Half-life is a good example.

It''s based off the Quake I engine for the most part. I think there was little bits and pieces of Q2 fitted in as well (not much). I remember reading somewhere that Gabe Newell said something to effect of, when all the coding was done and it was ready to shove out the door, about over 60% of engine was Valve code and very little of the original Q1 engine was left in.

If we look at Half-life it classifies as more of a whole new game than full conversion mod. I wonder if in hindsight Valve wondered if they would''ve been better off having "reinvent the wheel" when all was said and done.

TF2 isn''t out, but I remember they originally planned on use Q3''s engine, then found that their goal were beyond it being as practical. Then they opted for the UT engine and came to the same conclusion with that one. So their final decision was to make their own. TF2 isn''t out to assess whether it will pay off big or not, but Valve has proven they have some good insight into making games by giving us HL.

Reinventing the wheel isn''t always more work than is necessary, sometimes it''s less work in the long run. Sometimes it''s more advantageous to build your own engine since it can better support your out of the box ideas. Using an already built engine could present some limits to for your design that would mean recoding a good portion of the engine you''re licensing. If that much extra work is needed to be taken, why not do a little more extra work and go all the way and start from scratch. It''ll have the advantage of being built to your exact specifications and needs, you knowing it from the inside out, and you keeping all the profit instead of having to pay royalties. Also another fringe benefit, is that you may be able to license that engine you built, out to other developers and make even more money.

I think it all probably comes down to what it is you are trying to build. In some cases it may be more advantageous to license an engine, in other cases it may be a mistake in the long run. Obviously it''s going to take more work, if you start from scratch, but in some cases this may be what''s needed to accomplish the full goals. Especially if your gameplay is going to be radically different. Not to mention the side-benefit of all the extra understanding you have from the experience. That means all that you learned will be a useful mental tool to future games you make. As would the engine itself without the hassle of licensing.

I think however if someone doesn''t want to deal with all the hassles of coding but focus moreso merely on content, then probably and existing engine is the way to go. If a person wants to make a whole new game that is beyond the scope of anything else existing, perhaps it''s better to take the wheel concept and create a tank tread. Perhaps though, if you''re going to go to that extra effort then it would be a good idea to make it easily retrofitting to new fundimental features for future game designs. Design it in such a manner that it''s easy to replace specific fundimental parts of it for maximum life out of it. I think as design time is increasing, this is becoming more vital. Like it was a little different back in the days of pacman, and joust. Those games didn''t have nearly the development time as advanced games have now, so it was easily practical to design each game completely from scratch.

Ibanez: why not use 0''s and 1''s


The "language" a CPU "speaks" is not very understandable to
human, so we use translation programs to communicate with the CPU.

The closer the translator''s input language comes to
a human language, the easier it''s to tell the CPU what we want.
(well, somehow, it''s a compromise between
convenience / flexibility, compiling speed)

That''s very basical, you''ll just need it to do complex
software.
Don''t tell me you''re doing an entire game in pure asm nowdays.

Well, you could see a 3D-graphics engine as a "translation program"
to tell the CPU(/GPU) what "draw a polygon" means, but
that''s at a bit higher level, it''s not just a translation of
words (names for simple operations).


So, using a complete engine of some third party,
is not the same as using programming languages instead of 0''s and 1''s.

I don''t say you should not use other''s engines, to set creativity
focus on design, sure, you will have more time for the game then.

But I DO say, that, if you have to use someone''s engine because
your programming skills are not good enough to do it yourself,
then don''t try to program a game, improve your skills!
(for you probably won''t ever finish)

ment to say "draw a level" instead of "draw a polygon".