update. Original damage model for fall damage was pretty broken and you could crack your head open tripping on a chair as evident here for example:



So it was disabled because it needed a complex rework.

 

[Technomancer]

30/1/23

It took us longer than expected to get a test build out as we discovered a number of problems with AI behaving strangely. This is because there were very extensive changes to AI systems besides pathfinding, even though for the time being these should not have a very noticeable impact. It's out now, and it's been a surprisingly smooth first release, with few issues reported that we were mostly aware of / expecting or isolated and probably easy to fix. Amazingly we didn't need to do any urgent hotfixes, that has to be a first for a release like this. We've already gotten through some fixes and when we've gotten through a couple more we'll update the build.

We've had some good feedback and useful reports (thank you for that!) and the response to the changes has been very positive. We'll hopefully get through fixing the remaining issues quickly and then we'll gradually introduce some further improvements and additions to the test build. Either way we'll keep an eye on feedback and player submitted footage to gradually adjust and tune everything based on what the new features allow.

Check out the release notes for the test build along with instructions on how to access it here:
https://www.baremettle.com/forums/index.php?threads/exanima-0-8-5-beta-test.18222/

Both the new pathfinding system and AI behaviour framework represent major development milestones. As the game has evolved and grown in scope we've done several (mostly unplanned) big overhauls of engine systems which do represent big improvements, but don't really add much that is new by themselves. From this point on we feel we're finally back on track with things that we actually planned to do and just adding features with a more direct impact on the game. This is both exciting for what it brings to the game and also for all of us to be working on adding new things rather than redoing what we did before. Good times.

Have a great week!
-the BM team

 

[Technomancer]

Exanima 0.8.5.1 Beta is live

This brings a first round of fixes, mostly to cut down on reports of known issues and help home in on what remains. We are aware that a number of issues exist with AI pathing behaviour which we're working on.

• Fixed crouch not working

• Fixed movement in certain places being blocked for no reason

• Fixed some bugged archways in markets

• Inventory selection feathering now works accurately as intended

• Fixed a crash when AI characters progress to a new arena match round

• Fixed golems awakening when threatened

• Fixed over bright lighting on Intel graphics

• Compass overlay now fades out when thaumaturgy UI is shown

 

[Technomancer]

6/2/23

The test has been live for a week now and we've had a lot of detailed feedback and reports. We've since mostly been focusing on improving the AI's awareness and ability to traverse difficult or dangerous ground as well as avoid putting themselves in bad situations where they might fall somewhere they don't really want to be. We've had a lot of useful insights from players attempting all sorts of things and got some very nice results. We've also made some further improvements to player movement to solve some rare cases where you might find yourself stopping or hesitating with potentially dangerous results.

We've been fixing a lot of things, we already updated the test build once and will update it again soon. At this point most remaining issues are left over from previous versions or a result of not quite complete features. We'll be working through these too, but things are looking good overall. As with any changes on this scale we'll be tweaking and adjusting things over time, getting a head start on testing like this seems like a very good thing.

We've got a lot lined up already for another update to the test branch, but we're working on finishing up a few more fixes and improvements so further testing can be more meaningful and we get more known issues out of the way.

Have a great week!
-the BM team

 

[Technomancer]

13/2/23

We've updated the testing branch again with a lot of fixes and improvements. Most of this was related to pathfinding and movement, but we also took a deep dive into the parry and block system, resolving not only a new issue but a few long standing ones. Some of these new systems will need some tweaking and we'll keep an eye on feedback, but overall things seem pretty solid now.

We've been working on the environment damage system and we plan to introduce this into the testing branch very soon. We did have such a system in the game before, but we disabled it as at the time the physics engine was far less stable than it is now and some odd things could happen. The new system is much more refined either way and should be able to handle things gracefully even if the physics were to misbehave. It can handle locational damage and compound impacts, damage from objects hitting characters considering the masses involved and should generally be quite well behaved and realistic, but as always we'll tune it as we see it in action. This system will be particularly important with the introduction of Force, as that will introduce various ways to violently propel characters and objects.

There will be some smaller additions and we'll keep an eye on things, continuing to fix any issues old or new, but this will be mostly a wrap for this test version as we focus on things that support new content and the full update in general.

Have a great week!
-the BM team

 

[Mr. Magniloquent]

Technomancer So when do you anticipate release of force thaumaturgy? I don't know if it's just because of the types of enemies in Examina, but the mind thaumaturgy has been dreadfully underwhelming.

 

[Technomancer]

Technomancer So when do you anticipate release of force thaumaturgy? I don't know if it's just because of the types of enemies in Examina, but the mind thaumaturgy has been dreadfully underwhelming.

Hard to say, usually I estimate some distant date thinking I am being generous and it comes like four or six months after what I thought anyway. Well this year for sure but when exactly? My guess would be in two-four months. At the very least insider closed build.

As for mind thaum, its not even finished. Supposed to be easier to control party of undead with mass trust and similar stuff. Mass cast is not implemented and the last mind spell still remains a secret (for most).

 

[Mr. Magniloquent]

Technomancer Since this game is more of a pseudo-physics simulation, how are these less mental spells handled? Combat damage and force relate to the body type of the character, but is it known how mental scores/resistances are represented? I have done tests where I will perfectly target an eligible enemy and hold down the cast button for as long as it takes to illicit an effect (or I run out of mana). Sometimes it works, often times nothing.

 

[Technomancer]

Technomancer Since this game is more of a pseudo-physics simulation, how are these less mental spells handled? Combat damage and force relate to the body type of the character, but is it known how mental scores/resistances are represented? I have done tests where I will perfectly target an eligible enemy and hold down the cast button for as long as it takes to illicit an effect (or I run out of mana). Sometimes it works, often times nothing.

That's pretty obscure under the hood stuff, you best ask something like that from madoc on discord or steam. He is pretty good at explaining it and its usually not as simple as you would likely assume.

But I don't really understand your question? I found effects to be working consistently enough. At least on humanoids. On creatures, I just never bother. For humans I know that their mental state is a huge factor but I don't recall in what way. Probably extreme mental states change things quite a bit (panic, rage). Not really sure, never got that invested in mind myself.