🔒 Locked

Engine coding priorities?!

Started by Mephs Jan 24, 2003 at 11:45 AM 1 reply 200+ views

Original Post

Jan 24, 2003 at 11:45 AM

Heyas, This is probably less of a technical question than is usually found in a technical forum, but I''ve been busy creating an engine for myself and came round to wondering about setting priorities. So far I have a multitextured heightmap consisting of multiple meshes, I''ve added quadtree generation to help with the performance and just put the finishing touches to a makeshift particle system (it will need updating later as I have things in mind for it) that fulfills the role of creating a fired energy blast effect. Those are the major developments in my engine to date but I''m wondering what to work on next and how to decide. I''m not asking what I should make next, rather I''m asking how do you decide what aspect of your engine you are going to work on next? I''m thinking of perhaps making a quake style console next as this would help in-game debugging and offer a greater degree of control over the program, so it would be very useful.. but then my terrain is not yet lightmapped and I could implement continuous level of detail to further aid the performance, and of course my routine to generate a texture from multiple textures combined and blended according to the height of the map at that point needs some tweaking, and I will need a full blown particle system at some point... so many things that could be fun to work on... how do you decide!!? Steve AKA Mephs

Dreddnafious Maelstrom

272

Dreddnafious Maelstrom

Jan 24, 2003 at 12:07 PM

If you havent done any animation yet, thats always an impressive addition to an engine. Of course i found it really tough, but it definitely takes your engine to the next level.

Once you get your animation done your next step is a full meal deal billboarding implementation, which will allow for a bunch of cool stuff. (im assuming you are using point sprites for your particle gen. )

of course im basing these on eye-candy effect, and no real pragmatic reasoning.

jamessharpe

Jan 24, 2003 at 12:15 PM

I would personally say implementing the console will be a great productivity boost, esepcially if you can change game states on the fly such as map loaded, collision detection on/off, render states on/off.

Also try to implement it so that any messages that you send to a log file are mirrored on the console. This enables you to access any error messages that are logged without leaving the engine.

The way I''ve started to implement my console is as it being an ''EventManager''. This way it knows everything that is going on in the engine. All player movements, user input etc. passes through this one point, and then the messages get distributed to the appropriate ''clients''. The advantage of this system is that it will become easier to implement network stuff, becuase it just means extending my EventManager class to transmit appropriate messages across the network.

Basically what I am saying is that the most important part of your engine is it''s structure. A good structure will enable a much quicker implementation of features. An example of this is that all my messages have an option to be validated. I use this system to implement my collision detection since the player movement needs to be validated. I''m sure this system will also allow AI to be implemented relatively easily, because I know that all the state changes will pass through a single point.

Although eye candy is rather nice

, just don''t comprimise engine structure for eye candy that could be added later.

James

Topic Locked

This topic has been locked by a moderator. New replies are not allowed.