Browse freeroleplay: This Month's Archive • Main Archive Page • Related posts • Previous by Date • Next by Date
[freeroleplay] Re: Generics

From: Ricardo Gladwell <ax0n@xxxxxxxxxxx>
To: freeroleplay@xxxxxxxxxxxxx
Date: Tue, 11 May 2004 17:35:12 +0100
Per Inge Mathisen wrote:
> On Tue, 11 May 2004, Ricardo Gladwell wrote:
>>Alternatively, you could implement all the rules mechanics in XML.
> 
> I am not sure you understand the difficulty involved. A computer game
> doesn't just compute stuff, it also formats the results and display it to
> the player, fitting it into the rest of the graphics of the game. It needs
> to interpret the result and give it visual and audio results.

I agree, the issue here is that the by completely seperating the 
roleplaying rules and content from the actual code that runs the game is 
that you create an huge abstraction divide between the two. The code and 
programmer needs to know what to expect from the ruleset in order to 
represent it graphically.

For example, every RPG has widely different spells, and most CRPGs have 
very specific graphical effects for each individual spell. One system has a 
fireball spell, the other has a lightening spell... how do you know what 
spell to expect and how to relate that information to a specific visual 
effect if the list of spells itself is highly open and variable?

There are two solutions:

a) When writing your game you only ever use a specific XML rule-set with 
the RPG library so you know what to expect from it and can more tightly 
bind the game design to it. Because the XML is published in a completely 
different, non-software format and is only accessed at runtime it 
(probably) circumvents many of the legal difficulties surrounding 
distributing material with the open/free licenses such as the GPL. These 
problems are further lessened by distributing the code under the LGPL.

b) You have some very broad categories of 'effects' and 'states'. For 
example, all RPGs share states such as death, scaling damage (i.e. hit 
points) and sleeping, and all RPGs share effects like magic and combat 
attacks. In the example of magic, you would simply have a magic 'effect' 
global which could then be represented in the game with some sort of 
graphical 'twinkle'. This solution means that magic ends up looking very 
'samey', but it does allow for much greater flexibility and makes it easier 
to add new spells to the game.

> Can you make a computer game which reads a generic ruleset and shapes
> itself to conform? Yes, it is possible. However, it adds a lot to your
> work, and, the amount of extra work probably increases squared by the
> genericness of the ruleset beneath. The less assumptions the coder can
> make about the underlying ruleset, the more code he has to write to cover
> all the possibilities.

Alternatively, it might reduce the workload: far more of the rule-set is 
abstracted away and the work done for the developer, so rather than 
worrying about implementing a complex rule system the developer can get on 
with the work of writing the interface, AI, etc.

> It might be possible to write a software library which can cover several
> rulesets, and games that use it, but these rulesets would have to conform
> to some overall standard which says which mechanics are allowed and which
> are not. This would add strict limitations to the game rules - for example
> limits to the number of attribute types (attributes, skills, talents and
> statuses for Generics), perhaps a limit to the number of primary
> attributes, etc..

I would argue that most rulesets conform to some very basic overall 
standards: all characters have traits and can carry out certain actions, 
these traits affect the results of actions, and all actions carried out by 
characters effect other character's traits. There is no need to restrict 
the number of primary traits: just make the display of attributes capable 
of handling dynamic numbers.

There are some limitations: how would abstract an experience system for all 
cases? How would you be able to represent all the possible variations of 
rules and randomisers in XML? How would you represent rules that depend 
highly upon 'soft' calls such as GM decisions?

> If you convert it to byte/machine code, it would still be under the open
> content license. If you link it in to your software code, you probably
> (and according to the FSF) make them into one work, and get licenses
> conflicts.

Not sure about machine code, but byte code does not fall under the same 
restrictions. For example, there are many byte code Java APIs distributed 
under the GPL that work with the Sun copyright protected native API.

>>Perhaps it might be an idea to start an FRPGC project on exactly such a
>>library? Any volunteers?
> 
> I am already working on a library for Generics. It might be interesting to
> see if it could be made, uh, more generic...

It would be an interesting project. Anyone else with any ideas?

> PS One of the wiki links on the homepage does not work.

Oh dear, which one?

-- 
Ricardo Gladwell
President, Free Roleplaying Community
http://www.freeroleplay.org/
president@xxxxxxxxxxxxxxxx
Follow-Ups:

[freeroleplay] Re: Generics
From: Per Inge Mathisen

[freeroleplay] Re: Generics
From: Timothy S. Nelson
References:
- [freeroleplay] Re: Generics
  - From: Per Inge Mathisen
[freeroleplay] Re: Generics

Other related posts:
