The Woes of Design Documents
by Dan Mattia · 06/27/2006 (2:12 pm) · 16 comments
I have begun writing the design document for Tales of Fyrndell Tech Demo, the first project of Overdrive Games. Being an organized person, I have divided up the design document into separate .odt files (OpenOffice's version of .doc), which are then compiled into the "master" .odt file. After organizing everything for solely one faction in the game, I already have 20 pages of the design organized and planned.
This is my first design document, for what is to be ultimately a 30 minute to 1 hour long playable demo. Once I begin filling in the information in the design document, I can't imagine how many pages I'll have. It's a lot of work, but putting down my many thoughts onto paper (albeit virtual) really helps to organize what is solely an idea, into something feasible.
Tales of Fyrndell is going to be an amazing game. The demo is not going to even justify the actual game, mainly because we don't have the development time or funds to do what we want in 30 minutes to 1 hour. The demo is mainly meant to show to investors to request funds for us to design the actual game - X episodes of 6-8 hours each.
We have still not chosen an engine, but there are two engines we are looking at, with a third option available.
Firstly, I'd like the Offset engine, being developed on by Offset Software. I do understand they've been very "hush-hush" about the whole thing, so I'd like to get the full design document, concept art, and some models done before I inquire about using their engine. If we cannot acquire the Offset engine for the demo, or mod Project Offset (the game) to produce the demo, I still definitely want to use the Offset engine for the final game. It'll make things so much easier and provide us with the technology we need for Tales of Fyrndell.
A second choice is the Gamebryo engine, famed for being used by Bethesda for Oblivion. This is probably what the tech demo will be developed on, due to its moddability and availability currently. It probably won't be used on the final game, but it's a good choice if we cannot acquire Offset for the demo.
It's also been suggested that we modify Torque, though editing the engine for what we'll need to do may take away a ton of development time that can't be spared. It's still an option if all else fails, though.
No matter what, there's 10 months to get a 30 minute to 1 hour tech demo of Tales of Fyrndell completed. Will we do it?
...Why, of course we will. I expect nothing lesser.
This is my first design document, for what is to be ultimately a 30 minute to 1 hour long playable demo. Once I begin filling in the information in the design document, I can't imagine how many pages I'll have. It's a lot of work, but putting down my many thoughts onto paper (albeit virtual) really helps to organize what is solely an idea, into something feasible.
Tales of Fyrndell is going to be an amazing game. The demo is not going to even justify the actual game, mainly because we don't have the development time or funds to do what we want in 30 minutes to 1 hour. The demo is mainly meant to show to investors to request funds for us to design the actual game - X episodes of 6-8 hours each.
We have still not chosen an engine, but there are two engines we are looking at, with a third option available.
Firstly, I'd like the Offset engine, being developed on by Offset Software. I do understand they've been very "hush-hush" about the whole thing, so I'd like to get the full design document, concept art, and some models done before I inquire about using their engine. If we cannot acquire the Offset engine for the demo, or mod Project Offset (the game) to produce the demo, I still definitely want to use the Offset engine for the final game. It'll make things so much easier and provide us with the technology we need for Tales of Fyrndell.
A second choice is the Gamebryo engine, famed for being used by Bethesda for Oblivion. This is probably what the tech demo will be developed on, due to its moddability and availability currently. It probably won't be used on the final game, but it's a good choice if we cannot acquire Offset for the demo.
It's also been suggested that we modify Torque, though editing the engine for what we'll need to do may take away a ton of development time that can't be spared. It's still an option if all else fails, though.
No matter what, there's 10 months to get a 30 minute to 1 hour tech demo of Tales of Fyrndell completed. Will we do it?
...Why, of course we will. I expect nothing lesser.
About the author
#2
I do know that licensing the Gamebryo engine would cost $50,000 for the entire studio, but that's a lot of money we don't have right now. The whole purpose of the tech demo is to garner funds, so we may just end up modding Oblivion, which uses the Gamebryo engine.
06/27/2006 (2:28 pm)
We haven't asked Offset yet, and they probably don't even know we exist right now. I want to get a decent presentation together before asking, so that they know we're serious.I do know that licensing the Gamebryo engine would cost $50,000 for the entire studio, but that's a lot of money we don't have right now. The whole purpose of the tech demo is to garner funds, so we may just end up modding Oblivion, which uses the Gamebryo engine.
#4
06/27/2006 (2:51 pm)
Thank you Surge! It's appreciated! 
#5
06/27/2006 (3:46 pm)
I'd go with modding Oblivion if the sole purpose of the demo is to get some funding. It won't cost you anything but your time.
#6
06/27/2006 (3:49 pm)
Right now, that's what it looks like we'll be doing, for the demo, at least. Gamebryo provides us the best engine to work with, technologically, and Oblivion's a good game to mod for our tech demo.
#7
Wow Dan. 20 pages in 3 days? How much caffeine are you drinking?
06/28/2006 (2:33 am)
Well, Offset would be the best for the FPS side of things, at the least. But we do not know how it would handle massive numbers of unit AI's. Hypothetically, the game will be done when Dual core processors are common place. Then you could move all AI code over to the second processor.Wow Dan. 20 pages in 3 days? How much caffeine are you drinking?
#8
I wonder how you rationalize those kind of decisions. Based off some nice movies?
I'd stick with gamebryo for your game. At least you know that you can actually make it. I'm sure modding Oblivion would do it for demo purposes.
Forget the pretty pictures, you know nothing about the quality and reliability of thier technology. I'd be vastly wary of using tech that hasnt been shipped in a game yet, unless I'd been using it a while (ala Raven and ID kind of thing).
06/28/2006 (2:53 am)
So you think that an engine that you've never seen used in a game is more suitable than using TSE which admittedly hasnt itself been used in a game yet, but which clearly is built on a core that has been used a lot.I wonder how you rationalize those kind of decisions. Based off some nice movies?
I'd stick with gamebryo for your game. At least you know that you can actually make it. I'm sure modding Oblivion would do it for demo purposes.
Forget the pretty pictures, you know nothing about the quality and reliability of thier technology. I'd be vastly wary of using tech that hasnt been shipped in a game yet, unless I'd been using it a while (ala Raven and ID kind of thing).
#9
http://www.raphkoster.com/2006/06/26/40-ways-to-be-a-better-game-designer/
24: Giant design docs are useless
They are usually overelaborated piles of daydreams that nobody will actually implement. A bulleted list of specifics is far more fruitful.
06/28/2006 (6:30 am)
From Raph Koster's "40 ways to be a better game designer"http://www.raphkoster.com/2006/06/26/40-ways-to-be-a-better-game-designer/
24: Giant design docs are useless
They are usually overelaborated piles of daydreams that nobody will actually implement. A bulleted list of specifics is far more fruitful.
#10
Phil: Thanks for bringing that up. I'm going to email them with a request for information regarding the engine.
Rob: We're not trying for a huge design document, just an organized one. The 20 pages are mostly blank with headers and basic info. All information on it is then bulleted and includes concept art, if applicable.
06/28/2006 (6:40 am)
Bryan: No, no caffiene ;)Phil: Thanks for bringing that up. I'm going to email them with a request for information regarding the engine.
Rob: We're not trying for a huge design document, just an organized one. The 20 pages are mostly blank with headers and basic info. All information on it is then bulleted and includes concept art, if applicable.
#11
Now thats a design document ;p
ps:
So its only going to run on shader 3.0+ hardware and will probably
crawl on 32 bit machines (i.e. require 64 bit).
~neo
06/28/2006 (7:12 am)
Game that is damn fun to play and sells a million copies. The EndNow thats a design document ;p
ps:
Quote:
The Offset Engine uses a fully 64-bit floating point HDR rendering pipeline.
So its only going to run on shader 3.0+ hardware and will probably
crawl on 32 bit machines (i.e. require 64 bit).
~neo
#12
We may not use Offset, however--we've been discussing amongst ourselves a bit and collecting some feedback. There definitely needs to be more research done before we decide on an engine for the final game, though we have decided on one for the demo.
06/28/2006 (5:57 pm)
Well, the full game of Tales of Fyrndell will be pretty hardware intensive, unfortunately, so 64 bit will probably be used amongst our target market. Besides, 64 bit processors are becoming more and more mainstream, so by the time the full game is released (at least a 1.5 years), most of the market should have 64 bit.We may not use Offset, however--we've been discussing amongst ourselves a bit and collecting some feedback. There definitely needs to be more research done before we decide on an engine for the final game, though we have decided on one for the demo.
#13
06/28/2006 (7:23 pm)
Typically, the main advantage of 64-bit processors is memory space, not speed, even when doing 64-bit math. Unless the game will need 2G or 3G of RAM, requiring 4G of system memory, the speed advantage of a 64-bit processor would be minimal.
#14
around (not talking MMX, etc here) in memory in calculations it will probably be slower.
Perhaps not crawl but I wasn't really pushing that part of it ;p
~neo
06/29/2006 (2:16 am)
64-bit floats (doubles) are emulated on most 32-bit machines so if they're shoving a lot of emaround (not talking MMX, etc here) in memory in calculations it will probably be slower.
Perhaps not crawl but I wasn't really pushing that part of it ;p
~neo
#15
06/29/2006 (4:34 am)
Hry, Dan? I was right about how suitable the Jupiter engine would be. Wrong about the price though - same as gamebryo. 
Torque Owner Bryan Stroebel