Complete.Org: Mailing Lists: Archives: freeciv-dev: December 2001:
[Freeciv-Dev] Re: Development Strategies [Was Documentation, Usability a
Home

[Freeciv-Dev] Re: Development Strategies [Was Documentation, Usability a

[Top] [All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
To: Jules Bean <jules@xxxxxxxxxxxxxxx>
Cc: gregor@xxxxxxxxxxxxx, freeciv development list <freeciv-dev@xxxxxxxxxxx>
Subject: [Freeciv-Dev] Re: Development Strategies [Was Documentation, Usability and Development]
From: Daniel L Speyer <dspeyer@xxxxxxxxxxx>
Date: Thu, 6 Dec 2001 11:21:28 -0500 (EST)

On Thu, 6 Dec 2001, Jules Bean wrote:

> On Thu, Dec 06, 2001 at 11:16:06AM +0100, Gregor Zeitlinger wrote:
> > On Thu, 6 Dec 2001, Raimar Falke wrote:
> > > Than add an extra step between them:
> > > 1.5) collect features units can have
> 
> > as I said it's impossible unless it's a very generic description - a
> > scripting language :)
> 
> That's pessimistic.  It may be impossible to (no, it *will* be
> impossible) to cover everything a designer could imagine. But it would 
> be possible to cover a much wider range of features than currently
> supported, for quite an interesting system, without needing full
> scripting.

Do you really think we can anticipate more than a very small fraction of
the ideas potential-ruleset-makers may come up with?  

Building rulesets *should* be fairly easy -- and we should have a lot of
people doing it -- but it's boring if all you can do is recombine a short
list of features.  Now, we could write a longer list, but this means only
those with the skills and familiarity to write in the freeciv core will be
able to create interesting rulesets.  This is a major developement-model
error for a game with so many somewhat technical fans.

Furthermore, if new unit powers are written in C, then they can't be used
multi-player until everybody (roughly including
civserver.freeciv.org) upgrades.  This is a gross violation of
"release early, release often".

And why should new unit powers be written in C?  The language is slower to
write in and much more prone to bugs (no array checking, have to do
pointers manually...).  Now C does have its performance advantages, but
how often will this code be called (especially if the default units don't
use anything interpereted)?

Now, I suppose you could argue that central checking is nessesary to
ensure that unitscripts don't hang -- though this runs counter to the
general approach and philosophy of OSS -- but it would be possible to
write technical safeguards.  Recursion limits are common (255 levels is a
typical max), and iteration limits are possible (most loops can give a max
number of needed times).  Plus, if we have good built-in list-handling
(i.e. foreach, length, nth, firstWhich...) very little explicite looping
would be needed.

--Daniel Speyer
"May the /src be with you, always"


> 
> Jules
> 
> 



[Prev in Thread] Current Thread [Next in Thread]