[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Loki...

Josef Spillner wrote:

> A few years from now, when the relative short time of free entertainment
> development has caught up with the many years of say free server development,
> the barrier for non-free games will be even higher... a fact which I do
> appreciate very much.

Well, the problem right now is that the barriers to medium to high quality
*free* games are almost insurmountable.  We need commercial games in order
to have any halfway decent games *at all* for the Linux platform.

Without games, we'll have a very hard time in the desktop market.

That's why we *needed* Loki - and I'm sad to see them go. (I have every
game they ever produced - so I did my part to try to keep them afloat).

It's not a matter of free games development "catching up" with server development,
it's a matter of the basic difficulty of doing it.

The Source code for Quake-1 - including the server and the necessary support
tools - is 250,000 lines of code.  That's pretty comparable to a typical large
OpenSource project - Python (for example) contains about 120,000 lines,
Apache is 120,000 - GIMP contains 250,000 lines.

But Quake also has 57 megabytes of 3D models, textures, etc - just for the
basic demo levels.

Quake 3 is *MUCH* larger, return to Castle Wolfenstien is larger still.

So as good as the Linux server development is, just think about "rewriting
Apache *and* Python from scratch" as the magnitude of the task for *each*
good game you expect to see.

The difficulties with coming up with a good game like that in OpenSource
development are *many*:

  1) With most OpenSource packages, the developers write it because they
     want to *use* it.  Nobody wants to play the game they just wrote
     because they already know all it's little secrets and are heartily
     sick of the sight of it by the time they are done.

  2) A package like Python or Apache has dozens - if not hundreds of
     contributors.  Most games have to be written by a crew of between
     one and three people with maybe a couple of occasional contributors.

  3) Artwork is by far the biggest job in producing a modern game.  It has
     proven almost impossible to attract artists to the concept of doing
     work for nothing.  There have been *long* dicussions about this all
     over the place - the conclusions are always the same - there are almost
     no OpenSource artists out there.  Music and sound effects present similar
     problems.  Programmers *rarely* make good artists.  The artwork my son
     and I did for my Tux games is just pathetic compared to even the worst
     commercial work.

  4) Games are short-lived phenomena.  People play a game for a few weeks
     or months at most - and then move on.  It's rather depressing to spend
     two years of your free time writing a big OpenSource game - only
     to find that your fame lasts for a month and then nobody downloads it
     anymore.  Better by far to write a simple KDE utility in just a few weeks
     an see your work appreciated, used and extended for years to come.

  5) Only one in 35 commercial games makes a profit.  OpenSource games don't
     have to be profitable - but to be worth writing at all, you DO want them to
     have a reasonable "happy-audience-to-effort" ratio...which is kinda the
     same thing.  It's tough to go into an OpenSource development that's going
     to suck up all your free time for a year or more knowing that there is
     only a 3% chance that it will be loved.

  6) With non-games, you can release a version that's only partially complete
     and people will join the project - offer patches, etc until it's finished.
     If you release a game that's not "playable" - nobody will download it.

There *are* some things we can do to improve this situation (I have attempted
most of them) - but it's still pretty grim.  It's hard to point to a single
OpenSource-developed game that would sell for $29.99.  Almost everything
that's been written falls into the "100 Games on one CD for $5!" catagory.

The best OpenSource games would look good back in the days of the Amiga
and the Atari-ST - but they are hopelessly outdated by 21st century standards.

My efforts (Tux - A Quest for Herring, TuxKart and now "The Chronicles of
the Evil Overlord") have tried to circumvent these problems.

  * I'd hoped that by making a game where it was relatively easy to add
    levels of your own (TuxKart), I'd get lots of people contributing game
    levels - which would make the game bigger with little effort on my part
    - and also make the game more interesting for me to play.  I never got
    a single level contributed.

  * I put most of the code into a library that could be used (and therefore
    contributed to) by other Linux games writers.  Hence, TuxKart contains
    only 5,400 lines of code and the PLIB library has grown to almost 63,000
    lines.  That is the only thing that has enabled me to write three games
    in about 3 years...it would have been impossible otherwise.

  * I try to come up with game concepts that don't need much artwork.  That's
    proved to be almost impossible.  Games that don't have much artwork seem
    to always look crap - that's not really a suprise I guess.

  * I try to build games that are somewhat like successful commercial titles,
    without directly cloning them.  This gives the game a better chance of
    being liked.

But for all that, I don't feel that the process has been terribly fulfilling.

It's very depressing - but we *NEED* commercial games for Linux.

----------------------------- Steve Baker -------------------------------
Mail : <sjbaker1@airmail.net>   WorkMail: <sjbaker@link.com>
URLs : http://www.sjbaker.org
       http://plib.sf.net http://tuxaqfh.sf.net http://tuxkart.sf.net
       http://prettypoly.sf.net http://freeglut.sf.net
       http://toobular.sf.net   http://lodestone.sf.net