And then everybody says "But! It's a valuable Learning Experience! How dare you suggest that I shouldn't learn anything!" with LaMothian indignity.

Clue #1: you can learn just as many things from making games as you can from making engines. also, you learn quite a bit simply by using an engine, because you need to become familiar with how it works in order to use it effectively.

Clue #2: engines are made to be reused. think about that realistically and ask yourself if the engine you are making will actually be reused. Probably not. If, for some reason, you do actually finish your engine, you will use it for a single game, and that's it. Now you've got all of the time you spent on making the engine gone, and then all of the time you spent making your engine work in your game.

Clue #3: as I was writing this little rant, i looked up at the ad placed on the top of the page. guess what it was for? a GAME ENGINE! For

ONLY $100!

Clue #4: as much as you hate to admit it, you probably don't have the skills required to make a AAA game engine. just face it and move on.

But people forget. John is The Man. He has almost certainly sold more copies of his games than you ever will, and each of the games took less than a month to write. Don't listen to him, and instead go about making your Massively Multiplayer First Person Real Time Tactical Role Playing engine, because apparently *YOU* know better.

What is that repetative tapping sound??? Oh wait, it's you two raining on everyone's parade... [smile]

John makes a good point.

I recommend writing to a purpose not to an idea. If you want to create a game then create a game. If the complexity of the game leads you to need an engine then chances are you are better off buying or using an existing one. A good game engine takes an enormous amount of time to create.

As John has pointed out before time is not free. Even though you may not be paid for it at the time it is a limited resource and therefore has value.

Say you could write a decent engine in 3 months (likely a huge underestimation). That's 13 weeks. Assuming you worked 40 hours a week that's 520 hours. Assuming that your time is worth at least minimum wage (I think it's $5.15 at the moment) that's $2678. If you worked 60 hours a week for those three months that's $4017.

If you're writing a small game then just write to the game. After writing several games you'll see what you're doing repetitively so generalize that into a library and/or a tool. However, if you get to the point where something is going to take more than several days it's probably time to do a quick look around and see if you can buy it.

Daedalus: Good one :-)

BTW, if you really feel like you have to write a major engine the seriously consider contributing to an existing one by adding to the core framework or writing tools/plugins to support it. Stand on the shoulders of others. If you feel that something sucks about it then don't sulk off and write your own engine. Try to fix the existing one.

Excellent point, MauMan. Many of these existing engines are community works-in-progress. They're always moving in one direction or another. Many, like Torque or CrystalSpace, are getting new features and add-ons to make the product more capable. Some, like the old Doom engine, aren't getting more features but are getting ported to new platforms all the time (like PocketPC and Nokia S90 phones).

The point is that these products are far from static, and standing on the shoulders of giants (or at least very large midgets) is rarely a mistake.

I often hear people saying things like "I don't like Torque because it doesn't do <insert obscure feature here>". The fact is that top-notch stuff has been done with Torque (or the Quake engine or CrystalSpace, or SDL or whatever). Sorry to burst your bubble, but people have managed to make games that are BETTER THAN THE ONE YOU ARE WRITING without that aforementioned obscure feature.

Heh. [smile] I'm done my engine already; it's game time. But I have every intention of checking out the likes of Torque and Irrlicht (or whatever) when it's done. Writing your own engine from scratch is indeed a good learning experience ... but it's also a PITA! [lol]