How to create your own game company

How To: Create your own game company,

Posted May 19th 2006 8:00AM by Victor Agreda, Jr.
Filed under: Business, Design, Developer, Finance, Games, Internet, Utilities, Features, Windows, Macintosh, Linux, Windows Mobile, Symbian, Palm, Productivity, Web services, Shareware
how to create your own game company part twoIn part one I talked about a lot of the production tools: 3d apps, game engines, and so on. This time I want to tie that in to the business end. How do you run the company, and how do you make money?

First of all, I discovered PopCap Games offers their engine to developers for free. The license agreement spells out the details, but primarily all you need to do is acknowledge using the PopCap Games Framework. The practical upshot of using PopCap's engine? They might publish it. The practical upshot of getting an engine and look by a successful publisher? You can make money. You still have to file taxes, and do all the other business stuff you'd normally have to, but you get to skip the part where you worry about marketing, payment, etc.

One other development is the rather old Torque Exporter Plugin for Blender is now up to version .91. Since Torque is $100 puny bucks for an indie license, and Blender is free, that also seems like a reasonable choice for budget game developers. If you're still thinking 2D (or very limited 3D, like an isometric game), another option is the Arcade Engine for Runtime Revolution.

So lesson #1 in business: keep your costs low. I don't want to get into a lot of business advice, but if you want some great tips on what NOT to do, check out Ten Easy Ways to Screw Up Your Game Company, and Part 2 of the same. If you're really looking to make a company, and can get investors, there's a lot of funny and valuable lessons to be learned in those few pages...

And not to be redundant (which I'm good at), but Lesson #2? Pick the right product. I know, the author of Ten Easy Ways says the same thing, but this lesson is too important to miss. You can spend thousands on the Unreal engine if you like, but trying to make another Halo? Good luck. It's not impossible, but unlikely. Shoot for realistic early goals and clean up with original, playable content. Now, on to the software!
Building your team

So you've got an engine, you've got an idea, now what? Unless you're some kind of modern-day Da Vinci, you're going to need people to help out. I'm no HR expert, but I can safely say this might be the worst, most challenging, yet potentially most rewarding part of the process. For one thing, a great staff member can elevate the team to new heights. A bad worker can literally bring a project to its knees.

hiring peopleWhat to do? Be very honest with candidates. And force them to be candid with you. One of the coolest approaches I've seen to hiring involves a test in the form of a game from Procter and Gamble. You can start asking friends if they know anyone with the skills you need, and move outward that way. If you're lucky enough to live in a place with a major university nearby (as I do), another opportunity is offered in the form of student workers. And lastly, there's always online recruiting. Online is tough, even with the breadth of choices. It's just very difficult to gauge someone from a distance (kinda like it's hard to convey sarcasm in an email). The first place I usually look or post is Craigslist, and there's probably one in your area. If you're really desperate, there's always a professional solution like Elance.com. Mine those forums and consider targeted markets like GamesIndustry.biz and GameDev.net for talent with experience.

Personally, I've always been a fan of testing personnel. And since you're looking for people with a passion for games, why not host a LAN party, put them all on the block, and have a quick deathmatch? This won't necessarily tell you how well someone can allocate memory or manage UV texturing, but it'll tell you who the griefers are. Don't hire the griefers!

At the end of the day you need to at least have leads in the major development areas: Programming (or tech lead), Art, and Design. Programmers need to be drilled on their skills pertaining to your particular platform. If you need a Python programmer for Blender, don't quiz them on ActionScript. Artists should have a portfolio, and the ability to whip up some artwork on the spot. As for design, I'm referring to the game design. Chances are the game designer, or "vision keeper" is you. I'd still recommend bringing in someone with game design education or experience, even if they are only an advisor. Make sure to check this person's credentials though. Playing Kingdom Hearts for a week without sleep doesn't make you a game design expert.

Documentation

Game documents, who needs 'em? Everyone. Every TV show has what's called a Show Bible. This is the law and lay of the land. Is the main character a pedophile? That little detail might never be revealed on the show, but the actors and directors need to know it, as it would influence the characters and stories on the show.

The same applies to games, only more so. You'll need a Game Design Document first. The Design Document is essentially the vision, point, and basic rules, setup, and look of your game. It is truly the "bible" for your game. Extending out, you'll need an Art Bible, with examples of the look and feel of the game, the design of characters, basically any art in the game. And you'll want a Technical Design Document. This spells out things like your target platform, tools to use, and the potential problems with programming your game. If you're making a Flash game, how are your programmers going to simulate 3 dimensions?

The reason for these documents? Whenever there's an argument over what to do, consult the bibles. An Art Bible ensures a unified look for all the art assets, and makes your game look more professional. The Tech docs make sure the programmers know what to expect. And the vision absolutely must be spelled out. Listing everything you need for a game is an exhaustive process, but one you're going to have to do if you expect to ever finish the thing.

Project management and communications

Project management, in the simplest sense, is very simple. Take the end result desired (make a playable game), and start breaking this down into actual tasks, then plan those tasks across a realistic timeline. Check every day, week, month at different levels (big picture vs. details) to make sure those "milestones" are getting done as appropriate. Looking for an alternative to Microsoft Project? Personally I use a meatspace tool called a legal pad, but those of you wanting software should check out Open Workbench.

communicationsAs for communication with the team, the usual stuff works wonders. Instant messaging with Meebo allows you to be protocol-agnostic. Jotspot offers a pretty good wiki service. And Google's Calendar and Notebook might be just what you need to push dates and talking points. Although I personally think Basecamp is the bees knees. And then there's that old standby: email. I think Gmail has a pretty good advantage here too, due to Calendar and chat integration, plus a powerful search ability.

Each department is going to have specific needs however. Artists need a place to store and log graphics. Programmers need a CVS for managing code. Let's look at each department to see what they need.

Design: I consider design the actual game design, not art assets. Consider an all-purpose CMS like Drupal (which I'm a big fan of), or Typo3, which is actually used by some Blender nerds. What you're wanting to do is have a central repository for "the vision" of the game. A CMS is a good place to put those design documents.

Programming: these guys need a CVS! Essentially this is a way to check in or out code, without messing other stuff up. Managing code is a tough process, and I'm a hack myself. Besides, coders tend to be religious about ways to do this, so to avoid a holy war, I'll just point you to a great book on the subject-Open Source Development with CVS by Karl Fogel and Moshe Bar. Almost everything else is dependent on your platform of choice. Flash, as an example, has its own IDE for ActionScript inside the app.

Art: ultimately these are the guys with a need for a good filing system. This could be as simple as a file server, set up with a very strict nested folder system. For example, a folder with "Characters" could include subfolders of "NPC's," "Heroes," "Bosses," and so on. Inside those would be the model folders, texture folders, and perhaps backup folders for previous versions. You can also use CMS's like Drupal to set this up, as the nature of the CMS versatile enough to handle art assets.

Finally, there's another good ebook to check out: Producing Open Source Software, also by Karl Fogel. Karl's book is a wonderful introduction to the problems (and rewards) of working with a large group of squirrels, each one with their own agenda. If you're looking for more workflow ideas, check out this post.

Testing and marketing: mutually exclusive?

The QA portion of game development is generally a time when the time grows very large. Bug testing can be really expensive, but I don't think there's a need for all that. At least, not when you're the underdog! Years ago I read the book Guerilla Marketing, and it's had a profound influence on my view of marketing. This might be sacrilege, but I suggest you release the game as a limited public beta, and somehow reward bug finders. That's not a new idea (Google I'm looking in your direction), but might take some of the bug hunt pressure off your programming team. The key is to be very open about the stage your game is at. If it's really, really rough, maybe only test with some friends. Once you can more or less successfully navigate every level, that's a good time to open it to the world. If you're really wanting to spend the cash on QA, and I'm not denouncing that approach, Game Instict offers QA-for-hire services.

The bottom line is QA must be done. There are no shortcuts. During the development of Jak X: Combat Racing, the Naughty Dog team tested the online play every Friday night. They'd go home, dial in to a private server, and play against each other. Not only did this uncover code issues, it allowed them to collaboratively discuss gameplay issues. If you're really passionate, this won't even seem like work.

Making your money

So you chose a reasonable target platform that doesn't cost too much to develop for, you have a great team, and you've managed to test the game until the play is just perfect. The next step is to sell your game.

making moneyYou can find a myriad of advice columns on how to approach a video game publisher to sell your game. Some publishers come from a development background (like PopCap and now Gamelab) and might appreciate you more. Maybe not. The advantage of a publisher is that you get an expert in sales and marketing. That's the theory. As any musician will point out, whether or not that expertise is used is another matter... But the big point is to have a complete, playable game, free from licensing issues (like music). It helps to have an idea of how to sell your game too, especially if it is very unique. Publishers tend to be a little conservative...

So I say strike out on your own. Between guerilla marketing and online payment systems, I don't see why you couldn't at least generate a little cash yourself. This might even be enough to woo a tough publisher. So how do you sell online? I mean, aside from PayPal.

Unfortunately, Regshare.com isn't around any more (they're selling the domain, ugh), as they used to do comparisons of online payment systems. But probably the most recognized name in the business is Kagi. If you've ever bought a 3rd part Palm app, you might recognize them. MacDevCenter has a comparison of Kagi vs. DigiBuy and eSellerate, for those still looking to shop around.

You could also implement your own payment system, but that is as varied as the mechanism and platform you're developing on. My own company just uses Paymentech for processing credit cards, but I've found that's overkill for an indie shop. If you're just moving bits around, Kagi works well. If you're selling physical units, consider just setting up shop on ebay (or wherever you like) and go nuts.

Other resources

I am a regular reader of Gamasutra, and you should be too. Sloperama's advice pages are chock full of great industry info, on everything from game programming to the actual gameplay design. You might want to read and check out the blogroll on Greg's blog on Games, Design, Art, and Culture. If you have some time and want to have a little fun, check out GameGame, a card game about designing and pitching games.

If anyone has any other great sites (I know I've left out a ton), please post 'em in the comments. There's no easy solution to making your own game company, let alone a truly saleable game. But hopefully with enough understanding of the process you'll be a little closer to realizing the Next Big Thing on your own.
read more | digg story

Streaming MythTV to your cell phone

Streaming MythTV to your cell phone

Electrical tape wearing thin holding together my old cell phone (Samsung i500), I decided it was time to give in and upgrade to a new phone. I drank the kool-aid on a 2 year Sprint contract given the discounts both on the phone and in my monthly service, and in the end settled on the Samsung a920. It's an EVDO enabled phone with Sprint's $15 unlimited monthly data service. Playing around with the mobile TV functionality (that's an extra $10/month but I was enjoying the first month free), I had one of those lightning bolt moments.

Why not stream my own video to the phone? Better yet, why not just automate my MythTV to convert my recorded programs and automatically have them ready to be streamed whenever I care to watch them on the phone?

A bit of research later, I discovered SlingBox can stream your tv to your phone, but it needs to be a Windows mobile phone and then there's the monthly service fees and the box to buy. I also found random mythtv devotees with similar ideas at least as far back as January 2005, but couldn't otherwise find a concise guide or more information. Inspired by ZooVision, I knew it was possible for users to stream their own content to their phone, it was just a matter of putting the pieces all together. A couple hours of tinkering later, and I've got a working solution... my "tivo" on my cell phone wherever there's sprint evdo access. So here are the steps:

1. Get MythTV up and running. I won't go into that, but it's worth the hassle even if I couldn't stream it to my cell phone.

2. Follow these directions which will get you an updated version of ffmpeg (named ffmpeg3gp) and a myth3gp script which, in combination, allow you to take Myth recorded video and convert it to 3gp, a mobile video format standard that most multimedia phones support. However, instead of using the myth3gp script from there, you'll likely want to use this myth3gp script as I've included a couple of necessary changes. Save that txt file as /usr/local/bin/myth3gp on your Myth box.

3. Create a directory to hold your 3gp videos on your Myth system. Open up the myth3gp script from the previous step and change the line that says:

out="/mnt/drive2/myth3gp"

to:

out="/directory/for/saving/your/3gp/videos"

4. Install gpac. Once you've compiled and installed gpac, it will provide you with an executable called MP4Box which is used to convert the 3gp file from step two into a 3gp file that is streaming enabled. The myth3gp script calls MP4Box to take care of this for us.

5. Find an internet connected server which can house your 3gp videos and be available to stream them at will to your phone. On said server you'll need to install Apple's Darwin Streaming Server. Installing it can be a bit of pain, but just be sure to have port 554 open so that it can handle the rtsp streaming protocol. For me, getting Darwin to work took the most effort. If your MythTV box has a static IP and is internet accessible, that should work fine, but in my case I'm uploading the videos to another server.

6. We're in the homestretch. We've got programs to convert Myth recorded video to a phone friendly format, and we've got a server that can stream them to our phone. We just now need an automated process to convert the video and upload it to the server running Darwin and we'll be in phone tv nirvana. Myth can do this automation for us. You'll need to exit out of Myth and run:

mythtv-setup

From there, choose the "General" menu item and hit enter until you get to the "Job Queue" screen. Put a checkmark in "Allow User Job #1 jobs and continue to the next screen until you get the page with label "User Job #1 description". Give it a description like "Myth 3GP" and for the command, use (note the quotes):

/usr/local/bin/myth3gp %DIR%/%FILE%

"%STARTTIME%~~~%TITLE~~~%SUBTITLE%"

Save your changes, exit, restart the mythbackend and restart Myth.

7. With Myth restarted, go into Utilities / Setup -> Setup -> TV Settings -> General and click through until you get the "General (Jobs)" page. On this page, put a check in the checkbox next "Run User Job #1 On New Recordings". This will ensure that our process gets run after each new recording. Important Note: For all existing recurring recordings that you had prior to setting up Myth3GP, you'll need to manually edit the recording options and in "Post Recording Processing", you'll need to switch its setting so that it says "Run 'Myth3GP'".

8. If your using your local Myth system for Darwin streaming, just be sure that Darwin knows where to find the 3gp files on your system. Otherwise, edit the myth3gp script from step number two and set the darwin_username, darwin_hostname and darwin_dir as appropriate for your external server. At the bottom of the script, it uses scp to transfer the file from your Myth system to your Darwin server. Though, for everything to be automated, you'll need to set up scp to not require a password. Directions for password free ssh/scp are short and sweet. Once you can transfer files between systems sans password, you are good to go.

9. The final step. On your Darwin machine, make sure you have a web server running (apache), php installed and place this php script somewhere in a web accessible folder. It's job is to scan your 3gp movie folder for movies and generate a webpage with rtsp:// links so that you can access your recordings from your phone. Edit the php script and change $directory to the directory path for your 3gp videos and set $hostname to your hostname. With that installed, you're done. You can manually convert your old recordings by starting the "Myth3GP" job on them and any new recording will automatically get the mobile treatment. Fire up the URL for the mythmobile php script and start streaming.

Triumph. Indeed some beautiful uses of fair use. Fair use to record the tv program to my hard drive... Fair use to convert the video format to one viewable by my cell phone... Fair use to stream it to my cell phone for my own personal enjoyment. Now imagine trying to do any of this with the broadcast flag in place.
read more | digg story