Oh wow :drool:

Can't trust Youtube... The compression may be causing that.
(though there are moments of obvious 'low rez' looking textures)

Yes it is lol. Crytek has designed the engine around supporting massive landscapes! It has complex LOD systems, streaming memory features and more just to make huge open environments like whats in TES. Also you can have that quality and that large ended open world on both consoles and PC. CE3 scales according to the platform. With Consoles and up to high end DX11 PC's. Its a phenominal engine and was built for all genres of games. Just because the first game to use it is going to be an FPS doesnt mean it cant do everything we need from a TES game. its by far capable and yes in the case of large worlds designed for it from the start.

It's still being developed. Check out this video showing the water.
http://www.youtube.com/watch?v=X9B6rK0WwHI

I would rather Bethesda use off the shelf libraries & engines for for everything that it has in common with many FPSes and other games (graphics, scripting languages, most map loading, possibly map editing, most physics, most timing, input logic, and a lot of basic gameplay / combat elements). Bethesda should concentrate on the things that are unique to TES - crime and magic systems, distant NPC interactions, the plot, etc.

That, and I think that many of the people who have reverse engineered much of Oblivion / Fallout are a bit skeptical of Bethesdas ability and inclination to write an entire game engine.

There are far more things to a game engine than just graphics. But it must obviously be able to support huge open game worlds that aren't separated by levels and loading screens like Crysis.

But there's the AI (Oblivion keep track of it's ~3000 NPCs at all times), quest system, dialogue system, streaming capabilities, game building capabilities, debug features and so on. I'm not sure if many here can tell if whatever engine are well suited for large RPGs from that point of view, I sure don't. Instead the whole discussion here seem to be about making Skyrim pretty. But you can't base how well suited a game engine is for a certain game on graphics alone, which largely depends on the game artists anyway.

But if Bethesda used CryEngine, Unreal 3 Engine, Source or any popular engine, and built the game using the default tools that are supplied, the game world would be massive and would need several DVDs to be stored. In the past games Bethesda built their games with a plug-in system that proven to be very efficient when it comes to storage costs. What takes space in Oblivion are the voices and the textures, not the game world data (FalloutNV.esm is only 240MB). So they would have to rework several systems regardless.

I completely disagree. The whole reason they stepped away from the by hand creation was to try and add more content. Their incompetents in successfully doing this is obvious and very sad of course, but it doesn't add any less value to the efforts they made. Morrowind was all hand places, and as you could tell they had a lot more stuff crammed into much smaller places in Morrowind, but they spend just as much time mapping out the entire world as they did placing the objects inside the world. The procedural generation and creation of the world of Oblivion took very little time relatively speaking, as a result of programs like speedtree and other procedural generation technologies. This gave the developers more time to place content into the world, and work on their engines more concerned with actual gameplay. It's thus extraordinarily unfortunate that they focused so much of graphics and polish and less time on actually making MORE content, and thinking of better ways to implement it.

In any case, beyond shadow of reasonable doubt what developers NEED to start doing is considering better and more innovative ways to inject procedural generation into their games to add more variety, and more content to their games, but not just in the physical world, but in quests, and character interactions. Now in the old DOS game way though, but in character influence methods, quest lines, and genuine in game events. My 2 cents.

That is....magnificent.

Off-the-shelf middleware is the present, and the future of game design. Games are far too complex these days to rewrite everything from the ground up each time anymore, and when you are trying to do something already proven and established, like LOD generation, and physics, it just makes sense to use what is already available, and tool it to suit your own needs.

You might think that constantly using middleware will restrict games to only what this middleware allows,and restrict further advancement, like middleware is a kind of anchor. In reality, the middleware is constantly being developed as well, and will continue to push the envelope of what the hardware is capable of. Games will follow suit.

It's not like you can ask a level designer to start programming, or a programmer to write stories or design the game.
Those are all entirely different jobs.

I'm even pretty sure that the programming department has separate teams of experts working on various aspects (engine, effects, AI) of the game.
Game developers used to have small teams of multi talented people 20 years ago, but that's not how it works these days.

I think it's admirable for a game studio to develop a new engine from scratch.
A game has a lot more potential this way and I don't think it's fair to start criticizing that decision without even having seen one screenshot.

But they're always designed with 'traditional' games in mind and they're not likely to develop innovations which they're not sure anyone would like to use them.
For example: Suppose you want to ditch the usual motion capturing and replace it with an algorithm based system that allows millions of animations, or you want to be able for the NPC's to manipulate objects to build houses etc...

You're going to develop have these things yourself anyway and then you have to find out how to implement them in the engine without losing performance.
Alternatively you could hire some guys who used to program third party 3D-engines and let them start working on a new engine with these features in mind.

Also I think that if you can afford to have talented people in this field working exclusively for you, it's strategically better to do so than to have them run their own company where the entire industry can benefit from. This might also have something to do with it.

it was also admirable for the last clan of samurai to launch that final attack against the imperial Japanese army in The Last Samurai, but ultimately they were blown to pieces.
and its not like the employees would be doing a totally different job, more like adding and customising an already existing work.



yes, but all the npcs did was walk in a repeating circut.
and its not only making skyrim pretty, because they are probably already doing that. it is about choosing somthing that will make it pretty with a minimal amount of effort on their part so they can focus on other aspects, thus making the intended topic here to be how can bethesda maximise the content of the game while minimising time spent making that content prettier,

I'm sure Bethesda took into consideration, things like the Cryengine etc before they decided to use their own.

In-house engine is the way to go, even if it's not awesome with Skyrim, it gives them something they can work on and improve in their own way and tailor it to their needs without any restrictions.

Like a poster said before, Blizzard and Valve use their own engines and they are big players who make good games; it obviously makes sense to have your own engine or those guys wouldn't have bothered either.

Bethesda is a big player now and it's about time they had their own engine to tailor to their style of game development.

Then there's the money aspect; who wants to shell out millions to a 3rd party who will only place restrictions on what you can do when you have the capability to create your own and answer to no one on design issues?

I think any software company that incorporates it's own engine will usually succeed with the subjective requirements of their games.

It's all very well using an off the shelf engine, but one that can be adapted from creation to fit the needs exactly will always be better IMO.

You really think it makes their job easier to use an unfamiliar engine that they can't change at whim without having the authority of the original makers?

Why do you think Gamebryo was so buggy? Because Gamebryo was messing up and Bethesda couldn't just say "Hey you know what, this is buggy so well fix it" because it is not their engine and they can't just change the fundamentals of it. It left them stuck with a duff product.

It makes things easier if you have your own engine.

Indeed. This is what excites me. They were limited by the Gamebryo engine because it wasn't theirs. The sky is now the limit.

Because if you try to get any higher some dragons come and take you down...

If you look at the dozens of other games that use Gamebryo, and also find they have the same or similar issue with bugs, then your theory might have some weight. I think that would be worth some research.

I do find it unlikely that a middleware that would be notoriously buggy as you think and would be still have been as widely used as it is, will have lasted as long. :shrug:

I completely disagree.

In-house engine tailor made for the game is the way to go. Makes the game have it's own look rather than the after-taste of other games (cough UE3-engine cough)

Simply put, if the development house has the money and the resources to support it, an in-house engine is the way to go. Why? Because they're not limited by said engine. They can build it to support the kinds of features they want in their games in the way they would do them best, instead of finding convenient workarounds.

Just look at how Tribes and Tribes 2, two of the best online multiplayer FPS games ever made, are still avidly played by their fans; both made on the Torque engine (Well, Tribes was made on Torque's predecessor), designed from the beginning to support that kind of gameplay, with wide, open spaces. Tribes Vengeance, a not altogether bad game, had to make a lot of compromises with how the game plays in scale and physics, because it was made with the Unreal engine, and as a result lost a lot of what made the first two games so special. Nobody plays Tribes Vengeance anymore.


It's nothing but a good thing for TES.

I highly doubt that you can simply open the sourcecode of an "of the shelf engine" and start toying with it, therefor it will always have it's limitations.
Development platforms don't always do exactly what they're supposed to do according to the manual either, it's not like you'll ever be buying the perfect product to build your game on.

Also bare in mind that the team working on Skyrim is much larger than the one working on Oblivion.
If they want to program the engine themselves, then they should definitely do it. The base engine is probably finished by now anyway.

It depends on the company. If I heard LucasArts were making a new engine, I would tremble in terror. But I have confidence in Bethesda.

yeah..has bethesda ever made an engine from scratch?

i think lucas arts would actually make a pretty decent engine..........i think they designed the dark forces engine which as the time blew everything else out of the water. its the games themselves that lucasarts makes that usually svck lately. republic commando and empire at war are the last good games i can think of and both of those are practically from the 1980s.

Xngine for Daggerfall and Battlespire.

And it doesn't matter if they ever made one. What matters is the quality and creativity of the coders and designers they have on it now.