Information Antharion Kickstarter Relaunched

[JarlFrank]

Well, the countless hours you spent on random encounters in Wizardry 8 didn't really help the game, so sometimes less is more.

 

[quasimodo]

They just released an update with some new lighting and a some good dungeon music. Very nice!

http://www.kickstarter.com/projects/879484120/antharion-an-old-school-rpg/posts/364974

 

[Kosmonaut]

They just released an update with some new lighting and a some good dungeon music. Very nice!

http://www.kickstarter.com/projects/879484120/antharion-an-old-school-rpg/posts/364974

Uhm... that was two days ago dumbass.

 

[ajrs84]

Hey guys, Antharion is not all about combat. Creating an interesting world that feels alive is as important to us as creating a solid combat system and we'll demonstrate this in upcoming video updates. It's incredibly important to us that every inch of the world is interesting and has some purpose to it. Antharion is not large beyond the point where a tradeoff kicks in between size/quality - if this happened we'd quickly scale down in favor of quality, but is hasn't and everything so far has gone better than planned in this regard.

 

[Roguey]

I don't know about this. This is the type of game where, as long as the combat is varied and interesting enough, I want a lot of it. I loved KotC and that was all combat. This game seems to tend towards combat encounters, not unique locations, dialogue, etc. at least what I've seen to this point.

KotC also took me "only" about 25 hours and it was wearing out its welcome in the last few.

 

[mindx2]

KotC also took me "only" about 25 hours and it was wearing out its welcome in the last few.

Each to their own I suppose. I enjoyed the encounters and would have loved for the game to go on. Though it took me longer than 25 hours as I kept getting slaughtered until I got the hang of it.

 

[Kosmonaut]

New update with a video "demostrating" NPC and enemies AI.

http://www.kickstarter.com/projects/879484120/antharion-an-old-school-rpg/posts/368225?ref=activity

 

[BlitzKitchen]

Hey guys, Antharion is not all about combat. Creating an interesting world that feels alive is as important to us as creating a solid combat system and we'll demonstrate this in upcoming video updates. It's incredibly important to us that every inch of the world is interesting and has some purpose to it. Antharion is not large beyond the point where a tradeoff kicks in between size/quality - if this happened we'd quickly scale down in favor of quality, but is hasn't and everything so far has gone better than planned in this regard.

Are there any notable gameplay aspects that aren't about dungeon crawling and party combat?

 

[Hirato]

Finally got some cash, so I've backed.
I must say, the lighting/shadowing update gives me pretty strong fallout 1 vibes. I blame the music, good work and keep it up!


Will we be able to interact with the torches? (ie, to light, extinguish, or pickup)
Are there (dis)advantages for fighting in light or darkness?

As for the AI..
How does it handle a group moving to a set destination?
And what about 2 groups moving to 2 different destinations with a common intersection point in which they'd block the path for the other?

From one coder to another, I'd appreciate if you didn't skimp on technical details.

 

[ajrs84]

BlitzKitchen, Antharion is about way more than just combat and I'd refer to the above linked video as one tiny little example of this. Throughout the next couple of weeks we'll be releasing more video updates detailing non-combat facets of Antharion's world, including stuff like multiple-solution quests with tangible consequences.

Hirato, Torches don't extinguish themselves but you can light unlit wall mounted torches (one of our backers came up with this suggestion and we took it). A dungeon that's pitch black will be completely different from a dim dungeon in terms of the combat strategies it necessitates - in such a situation having a torch or having a light spell cast becomes crucial.

As for the technical pathfinding stuff, Antharion uses a preprocessed waypoint system. Basically, at some point prior to run-time we generate a bunch of waypoints, and then run an all-pairs shortest path algorithm (Floyds) on them. This creates a table containing all shortest path solutions for every pair of waypoints. The main advantage of this method over A* is speed, since all the slow calculations are done prior to run-time. Once pathfinding is active for a NPC, the pathfinder will continually update that NPCs path until he reaches his goal. Before giving a path to an NPC, the pathfinder steps through it making sure that no dynamic objects (such as other NPCs) are blocking it - in certain situations NPCs are allowed to swap positions with one another if it is in both of their interests. Hope this explanation was clear.

 

[Hirato]

Avoiding A* for speed reasons is somewhat silly. My own experience with it suggests that you could easily use it to generate thousands of routes per frame with very little impact on the framerate (my test was running at 200 Hz).
IMO, the greatest speed issues come from finding the closest waypoint to various points of interest; without an easy and implicit means of doing it (like you'd have with your grid), you don't have much choice but to brute-force it and store the results in a cache, which at that may just break the world up into smaller more manageable sectors to limit your search.

Looking at Wikipedia's description of the algorithm, it looks as though it basically points you to the next waypoint, and uses the information in that waypoint to direct you to the next one, until you reach the end.
I had thought of building a cache like this, but decided against it due to space reasons, though I was still at the "cache the full route" stage.
But this specific algorithmn would end up at around 16MB for my standard use case (2000), and 324MB in the more extreme ones (9000), for the record, A* still performs really well here.

I imagine that Antharion would have considerably more waypoints per level than 2000, so what steps do you guys take to prevent the memory/space requirements from becoming untenable?

 

[ajrs84]

Avoiding A* for speed reasons is somewhat silly. My own experience with it suggests that you could easily use it to generate thousands of routes per frame with very little impact on the framerate (my test was running at 200 Hz).
IMO, the greatest speed issues come from finding the closest waypoint to various points of interest; without an easy and implicit means of doing it (like you'd have with your grid), you don't have much choice but to brute-force it and store the results in a cache, which at that may just break the world up into smaller more manageable sectors to limit your search.

A* is very implementation specific. I can tell you that we got measurable performance increases in terms of fps when testing on an older-gen iPad unoptimized, where even a 2-4 fps gain can be meaningful (not to mention the saved memory). That said, speed was not the main reason why we went with pre-processed waypoints, it just worked better given how the engine was setup.

Looking at Wikipedia's description of the algorithm, it looks as though it basically points you to the next waypoint, and uses the information in that waypoint to direct you to the next one, until you reach the end.
I had thought of building a cache like this, but decided against it due to space reasons, though I was still at the "cache the full route" stage.
But this specific algorithmn would end up at around 16MB for my standard use case (2000), and 324MB in the more extreme ones (9000), for the record, A* still performs really well here.

That's exactly how it works, yes.

I imagine that Antharion would have considerably more waypoints per level than 2000, so what steps do you guys take to prevent the memory/space requirements from becoming untenable?

The way around those huge table sizes is in breaking the world up into a series of smaller manageable sectors, each sector with its own solution table and associated pathfinding data - then you'd just cache this data as needed. A small sector with 500 waypoints will have a solution table that's only 500kb. And a single dungeon, for example, requires way under 500 waypoints.

 

[Hirato]

Both interesting and unexpected.
I take it you don't treat each and every cell as its own unique waypoint then?

That sort of would explain why the NPC in the AI update didn't walk along the wall, but through the centre of the corridor.


Thanks for taking the time to answer, appreciate it.

 

[ajrs84]

That's correct. Antharion uses a system of strategically generated (or sometimes manually placed) waypoints, so you'll usually see them near doorways, in the center of the room, around obstacles. Another benefit of waypoints is that it's easy to control how NPCs/monsters move about in a complex dungeon corridor. Also, you don't have to deal with non-realistic looking behaviors like wall hugging that sometimes accompany the actual shortest path between cells (as opposed to the shortest path between waypoints).

Always happy to engage/answer questions!

 

[ajrs84]

We just posted a new video update demonstrating Antharion's day/night cycle...

Video Update #8

 

[Hirato]

I can't help but notice that the shadows and shading remain completely constant throughout the day/night cycle.
Looks rather off, though I suspect that it won't be noticed during actual gameplay.


May I suggest that interior windows act a bit more like a directional light source during the day time?
It's hard to tell, but it currently looks as though the whole scene becomes brighter as a sole unit.

 

[ajrs84]

Not sure if you're referring to the drop shadows or dynamic shadows. In the videos the drop shadows are constant, but in the final version they'll attenuate along with the outdoor ambient light intensity. The whole scene doesn't lighten uniformly, but the lack of directional lighting may be what gives you this impression. There's no directional lighting in Antharion, I'd love to do stuff like soft shadows with penumbras, normal mapping, etc. as I'm really into all of that stuff, however, I made a conscious decision early on to limit the scope of the graphics engine so that I could concentrate mainly on game play.

 

[Hirato]

For the shadows/shading part, I was referring to the tree outside as well as roof of the building, one side of it is always shaded brighter than the others, and the tree's shadow was always being cast in a specific direction.

May I suggest that interior windows act a bit more like a directional light source during the day time?
It's hard to tell, but it currently looks as though the whole scene becomes brighter as a sole unit.

What I meant with this was, have the windows, during the day time, light up the area nearby, whilst going much further in the window's direction than the others, like a spotlight.
Not necessarily for casting shadows, unless you wanted to.


I'm just a bit bothered by how uniform the lighting looked during indoor day/night cycle.

 

[ajrs84]

For the shadows/shading part, I was referring to the tree outside as well as roof of the building, one side of it is always shaded brighter than the others, and the tree's shadow was always being cast in a specific direction.

What I meant with this was, have the windows, during the day time, light up the area nearby, whilst going much further in the window's direction than the others, like a spotlight.
Not necessarily for casting shadows, unless you wanted to.

I'm just a bit bothered by how uniform the lighting looked during indoor day/night cycle.

Usually for 2d isometric graphics you'll color one face of walls, roofs, etc. darker than the other to exagerate a sense of depth. Static drop shadows are a work in progress and will change drastically before final release, we may even end up using dynamic shadows for trees. I tried doing versions of what you suggested with lighting the windows surrounding area but in the end it just didn't look the way I wanted. Also, you'll never notice the uniform lighting because the color/lighting changes occur imperceptibly slow over the course of many turns in the actual game.

 

[Bulba]

stop bitching about the lighting already, you sound like a little spoiled graphics whore.

 

[Kosmonaut]

stop bitching about the lighting already, you sound like a little spoiled graphics whore.

Yeah, typical bitching from weeaboos.

 

[Hirato]

I can't help it!

 

[Hirato]

They literally just pushed out update #9.
It's a video of the iPad version of the game.

http://www.kickstarter.com/projects/879484120/antharion-an-old-school-rpg/posts/371915

 

[ajrs84]

For those interested we've managed to push out another two updates. The first outlines three new rewards tiers while the second is kind of a misc. progress report which includes a screenshot of a new interface skin, a new amazing sounding combat track and a couple of new monsters.

Update #10: New Reward Tiers
Update #11: Progress Report

 

[Hirato]

Looking (and sounding) good, I've upped my pledge a little bit.