Metacosm is obsolete ! (Re: [Metacosm-dev] Informal IRC meeting summary 10/9/03)

[Cyril Hansen]

Chris Laprun wrote:

> As far as I am concerned, I don't see metacosm as relevant anymore. 
> The less ambitious new goals render the project pretty much similar to 
> a number of other ones out there. If you want to use C++ and remove 
> the framework aspects of the project, join one of the gazillion MMORPG 
> projects that already do that. Why start a new one?

The ideas of dynamic description, environmental evolution  through 
simulation and dynamic quests generation are still ambitious. I believe 
in incremental development of software, and that the hardest part is to 
get the design issues right. For this we need to experiment, to build a 
working prototype quickly.

I feel confident that once we got a world up and running  in any 
language, rewriting the whole code base in any other language is an 
order of magnitude less work than build it in the first place. And i 
also belive (this is not a rational claim, just an opinion) that we 
should code for today (prototype) not tomorrow  (framework) if we want 
to achieve something. Extending / rewriting the code base as a framework 
can be done later, with an overall smaller cost. At least, this is the 
mantra i currently believe in.

> Another relevant issue is the point made by Julien. Java is more and 
> more used in the gaming space. Actually, it seems that metacosm is 
> retrograde with respect to what's going on in the industry. When 
> everyone was doing stuff in C++, we chose Java. Now that people are 
> starting to realize that Java is not so bad, you want to go back to 
> C++? This change of heart is four years late! Not really an inspired 
> choice... But, then again, choosing to develop yet another mmorpg is 
> not that original either.

I don't know what the others think of that, but I would be really happy 
if Metacosm implemented some important ideas we choose as objective in 
the first place. I would be very happy to have a user base, but i would 
be already happy if we correctly implemented and *demonstrated* some of 
our objectives even if the game is not really playable because it is 
still incomplete. For me, it would already be really *fun*.

Incompleteness is not avoidable during dev,  but of course it is better 
to use the tools we feel confident and productive with.  The problem 
with Java is that we had a lot of expectations on the productivity 
departement, and reality was less fun than expected. As the choice was a 
compromise, we thought about it again. But anyway, we agree to restart 
from the existing code, which means a Java server.


(off topic - my biased and uninformed opionion on the java trend for game)

The trend towards Java game in mobile phones and PDAs make me feels 
strange. Current Palm offerings are based on DragonBall processor and 
ARM compatible processors. All of them have many times the processing 
power and memory of an original Amiga 500, their screen has a comparable 
resolution, but to my knowledge no amiga emulator is available for them 
(This would generate a lot of sales of harware but make software 
developpers unhappy due to unfair competition and make problems with the 
copyright olders for the old games). My point is that there is terrible 
problem of hardware ressource waste in this area.

The problem is that the mobile market is fragmented (Pocket PC / Palm & 
compatibles, Psion, Zaurus, different OS versions, smart phones and 
their different OS). So the developpers cannot afford to spend days and 
months to build a dedicated version for each platform variant. 
Traditionnal architectures like Palm OS are suffering of the compatibity 
weight : Most software must be compiled for the lowest CPU, cannot make 
use of the functionalities of new OS (higher resolution, true colors, 
sound...) while retaining compatibility  with all  the existing user 
base. Their is a gap between Palm OS 4 & 5 for example, which means that 
Palm Os 5 has a lot less software available than Palm OS 4 had.

As the computing power is already sufficient, the idea of a virtual 
machine solution is really interesting. In these hard days of hard 
business, it has the advantage of protecting the investment and allows 
for a larger income : A game developped originally for a PDA can be 
retargeted to a mobile phone at the last minute without too much loss. 
So Java flexibility and universality is not just for developpers - their 
bosses make use of it too.

For the anecdote, the first product that Amiga Corp. has put into the 
market these last years was a bunch of games for Pocket PC. These used a 
proprietary virtual machines, which is supposed to be available on their 
future home machines. So Java is not really the only solution. And it is 
not really an engineering solution, it is an economic opportunity 
related to the current state of the market.

Regards,
Horus