Game Development at Microsoft (Bungie)

[SanGreal]

http://research.microsoft.com/asia/ur/workshop05/downloads/Game Development At Microsoft.pdf

Nothing terribly exciting, but it gives some interesting insight into Bungie's development process and some of the problems they encountered with Halo 2.

 

[Wunderchu]

cool,

thanx for the heads up

 

[pc999]

I did not liked this part:

"Breaking new ground is not a requirement".
Although being comparable to academic research, it is always a good thing.

PS/OT: how do you put quotes in this new forum.

 

[BRiT]

I did not liked this part:

"Breaking new ground is not a requirement".
Although being comparable to academic research, it is always a good thing.

PS/OT: how do you put quotes in this new forum.

The same way you did in the old forum. It uses the same exact syntax "

" pairs. You might need to change your editor preference to Advanced or WYSIWYG for all the other features.

 

[aaaaa00]

I did not liked this part:

"Breaking new ground is not a requirement".

Simple reason.

These:

- Concrete requirements, rigid timeline
- Results go straight to production

...tend to be incompatible with "breaking new ground" -- if some new-fangled research code is on your critical path, that is a recipe for schedule slip and project disaster, hence:

- Each R&D task must have at least one clean fall back

Then the question becomes, "what features do I cut in order to allow some guys to work on this research project (which may not ever ship and end up being wasted time), what are the potential payoffs and are they worth the risk?"

 

[RancidLunchmeat]

Actually, I found the approx 27 month timeline for game creation to be rather interesting.

Especially considering the OT discussion the "PS3 Expensive thread" about whether or not Bungie could actually release Halo 3 timed to counter the PS3 launch.

If it really takes Bungie 27 months (as low as 21 months), from start to finish on a game, I don't see why they can't have Halo 3 out the door in time for the PS3 launch. Especially considering it's a sequel, so you'd ass/u/me that their developmental time would be on the low rather than high end of their estimations.

That white paper also confirms what Xbdestroya said about Bungie having to scrap their engine halfway through the development process of Halo2 because the Xbox couldn't handle it.

My question would be: Does that actually give them a leg up this go around? Seems to me they developed an engine that was too complex for the last generation hardware to handle efficiently.

Did they develop the engine and just wait untl MS dropped the platform (the X360) into their lap?

 

[pc999]

Thanks Brit.

aaaaa00-
But they , sometimes, need to break new ground, likee (so they say, at least, I had not tried yet) MS had did with the drivatars in Forza, If I am not wrong Peter Molinoux (? Fable/B&W creator), also sayed, that they will started facingacademics problems (in AI I suposse).

e.g.If we really want to see an advance in AI, they need to, and great AI do wonders to the game, so I think it shoud be a riquirement, even if hard, and if it is not readdy to the game it would be for others games/estudios that could work togeder.

New ground is one big reason why we want next gen, to play the same we already have this gen.

 

[Carl B]

I'm thinking along the lines you're thinking Rancid: That if they chose to, Halo 3 development could likely be fairly accelerated if Bungie decided to use their existing engine assets from the Halo 2 development. Still, I wonder if they might not choose to develop a new engine for Halo 3 and use the previous one as their 'fall back,' as per the presentation - something that would still give them a graphical leap over Halo 2 then regardless.

The whole white paper itself is interesting, and especially timely considering my recent skirmish with BenQ. I'm not sure how many other dev houses would agree with the Bungie philosophy written within it though - it seems that other teams would approach projects with a different set of framework ideals; maybe some devs here will chime in on the subject. Whatever the case, the Bungie approach certainly seems sensible.

It does also lend some support to the theory that Halo 3 may be releasable come PS3's launch. Hopefully we'll gain some insights into the engine they're using come the release of a trailer, one would imagine in the very near future.

 

[richardpfeil]

Interesting read. I would have loved to see the presentation.

I wish there was more info in there about the impact of technical choices on the art pipeline. That's the real killer. Did any of these cut features cause artwork to be recreated?

 

[pc999]

IMO if they want H3 to go against PS3 itself, they will want the best gfx possible, not a fall back from XB, or PS3 fans will have the day of their lives.

 

[Sonic]

I'm sure Bungie knows the limits of their specific engine and if it will be a good idea to rework it for the Xbox 360. If they determine that to be the case then it should be just fine when being compared to PS3 fans having the day of their lives.

 

[pc999]

Taking in acount what some dev had said (most recent Gabe Newell), about old code to those multi-core CPUs, and a SM1.1 to SM3+++, just update the old engine is not a good thing IMO, look at PDZ and at GoW, that what I say about it just look sub-par, period. And H3 being the flagship from XB360 that is not good (hence the PS3 fans coments, and I am a XB fan), on the other hand Rare games are launch games should be harder to do big changes, and meybe Bungie can.

 

[jpr27]

Halo being one of MS flagship titles for the Xbox (and Xbox 360) I cant see them falling back to their old engine. If I remember correctly it was mentioned (I'll try and find the link if someone else knows please list it.) that Bungie was doubling the size of the development team for Halo 3 to try and meet its proposed release time and maybe steal some of the thunder from the PS3 launch.

I'm sure MS will ensure Bungie has everything possible at its disposal to try to not only make Halo 3 a game the fans expect but also meet the launch date against the PS3. Now whether the game is a huge success is anyones guess but I think MS and Bungie know that if they want to build on gaining marketshare against Sony then their flagship title Halo will have to be top notch to compete against Sony franchises not just in graphics but gameplay as well.

I just hope they try to put more emphasis on the single player aspect of the game this round (Although not a terrible story line, I thought it was not as good as the first. I want to play more Master Chief ! ) because as far as MP goes, the people playing Halo 2 on LIVE speaks for itself. Halo 3 MP will probably be an evolution of this not to mention enhanced graphics, and who knows what else they can come up with. Hopefully something that is truely new and a worthy successor to the Halo 2 MP.

 

[Hardknock]

I was under the impression that MGS was using UE3 for all it's 1st party games? I'm pretty sure that's the engine they'll use no?

 

[Platon]

I was under the impression that MGS was using UE3 for all it's 1st party games? I'm pretty sure that's the engine they'll use no?

Maybe, maybe not. MS has licenced the engine so there is a big chance that Bungie will use that one, although they are more than capable of making their own. I suppose that using the UE3 but maybe heavily modified so that it takes advantage of the xbox 360 would be the best think, than spend all the time making a new one...

 

[AlNom]

pretty interesting R&D outlines.

How much memory would they have saved by not using shadow buffers? (Ballpark figure)