http://www.kickstarter.com/projects/obsidian/project-eternity/posts/390987?ref=activity
Update #39: Non-Core Classes, Cooldowns, Attack Resolution, Damage vs. Armor and a Tileset!
Update #39 · Jan. 23, 2013

Hail, elves. It's been a while since we've had a mechanics update, so I'd like to cover a variety of topics today: the basics of our "non-core" classes, our cooldown system (or lack thereof), an update on how attacks are resolved, and another update on the evolution of our armor system. I'd also like to show you a dungeon tileset test render and some sweet shakycam of some of the combat basics running in engine.
Non-Core Classes

We've previously discussed the design of our "core four" classes: fighter, priest, rogue, and wizard. The non-core classes are the other seven: barbarian, paladin, ranger, druid, monk, chanter, and cipher. Like the core four classes, the non-core classes all start the game with two active or modal abilities and one passive ability.

When it comes to the balance of active/modal and passive options, the classes generally reflect their D&D counterparts, with spellcasters having more active use abilities and weapon-based classes being oriented toward more passive or modal abilities. Even so, it will be possible to push a spellcaster toward more passive talents and to optionally buy more active/modal abilities for traditionally low-maintenance characters. While all classes will have many more abilities as they advance, here are some basic elements for each of the seven classes.
Barbarians can use Wild Sprint a limited number of times per day, allowing them to rapidly rush across the battlefield to a distant target while ignoring hazards along the way.
Paladins have limited healing capabilities, but their Revive command allows them to instantly snap an unconscious ally awake with a large Stamina boost.
Rangers' animal companions are so closely bonded to their masters that they share Stamina and Health pools, which can be both a blessing and a curse.
Druids can Shapeshift into animal forms, gaining natural -- and some supernatural -- abilities associated with those creatures.
Monks absorb a portion of incoming damage and convert it into a Wounds resource they can use to power their soul-based abilities (such as Stunning Blows) through any weapons they use, including unarmed strikes.
Chanters begin the game with a number of phrases they can arrange to form songs with different effects. Aefyllath Ues Mith Fyr is a phrase that causes allies' weapons to emit magical flames.
Cipher powers often gain intensity as they maintain focus. Their basic Mind Jab starts as a minor irritant but can build to inflict devastating damage.
Cooldowns

Early on, some folks asked about cooldowns and both Tim and I agreed that we weren't opposed to using them in some form if it made sense for our mechanics. To be more explicit about it, the only way in which we are currently using anything cooldown-like is for per-encounter and per-rest abilities. Per-encounter abilities can be used a number of times in an encounter and are then disabled until combat ends. Per-rest abilities can be used a number of times after resting before you must rest to recover them. We've previously discussed grimoire-switching for wizards possibly invoking a cooldown. It's more likely that grimoire-switching will be limited through the inventory system and not by a cooldown.

We also have modal abilities that can be turned on and off at will, with some abilities being exclusive to others, meaning you can only have one active at a time.
Attack Resolution

I've talked about this a bunch on the forums, but not in an update. All attacks in Project Eternity compare the attacker's Accuracy value to one of four defenses: Deflection (direct melee and ranged attacks), Fortitude (body system attacks like poison and disease), Reflexes (area of effect damage attacks), and Willpower (mental attacks).

A number between 1 and 100 is generated to determine the attack rules. If the Accuracy and target defense are the same value, these are how the results break down:
01-05 = Miss
06-50 = Graze
51-95 = Hit
96-100 = Critical Hit

A Hit is the standard damage and duration effects, a Graze is 50% minimum damage or duration, a Critical Hit is 150% maximum damage or duration, and a Miss has no effect. In a balanced Attack and defense scenario, the majority of attacks wind up being Hits or Grazes. If the Accuracy and defense values are out of balance, the windows for each result shift accordingly, while always allowing for the possibility of a Graze or a Hit at the extreme ends of the spectrum.
Damage Type vs. Armor Type

We've previously talked about how different weapon damage types (Slash, Crush, and Pierce) fare against Damage Threshold (DT) in the game. We implemented that system and found that while it worked well on paper and scaled well, it was unintuitive when put into the game. It was not possible for players to make informed decisions about what weapons to use against a given armor type because doing so required making relative damage vs. DT calculations for all weapon types, i.e. having a spreadsheet open for comparison at all times.

In light of this, we are going to try a more explicit damage type vs. armor type model where armor, regardless of its DT, has a familiar weight classification: Light, Medium, and Heavy. Damage types are either good or bad against a given weight classification. When a damage type is "bad" against an armor type, it does half damage before DT is applied, making it very inefficient. Within the "good" types of damage, there's still an efficiency curve against DT for meticulous players to figure out, but it has less impact than avoiding "bad" damage types in the first place.

Energy-based attacks (like most spells) oppose a different characteristic of the armor, its substance type (Natural, Armor, or Spirit) and like damage types, have good and bad opposition characteristics. Weapon bonus damage that is energy-based is applied to the target separately, but at a fractional DT value matching the bonus damage. E.g. if a sword has a fire effect that does +15% the sword's damage, it is opposed by 15% of the target's Damage Threshold.
Tileset Trials and Tribulations

Link to Larger Version

Environment artist Sean Dunny has been experimenting with building tilesets for our dungeons. "Tilesets?!" you may be saying (or thinking). It may be a surprise, but many Icewind Dale and Planescape: Torment levels started from a tileset or modular unit base. We use these tilesets to generate basic renders for testing layout, navigation, and combat. Once we like the basic layout, we refine it by adding additional "meta" (special) tiles, modifying the tiles individually in the layout, adding lights, and of course having an artist do a 2D touchup pass.

That's all for this week. Thanks for reading!

Really informative and interesting.

This really does sound like it's going to be a fantastic, sprawling RPG.

Anotherr update:

As some of the keen-eyed among you noted from last week's update, there was an unfamiliar portrait in our work-in-progress tileset screenshot. We read the debates and viewed the Blade Runner-esque enhanced images that followed with interest. Good work, sleuths, the character pictured is, in fact, an orlan. This orlan is engaged in some important work in one of the Dyrwood's busiest cities. Here's the full-sized portrait for your continued speculation!
Ziets on Pantheon Design

George Ziets

Hello all. This week, I will be writing the update, and we’ll be starting to talk a bit about world development. For a designer, this is the fun part, and it’s a surprisingly rare opportunity.

On all my previous Obsidian projects, the team has worked with an established IP (intellectual property) like the Forgotten Realms or Fallout. We’ve always had a wealth of existing lore to draw upon – cities, towns, characters, history, gods, etc. Sometimes we’ve worked in an area of the world that hadn’t been seen in a CRPG before (as in NWN2: Mask of the Betrayer), which gave us the opportunity to extrapolate beyond what was already established... but otherwise, we were working with established material and trying to be true to existing lore.

In contrast, Project Eternity is an entirely new setting that we’re creating from scratch. And at the start of development (around the end of the Kickstarter campaign), we didn’t have much more than what KS supporters have already seen: a map, a few high-level ideas about races, nations, and technology level, and the idea that souls play a major role in this world.

So where do we go from there?

Building a Pantheon
One of our first steps was to think about gods. Deities can be a good starting point when developing a world. They reflect the views and beliefs of the world’s inhabitants, and they can inspire ideas for characters, organizations, and conflicts.

You’ve already heard a few of our gods mentioned in passing: Magran, goddess of fire and war; Berath, god of cycles, doors, and death; Eothas, god of light and redemption. Josh invented these gods when he was first developing the world, and they play important roles in the region where the game will be set. But we’ll need a lot more gods to fill out the pantheon.

Here are a few of the elements we consider for each new deity:

- What is the god's name, and what are his/her "aliases" (e.g., "The Twinned God" for Berath).
- What is the god's portfolio? That is, what aspects of life or the world do they represent (e.g., mortality, greed, summer, commerce)?
- What allies and foes do they have amongst the other gods?
- What are their symbols?
- How do they manifest in the mortal world?

We list this information for each deity, as well as providing a detailed description. Players won’t necessarily get to see all this stuff, but it’s useful background for the art and design teams, so that the world feels like a consistent, coherent whole.

One other thing to bear in mind: for the most part, our deities aren’t good or evil. They’re somewhere in between – closer to the multi-dimensional gods of the ancient world. Every deity has his or her own agenda, which isn’t bound by notions of alignment. Sometimes they can be helpful and benevolent. Other times – not so much.

Woedica – "The Exiled Queen"
All the preceding info is important, but I wouldn’t want to leave you without revealing an actual example...

A lot of my design ideas start with a visual image. That applies to characters, locations, even narrative. I’m not sure where most of them come from, but examining them more closely will usually lead me to develop stories to explain who they are and where they came from.

One of the first images that sprang to mind was an old woman – a dethroned queen – wandering along an empty road in tattered finery. Despite whatever horrors she had suffered, she maintained a certain stubborn dignity, and she carried a heavy book of law.

I felt like this goddess could cover a range of portfolios, and I liked the idea that the deity who was the "rightful ruler" of the gods (in her mind, at least), had lost her throne. That may have some interesting implications for the way in which mortals view the world.

So what was her story? According to her followers, she had once claimed rulership over all the other gods. But if that was true, she was cast down in the far distant past. Among the other gods, she has no real allies, believing that all the gods owe her fealty. She claims the portfolios of law, rightful rulership, memory, and vengeance. And she manifests in the world as the Strangler, a leathery-skinned old woman, always clad in tattered finery, who appears on an empty road or abandoned alleyway to murder those who break a solemn oath.

Her Aedyran name is Woedica, which evokes the “Old English” feel of the Aedyran language. (Maintaining a distinct sense of national/ethnic language and culture is important to us – more about that in a later update.)

There’s a lot more to tell about the Exiled Queen and the other gods (some of which aren’t even “human”). But that should give you a taste of our creative process. We’ll have plenty more to say about world-building in future updates.

I really, really like the concept of that god. She sounds to me like she might also be a logical fit for certain kinds of mental disorders: wilful ignorance, dementia, cognitive dissonance, that sort of thing. Sort of like a lighter variant of Delirium from Sandman.

Of course, we don't know how the gods are supposed to work yet. It's common for gods in fantasy to have a darker side to them as well, such as a god of leadership also being a god of tyranny.

Or maybe I've just read too much Song of Ice and Fire.

Ive been lazy and haven't read too many of the updates, but at least they are pumping them out to the KS backers (unlike a lot of other devs). Really looking forward to the game when its done. But til then.. i better get reading hehehe.

Update #41: D&D: Dwarves and Doors
Update #41 · Feb. 06, 2013 · 17 comments


Adam Brennecke

We are another month into preproduction and have been making awesome progress on all fronts. This update covers dwarves and doors, two of the many accomplishments in the month of January, and gets into the finer details of development on Project Eternity.
The Creation of the Dwarf

One of the goals in preproduction was to figure out how we could make character modeling pipeline be as efficient as possible. The problem is fairly complex: All of the six playable races, human, elf, dwarf, aumaua, orlan, and the god-like can wear armor, boots, gloves, helmets (...well, some have trouble wearing helmets, but we will talk about that some other day...) and have other options that the player can customize like facial hair, hair style and skin color. We also have tons of armor variations and types of armor, like plate, brigandine, leather, and mail. (Josh loves his armor). Ideally, our artist would only need to model one armor piece - let's say plate body armor - and have it fit all six of our playable races even if the races are all of different proportions and body structure. At the end of the day the same model for plate armor could fit a slender four-foot-tall orlan and a burly seven-foot-tall aumaua. The goal for January was to build a system to allow us to do this very thing.

During January, we've developed a new system to allow our human bipedal skeleton to be shaped and morphed into the other playable races and have armor be shaped and morphed along with the skeleton. The character modelers have fine control over the proportions of the races, and only need to model armor pieces once and not six times over. In preproduction we look at developing systems like this. It may cost us time up front, but will save us hundreds of hours down the road in production. The dwarf ended up being our first test case, and now we have dwarves as playable races working in game.

Pictured at the front of this update is a high-poly dwarf head that Dimitri Berman (lead character artist) modeled in ZBrush. The high-poly head is used for making normal maps which aid in lighting the character models. A simplified mesh is created from the high-poly head is used in game.One of the goals in preproduction was to figure out how we could make character modeling pipeline be as efficient as possible. The problem is fairly complex: All of the six playable races, human, elf, dwarf, aumaua, orlan, and the god-like can wear armor, boots, gloves, helmets (...well, some have trouble wearing helmets, but we will talk about that some other day...) and have other options that the player can customize like facial hair, hair style and skin color. We also have tons of armor variations and types of armor, like plate, brigandine, leather, and mail. (Josh loves his armor). Ideally, our artist would only need to model one armor piece - let's say plate body armor - and have it fit all six of our playable races even if the races are all of different proportions and body structure. At the end of the day the same model for plate armor could fit a slender four-foot-tall orlan and a burly seven-foot-tall aumaua. The goal for January was to build a system to allow us to do this very thing.
Open, Close, Lock

On the other end of the pre-production spectrum, the programming team has been writing the building blocks for the area design toolbox. One of the essential things that all areas need are doors. From past experience we know that doors always present difficult problems with pathfinding and are a big pain in the arse. Getting a potentially risky, yet required, feature out of the way now seemed like a pragmatic goal, so Steve Weatherly (game programmer) and Sean Dunny (environment artist) set off on a quest to get doors working in the game.

We first tackled this problem creating a list of all the features that doors need to have. It's easy with doors since we all know how doors work: Doors have a few states, like open and close. Doors can be locked, and be unlocked with a key (or skill). Doors can be used, meaning the player can click on a door and the selected character will be commanded to go and "use" the door. Doors can animate to match the open/close state. Doors block character pathing when closed, and don't block pathing when opened.

We even listed out minor details such as doors can change the mouse cursor to a different state when hovered over, and doors should always open away from the character using the door.

Tasks were made from this list, and the work began. Steve was able to get a working prototype of a door ready to test quickly. At this stage we could see how the door looked and felt in game, and if there are any unexpected problems that came out of the prototype.One of the goals in preproduction was to figure out how we could make character modeling pipeline be as efficient as possible. The problem is fairly complex: All of the six playable races, human, elf, dwarf, aumaua, orlan, and the god-like can wear armor, boots, gloves, helmets (...well, some have trouble wearing helmets, but we will talk about that some other day...) and have other options that the player can customize like facial hair, hair style and skin color. We also have tons of armor variations and types of armor, like plate, brigandine, leather, and mail. (Josh loves his armor). Ideally, our artist would only need to model one armor piece - let's say plate body armor - and have it fit all six of our playable races even if the races are all of different proportions and body structure. At the end of the day the same model for plate armor could fit a slender four-foot-tall orlan and a burly seven-foot-tall aumaua. The goal for January was to build a system to allow us to do this very thing.

One issue that came up was door placement. We found that it was not easy to place a door in the exact space to fit a dungeon doorframe. Steve and Michael Edwards (senior technology programmer) coded a system for doorframe "snap points" that makes the door pop to the exact place that we want it to go. Designers can now place doors efficiently. Hooray!

We love being able to share our progress with you all, and we hope you enjoy reading these production updates. If you have any questions about development, please post them in our Project Eternity forum.

Thank you!

I think something or someone messed up there :P

The newsletter email also had that duplicated paragraph.

http://forums.obsidian.net/topic/63354-update-42-avellones-lets-play-arcanum/

Not up on KS page yet. He has started his playthrough of Arcanum sans mods.. :)

Why does this make anyone happy? I mean, I assume he's a bit ignorant as a video game writer if he hasn't hit the major "touchstone" games of every generation, but I don't get the giddiness. Arcanum is one of my favorite games of all time.

Eh, I respect him as a writer and Arcanum is one of my favorite games of all time, so it's just fun to see 'em in conjunction, y'know?

I guess not, but I'm glad you're having fun:)

Fair 'nuff. :)

(Whats this mention about Project Eternity/Trention???)


Update #43: Pretty and Technical
Update #43 · Feb. 20, 2013 · 21 comments


Rob Nesler, Project Eternity's Art Director

Hello everybody it's me again, Rob Nesler, Art Director on Project Eternity. I had intended on presenting our Art Style document to the world by this update, but it still needs work. So, you have to wait a bit longer for it. So that’s sad. However, with the last art update, I glossed over our Technical Animator Antonio's work 'cause it was 3 o'clock in the morning and I couldn't think about how to describe intelligently what Antonio does for us. Some of you professed extensive knowledge of rigging and skinning, as well as profound disappointment in my patronizing tone, and demanded to be better informed of this horribly complex facet of game development. So...okay!

Before that we have some eye candy.

A Godlike concept by Polina (click to see full size image):

And noooooo...you can’t have a larger version. Google "estoc" if you want a sense of the weapon she's holding.

The godlike are the children of humanoids (most often humans) who have been "blessed" (or cursed, depending on personal or social view) with the physical manifestation of a divine spark granted by the gods. Godlike manifest their divine heritage in a variety of ways: wings, horns, strange birthmarks, talons, odd eyes - but they always manifest it somehow. BTW, in case you missed it, this is a playable "race". Below are some further examples of the possible deviations/conditions they inherit:

Okay back to technical animating.

If you recall, in the last update, I wrote this little nugget:

"Antonio Is our Technical Animator. He makes rigs, writes scripts that make rigs, and rigs the rigs. It’s all very technical. You wouldn’t understand. He’s a professional."

Some of you wanted more... so... here’s more.

His primary responsibility is rigging and weighting characters for skinned forward kinematic and inverse kinematic animation. Daily, he also uses his expert technical knowledge of animation techniques and the underlying idiosyncrasies of 3D animation software to assist our animation team as a problem solver, a mentor, and a scripter to improve animator workflow and overcome deficiencies in the applications we use. So that sounded intelligent; vague and awkward, but intelligent.

So, specifically ROB, what does Antonio do that will make Project Trenton/Eternity awesome?
Generally, technical animators program scripts using languages like MEL or Python to extend or enhance the functionality of existing off-the-shelf apps like Maya, Max, or Softimage. These scripts are often programs running within the larger app, but sometimes they are utilities that exist outside the application to assist in conversion or batching operations. These often can be purchased, but sometimes they themselves don’t have all the features our animators want so...

NO! Specifically ROB! What the hell does he do???
Okay, okay, for Project Trenton/Eternity Antonio has written, a few, and re-written a couple times the following, and it's all called DNA (Design New Actor).

Firstly, there is the Export Rig, and this is not so unique to 3D character animation. This is a highly optimized skeleton that represents only the bones that the actual actor mesh is weighted to. For a humanoid character these bones would be named: Pelvis, Femur_Rt and Femur_Lt., etc. Vertices of the visible textured geometry, that the player sees, are all attached to these bones with various amounts of strength, so the character will appear to bend and flex more naturally. Sometimes these bones will have physics applied to them, like pony tails. This is less a rig and more of: simply the skeleton that the other rigs interact with, but these are the only bones that go into the game, and we call it the "export rig". Antonio builds this skeletal hierarchy to fit the expected proportions of the character, and character modelers build character meshes to match visible body parts to the locations of bones, so that -- for example -- the bendy part between the upper_arm_rt and lower_arm_rt ends up being where a humanoid creature, as designed, would expect an elbow to be.

Then there is this nifty rig that was discussed a little bit, a couple updates ago — fricken’ Adam (PE's Executive Producer) always stealing my thunder! I think we call it the “stretch rig”. As he mentioned, it allows Dimitri, our dutiful Character Lead, to scale, even non-proportionally (with volume adjustments) meshes with the export rig already weighted. This way we can make a human character mesh, rig him to a human-proportioned skeleton, animate him, then take that same character and deform him to a new size and proportion, export him and use the same animations that were created for the human on this newly scaled/deformed mesh. This is a very useful asset multiplier, which allows us to leverage costly armor set and animation development across the spectrum of our races.

(Click for a larger image)

...And last, but most certainly not least, we rely on Antonio for creating rigs that provide controllers that allow the animators the maximum amount of flexibility when creating animations. These 'control rigs' are specialized for each character/creature type, after they've been stretched. They provide controls for jumping, crouching, twisting, grabbing, etc. They provide inverse kinematic and forward kinematic transformations, physics blending, following, squeezing, etc.-- whatever is required, actually. These controllers and extra bones are for animator manipulation only; they do not get exported. That's why when I said, "he rigs the rigs" I wasn't actually joking. These are rigs of rigs. These are the animator’s most essential bread and butter. Using these tools is how they actually create character animation.

And that's it. The update is over. I know, not so fun, but this is serious stuff.

We’ll show some animated examples sometime in the not-too-distant future.

Rob Out!

http://forums.obsidian.net/topic/63432-update-44-the-rules-of-melee-engagement/
(Kickstarter hasnt been updated yet

Update #44

Update by Josh Sawyer, Project Director and Lead Designer



Last week, our art director, Rob, showed you our godlike concepts and dazzled you with an in-depth technical breakdown of how we're doing animation rigging on the project. This week, we'll be talking about a different technical subject, but one that's more connected to gameplay: engagement -- specifically, melee engagement.

Melee engagement is a solution to two common problems in the Infinity Engine games: melee characters' inability to control an area and ranged characters' ability to "kite" melee characters. In the Infinity Engine games, melee characters could be quite powerful in toe-to-toe combat, but many opponents found ways to foil those characters with little difficulty. Fast characters could easily rush around a slower melee character with impunity and ranged characters could backpedal perpetually out of reach.

If you're familiar with D&D 3E/3.5/4E/Pathfinder's attack of opportunity mechanics, Project Eternity's melee engagement fills a similar role by making melee combatants "sticky". Coming near a melee combatant means being drawn into Engagement with him or her, a state that can be risky to get out of.

Here's how it works: when two opposed combatants come near each other and one of them a) has a melee weapon equipped b) is not moving and c) is not currently at his or her maximum limit of engagement targets (the standard is 1), the other character will be Engaged.

When an opponent is Engaged by an attacker, moving any significant distance away from the attacker will provoke a Disengagement Attack. A Disengagement Attack has an inherent Accuracy bonus, does significantly more damage than a standard attack, and will call a hit reaction animation while momentarily stopping the character's movement.

When it's initiated, a Disengagement Attack automatically breaks Engagement on the target, but if the target is also the attacker's current melee target, the attacker will typically be able to re-establish Engagement before the target can move farther away. In this manner, melee combatants, especially ones that have high Accuracy and damage per hit, have a solid mechanic for keeping enemies close to them -- or making the cost of escape extremely expensive.

Of course, there are other ways to end Engagement. If the attacker switches to a non-melee weapon or performs a non-melee-based action, Engagement immediately ends. If the attacker moves away from their Engagement targets, is paralyzed, knocked down, or otherwise prevented from maintaining a threat, Engagement will also immediately end. If the attacker has a limited number of Engagement targets (as most do) and switches his or her attack focus to a different character, Engagement immediately ends.

We believe that Engagement can give AI a clear "decision point" where they can evaluate the threat of their new status and choose the appropriate course of action. For player-controlled characters, it makes melee enemies more potent threats and presents players with tactical challenges to solve.

We want Engagement to be a mechanic that players and enemies can mess with using a variety of class Abilities and general Talents, so we will be experimenting with a variety of elements to that end:
Fighters' Defender mode allows them to engage two additional targets and increases the range at which they engage targets. This gives fighters much greater capability to control the area around them.
The limited-use Escape ability lets rogues break Engagement without provoking a Disengagement Attack. It is generally best used when the rogue's enemy is preoccupied with another target.
Barbarians can use Wild Rush to temporarily ignore the movement stop and hit reactions from Engagement and Disengagement Attacks, respectively -- though they can still suffer massive damage while powering through.
The wizards' Grimoire Slam allows them to attack an enemy in melee with their magically-charged grimoires, unleashing a concussive wave of energy on contact. If it hits, the attack knocks the target back, usually far enough to break Engagement in the process.

Additionally, creatures may have their own special abilities related to Engagement and Disengagement Attacks. We hope that the system itself is easy to understand but allows for increasingly complex tactical considerations over the course of the game.



That's all for this week! Let us know what you think of the mechanic on our forums. Your feedback, as always, is appreciated. In our next update, we'll be continuing our thrilling coverage of Chris Avellone's playthrough of Arcanum: Of Steamworks and Magick Obscura.