Re: [Adonthell-devel] Schedules

[Alexandre Courbot]

> The gametime class keeps track of the (gametime) minutes that passed
> since the start of the game. It also allows to specify how much cycles
> make one minute at runtime. So we can for example slow down gametime
> while the player is examining the inventory or stats or during
> dialogues. We should also be able to pause the clock.

Mmm, not sure about whether slowing the game time during dialogues and
such would be pertinent. The characters behind would still continue to
move, do you want to slow them too? Also, think about multiplayer...
time can't run at a different pace for other players. I guess you want
to slow down the game for realism (that is, so you can't spend a whole
day talking to a person). Well, maybe it won't be very realistic, but
it's the best we can do in that respect. And it's already nice! :)
Actully it depends on how long a day would be. I'm in favour for making
them quite long (a few hours).

> Then there would be a date class that keeps track of the total amount of
> gametime spent in the game. That means it would be able to save and load
>
> The date class further provides a number of utility functions to turn
> the gametime minutes into hours, days and weekdays, test whether it is
> day or night and maybe other stuff. Each time one of those functions is
> called, it will update itself from the gametime class. That way, the
> gametime class will remain independant from other stuff. Things like a
> time_event or the schedule manager would make use of the date class
> instead.
> 
> Does that sound okay?

Fine by me.

> Oh and another thing. Since all that stuff is closely related to the new
> mapengine, I would add it to the 0.4 branch, instead of making a new
> branch. There's no use in adding it to 0.3, but it will need a mapengine
> to test the schedules.

It's a matter of (a few) weeks before 0.4 characters are scriptable. You
won't have path-finding and such immediatly, but I guess it would be
sufficient for testing purpose.

I'm currently re-thinking about the collision tests (a pain, especially
with the pixel-based tests) and about having a perfect, fast rendering.
I guess once the engine basics are running, even if the rendering isn't
perfect yet, you'd still be able to use it.

Alex.
-- 
http://www.gnurou.org