Developer Story Time with Old Man Davis

[Anthony Davis]

Ok, it seems to be popular, so let's talk about sequels. I might get some details wrong because my memory isn't what it used to be.

Obsidian has done three four sequels:

KOTOR2: The project was 12 months long.
The game uses the CONSOLE version of the Aurora engine. This engine was not heavily modified, though Jay Fong, Frank K, and Rich Taylor did do a lot of work on it, like adding Bink support, improving the map and fast travel system, oodles of pathfinding bug fixes, dual wielding weapons, new UI, modified inventory system... and that's all I can remember off the top of my head. So a fair amount of modification, but nothing too crazy.
Frank K was also responsible for a lot performance improvements, for example he fixed a HUGE bug on grass rendering and alpha testing/blending that really improved the frame rate. In general tests, Frank and Jones (I think) improved the performance of the game across the board by about 30%. However, we also added about 30% (or in some cases more) objects and other content to the game.

We reused SOME assets, like the Ebon Hawk and some character models, textures, creatures, etc. A LOT of new art assets as well as sound effects were created from scratch.
New animations were added, and the animation system was worked on.

There is actually a pretty good post-mortem of KOTOR2 in some old issue of Game Developer that was written by Kevin Saunders. I don't know if anyone has it, or if there is a digital version somewhere.




NWN2: This project was about 22 months long... maybe 23, I can't be sure.
The game uses a DRASTICALLY modified version of the PC version of the Aurora engine.
The entire renderer was re-written from scratch and called Electron.
The entire UI was re-written.
The entire script system, sound and music system, character system, RPG mechanics, and more were all drastically overhauled.
EDIT: I remembered some more. We completely changed the way worlds and levels were built.
We (mostly Erik Novales with some good help from Brock Heinz) created a brand new toolset with supports actual in engine rendering and scores of new features.

The changes we made were very far reaching. For example, this also includes the party mechanics and the party companion system. In NWN1, your single companion you could pick was not an actual character. He didn't really "level up". He didn't even have his own inventory. It was all faked. Rich Taylor and I fought VERY hard to get the time to redo all of that completely, especially considering the complicated nature of the companions written for NWN2. Brock Heinz and I worked on that, and I remember vividly when we got it working and using a script command we added a bugbear to the party and then proceeded to level him up and give him equipment. This is also why the druid pets now actually level up too. It was super awesome and I really wish we had been able to add more monsters to the party in the original campaign.

No art assets were reused, obviously.
Sound and Music from NWN1 were kept, but also a lot more music and sounds was created for the game.

I am not aware of a post-mortem that was done for NWN2 (outside of our internal post-mortem).

The next two expansions for NWN2 would continue to see drastic improvements and new features added to the game.



Dungeon Siege III:
I cannot remember how long this project was, even though I was on it the whole time it was in development... I *think* it was 18 months, but it could have been a lot less.
The game uses Obsidian's proprietary engine, Onyx.

The game uses zero assets at all from previous Dungeon Siege games, obviously. Except for the name, it is an entirely new game.

We had no real development problems on this game. Sure there were bugs and stuff that we had to work on, sure there were disagreements and discussions, but for the most part this game was very smooth, which is what Feargus and the other owners wanted. It was stressful despite how smooth it went precisely because we WANTED it to be smooth and as bug free as possible.

I am not aware of a post-mortem on this game (outside of our internal post-mortem).


DERP, I forgot New Vegas.
Fallout New Vegas:
I actually am not as familiar with this project as I am the other three because I never worked on it in an official capacity, I was working on DS3 at the time. I think this project was about 14 or 16 months long.
The engine was the same one from Fallout 3.
While not originally scoped, the small programming team for FO:NV actually had to do a LOT of work. Almost every single SDK had to be upgraded in order to pass certification.
The companion wheel was added.
A new dialog editor was created/added.
Obsidian superstar Mike Bosley created Obelisk, a powerful memory/performance tool. This is what was used to fix many PS3 memory specific problems due to the PS3's terrible non-unified memory. Bethesda used Obelisk to get Skyrim working on the PS3 also.
Thirst, Hunger added to the game.
Faction system drastically revamped.
Weapons, combat, and VATS were drastically overhauled. One of the goals of the game was to be able to play through the game without using VATS. Josh succeeded on this goal hands down.

There were a lot more features added... but they are slipping my mind at the moment.

Almost no original art was used as it was, much of it was new, or heavily modified because the color palettes and land geometry is so different compared to the east coast.
A lot of old sounds were kept, but a ton of new sound effects and music were created/added to the game.

Obviously a tremendous amount of content was added as well.

Lots of performance and bug fixes, many of the bugs were YEARS old.

It was a great project with lots of awesome DLC. I really hope Obsidian gets a chance to make another one because everyone loved working on it and I think the fans really liked it too.

 

[AN4RCHID]

Ok, it seems to be popular, so let's talk about sequels. I might get some details wrong because my memory isn't what it used to be.

Obsidian has done 3 sequels:

New Vegas too.

 

[Anthony Davis]

Ok, it seems to be popular, so let's talk about sequels. I might get some details wrong because my memory isn't what it used to be.

Obsidian has done 3 sequels:

New Vegas too.

Good reminder, I'll edit the original post. Thanks.

 

[Cosmo]

3. I hope with all my heart that Obsidian someday gets the chance to apply what they learned from the mistakes and gets to make Alpha Protocol 2: Protocol Harder. I know this is a slim hope at best, but stranger things have happened.

Don't worry. After the success of South Park i predict they'll be hired to make an Archer RPG...

 

[AN4RCHID]

Ok, it seems to be popular, so let's talk about sequels. I might get some details wrong because my memory isn't what it used to be.

Obsidian has done 3 sequels:

New Vegas too.

Good reminder, I'll edit the original post. Thanks.

Cool, thanks. Fascinating stuff btw.

 

[Lilura]

In NWN1, your single companion you could pick was not an actual character. He didn't really "level up". He didn't even have his own inventory. It was all faked. Rich Taylor and I fought VERY hard to get the time to redo all of that completely, especially considering the complicated nature of the companions written for NWN2.

This is true of the OC but in SoU and Hordes you could access henchmen inventories and had limited control over how they leveled. You could, for example, ask a companion to stop leveling in one class and start in another.

But yeah, it was a welcome addition in NWN2 to have direct control of companions in combat, with the ability to select their level-by-level feats and skills (both optional, of course).

I thought SoZ was conceptually a high-point of NWN1/2 series, total control in combat and leveling, the overland map, trading and easy crafting (!), the party dialogue system, the oldskool feel. SO MUCH MORE fleshing out needed in dungeons and combat, but a solid framework is offered for modders to work with at least.

Incidentally, why do you think the NWN2 community was never as successful as NWN, I would say one reason is no Linux support for servers and the like?

 

[pakoito]

My thread ... What have you done?

I suppose it was just a matter of time.

Noooooooooooooooooooo I want the class header and data driven discussions before the Codex eats up the thread

 

[Anthony Davis]

In NWN1, your single companion you could pick was not an actual character. He didn't really "level up". He didn't even have his own inventory. It was all faked. Rich Taylor and I fought VERY hard to get the time to redo all of that completely, especially considering the complicated nature of the companions written for NWN2.

This is true of the OC but in SoU and Hordes you could access henchmen inventories and had limited control over how they leveled. You could, for example, ask a companion to stop leveling in one class and start in another.

But yeah, it was a welcome addition in NWN2 to have direct control of companions in combat, with the ability to select their level-by-level feats and skills (both optional, of course).

I thought SoZ was conceptually a high-point of NWN1/2 series, total control in combat and leveling, the overland map, trading and easy crafting (!), the party dialogue system, the oldskool feel. SO MUCH MORE fleshing out needed in dungeons and combat, but a solid framework is offered for modders to work with at least.

Incidentally, why do you think the NWN2 community was never as successful as NWN, I would say one reason is no Linux support for servers and the like?

In SoU and Hordes it was still kind of fakey fakey. Their inventory for example, was actually part of the players inventory. Don't get me wrong, I'm really impressed how much they were able to stretch that old system.


There are several reasons why NWN2 was not as popular as NWN1.

1. It was drastically different from NWN1. The tools were more complicated, and the game was more complicated. People who are into NWN1 already had spent YEARS developing their skills for NWN1, and it clearly wasn't about graphics for them. Very similar to how Everquest 1 was much more popular than Everquest 2 - people who sank all that time into EQ1 did not want to flush it away to jump into a brand new grind.

2. A lot of the DM features and server features didn't become fully functional and till the first expansion. I think many of them got improved, but it wasn't enough.

3. We had too many problems with the 3rd party tools which inhibited game content creation (ie. Granny).

4. The gaming world had also changed. We now had much higher quality MMO's like WoW and EVE to contend with. The multiplayer gaming scene was changing very rapidly.

Those are just a few half-thought out guesses.

 

[J_C]

Incidentally, why do you think the NWN2 community was never as successful as NWN, I would say one reason is no Linux support for servers and the like?

I think the NWN toolset was so much simpler and easy to use compared to NWN2's toolset.

 

[Anthony Davis]

3. I hope with all my heart that Obsidian someday gets the chance to apply what they learned from the mistakes and gets to make Alpha Protocol 2: Protocol Harder. I know this is a slim hope at best, but stranger things have happened.

Don't worry. After the success of South Park i predict they'll be hired to make an Archer RPG...

how cool would that be? Avellone loves Archer too and he totally gets it. it would be a thing of beauty.

 

[Anthony Davis]

Incidentally, why do you think the NWN2 community was never as successful as NWN, I would say one reason is no Linux support for servers and the like?

I think the NWN toolset was so much simpler and easy to use compared to NWN2's toolset.

Definitely true, but our goal was to enable more "organic" levels. Oh well.

 

[Roguey]

How long was Kotor 2's development time and how long was Kotor 1's?

About a year; about 3.5 years.

So why does Obsidian get these deals where they have to develop games in a short time compared to the game before it? Bethesda spent 5 years on Skyrim.

Publishers don't like spending a ton of money when it comes to contracted outside jobs, especially when the contractors don't have a big hit to their name.

You'd think Bethesda would have done Q&A for the game so it wasn't as buggy so the game would be more successful.

They did do QA, but they also moved the release date up according to Brian Fargo. It'd be against their best interests if Obsidian were to meet their bonus goal.

The fact of the matter is Lucas Arts destroyed the franchise. There was no interest in a Kotor 3 after Kotor 2 failed to do well, so it was turned into a MMO. Or word of mouth was so bad with Kotor 2's reception that they thought a Kotor 3 wouldn't do well.

They had an internal studio working on kotor 3 pre-production the same time they had Obsidian working on the second. It was canceled because of financial difficulties.

 

[Davaris]

Noooooooooooooooooooo I want the class header and data driven discussions before the Codex eats up the thread

http://gameprogrammingpatterns.com/component.html

 

[Infinitron]

In SoU and Hordes it was still kind of fakey fakey. Their inventory for example, was actually part of the players inventory. Don't get me wrong, I'm really impressed how much they were able to stretch that old system.


There are several reasons why NWN2 was not as popular as NWN1.

1. It was drastically different from NWN1. The tools were more complicated, and the game was more complicated. People who are into NWN1 already had spent YEARS developing their skills for NWN1, and it clearly wasn't about graphics for them. Very similar to how Everquest 1 was much more popular than Everquest 2 - people who sank all that time into EQ1 did not want to flush it away to jump into a brand new grind.

2. A lot of the DM features and server features didn't become fully functional and till the first expansion. I think many of them got improved, but it wasn't enough.

3. We had too many problems with the 3rd party tools which inhibited game content creation (ie. Granny).

4. The gaming world had also changed. We now had much higher quality MMO's like WoW and EVE to contend with. The multiplayer gaming scene was changing very rapidly.

Those are just a few half-thought out guesses.

You're forgetting:

5) The engine and interface were and still are to an extent kind of slow/clunky. The game didn't feel responsive. It had a broken, laggy feel to it. Character movement looked weird. And of course there's the endless Codex complaints about the camera.

 

[Anthony Davis]

In SoU and Hordes it was still kind of fakey fakey. Their inventory for example, was actually part of the players inventory. Don't get me wrong, I'm really impressed how much they were able to stretch that old system.


There are several reasons why NWN2 was not as popular as NWN1.

1. It was drastically different from NWN1. The tools were more complicated, and the game was more complicated. People who are into NWN1 already had spent YEARS developing their skills for NWN1, and it clearly wasn't about graphics for them. Very similar to how Everquest 1 was much more popular than Everquest 2 - people who sank all that time into EQ1 did not want to flush it away to jump into a brand new grind.

2. A lot of the DM features and server features didn't become fully functional and till the first expansion. I think many of them got improved, but it wasn't enough.

3. We had too many problems with the 3rd party tools which inhibited game content creation (ie. Granny).

4. The gaming world had also changed. We now had much higher quality MMO's like WoW and EVE to contend with. The multiplayer gaming scene was changing very rapidly.

Those are just a few half-thought out guesses.

You're forgetting:

5) The engine and interface were and still are to an extent kind of slow/clunky. The game didn't feel responsive. It had a broken, laggy feel to it. Character movement looked weird. And of course there's the endless Codex complaints about the camera.

I partially agree, but I think the context menu was hands down better than that radial menu. I also understand that that is subjective.

Also, one of those camera settings was the EXACT camera setting for NWN1. Unfortunately it wasn't the default one.

 

[Infinitron]

I partially agree, but I think the context menu was hands down better than that radial menu. I also understand that that is subjective.

Also, one of those camera settings was the EXACT camera setting for NWN1. Unfortunately it wasn't the default one.

The setting may have been the same as NWN1, but the different type of maps that NWN2 had (more organic) meant it was no longer as appropriate. The camera was also laggier, as I said. You'd move your mouse to the edge of the screen and it would take a half second before it started moving. It was really annoying, particularly on the PCs of the time.

Overall NWN2 seemed like it was made by people who had never played a 3D game with a top down perspective and didn't quite know how it was supposed to feel and control. Either that or they didn't have time to tune/optimize it.

 

[Lilura]

UI is more powerful than NWN, though IMO not as intuitive, responsive or as pleasurable to use. Plus all the inventory icons were the same-sized little square, NWN had the most well-drawn icons of any cRPG out there.
Framerate is still pretty bad even on today's high-end rigs, when FRAPS reports high and consistent FPS there are still inexplicable jitters and jumps?
Load-times can be wrist-slicingly long, I even unzipped the compressed files in /data and re-zipped them as uncompressed and it made no perceivable difference. My rig isn't your typical Codexist lump of shit either, it handles heavily modded STALKER and Crysis.
Conclusion: poor optimization.

One thing that stuck out for me about character movement (coming off NWN) was how enemies would occasionally lock on and sort of glide towards you without a walk anim, and how great cleave lacked anims for each hit. Melee characters also didn't shape up and move around naturally and with realism as in NWN, in fact characters in general didn't distance themselves properly in fights, especially in mobs, sometimes swinging thru each others' models.

Most camera complaints are probably unfounded and based on pre-patch. There are three modes with configurable sensitivity and camera view and focus options, I can't think of a cRPG with more flexibility in this area, and since the patches the only problems I had were with exiting a building and the camera being behind you, so all you saw was a black screen. Bloody annoying, but I can think of 20 little things about the engine that annoy me more than the camera.

Bitching and nitpicking aside, Obsidian must have done enough things right, since I recently re-played all the official stuff and enjoyed most of it (mostly SoZ).

 

[Volourn]

"I partially agree, but I think the context menu was hands down better than that radial menu. I also understand that that is subjective."

You are insane. the radial menu is vastly superior. The only things NWN2 did better was graphics and 'more stuff' like classes, skills, spells, etc.

So, yeah, the toolset allows you to make prettier backgrounds but big deal. As a NWN 'fanboy' I would have LOVED if NWN2 was better. That's why i bought the game with the number 2 atached. I wanted a batter, updated NWN. I didn't get it so EPIC SAD FACE.

P.S. The NWN1/2 OCs are different case. theya re both slightly abover average RPG efforts. I might give the edge to NWN2 in that regard. But, the enmtire package/ NWN2 is EPIC FAIL.

 

[Lancehead]

Nice little derail in the thread, Mrowak "the devil's advocate" vs. The Tron.


Anyway, Anthony, can you tell us about any really cool ideas that got cut from the final game for whatever reasons?

 

[Jasede]

As someone who actually played years of NWN and NWN 2 online I can tell you first-hand that the NWN 1 interface, as Volourn correctly states, is MUCH better for sustained play.

 

[Jasede]

Of course they're subjective; that doesn't mean we shouldn't voice them, especially here on the Codex.

I had no issues with the menu. I always managed to select spells quickly and efficiently, even in real time. What helps a lot is using the hotbar. If I played a spellcaster I'd have a bar set to offensive magic, one for defensive magic and use the radial menu for everything else. A lot of spells are cast during "safe" situations (buffs, which is like half of your spells if you play efficiently), so you can easily radial-menu them. In combat you can use the hotbar- though honestly, since you have a whole 6 seconds to select your spell, that's plenty of time for me.

3 seconds if you're hasted might be a bit hard though.


And yes, we needed diverse spellbooks. The dungeon masters love nothing more than seeing that you only memorize fireball and haste and then making sure you have to fight some dispelling fire elementals, because fuck you players for thinking a mage would only bring two spells.

 

[Kane]

EDIT: I remembered some more. We completely changed the way worlds and levels were built.

And yet, the levels are still all rectangular.

Risk management and overrules other considerations. It's just the reality of business.

As a consumer, I can't accept that excuse.

You can not accept that all the way you want, but that won't change reality. The publisher will prefer going home and still having a job over going home and having no job.

The game uses Obsidian's proprietary engine, Onyx.

Will we see future games actually running on that thing? (Possibly even with non-rectangular levels???)

Weapons, combat, and VATS were drastically overhauled. One of the goals of the game was to be able to play through the game without using VATS. Josh succeeded on this goal hands down.

FWIW now, you should've gone all the way thru and turn that into a full FPS mode, like adding sprint, a grenade hotkey and stuff. There are mods now that add that to the game, but it's kinda beyond me why you didn't took it that far.

I think most of them who do are programmers.

The soldiers entrenched at the front line can handle the heat.

 

[Volourn]

"Radial menus in NWN1 made playing a spellcasting class a complete chore, especially if you aimed for diverse, utility spellbook and not for typical munchkin memorizing."

BULLSHIT. The radial menues PERFECT for ALL spellcasters.

 

[Gurkog]

It is always the head producer's fault if a project is unfinished. He shouldn't be making promises with no intent to back them, and should easily be able to see the effects of making such false promises while doing routine checks on progress. The producer let them shift focus and it is his fault they did not make the deadline. That is the guy's damned job after all.

 

[Kane]

and should easily be able to see the effects of making such false promises while doing routine checks on progress.

Yeah, thing is, that stuff isn't easy to see. Unless you're a wizard with a crystal ball.