Re: How to continue

[Ingo Ruhnke]

David Philippi <address@hidden> writes:

> I've never used jabber, but are you using icq under Linux? I've got
> the problem that using licq I may receive messages without problems,
> but if I sent messages it's more like random wheter they arrive or
> not (I get an ACK package every time).

I am using icq through a jabber-gateway. The problem with random
message loss is a know problem and had happend due to a protocol
change, its fixed in licq-cvs, ickle and the jabber-gateway IIRC.

>> for the basic concept (disable java-script to get to that page without
>> a login).
> Is this an intended feature or a stupid webmaster?

Don't know, but it allows to search through the articles by using
google, which is quite nice.
 
> Using pause quite often to solve the game isn't something I'd
> consider a bug on the other hand. If you still have fun while you
> "cheat", it's ok for me.

Its cheating and its something that you won't find in any real
(commercial) game, therefore I am searching for a workaround.

> How many layers are there actually?

Basically every WorldObj is a layer in itself, but the main layer types are:

 * gfx-map, the bitmap on which the pingus walk (or at least seem to
   walk, in reality they walk on the colmap), there is only a single
   layer of this type

 * pingusholder, thats the layer which contains all the pingus, there
   is only a single layer of this type

 * particlemanager, for particle effects, exploding pingus, rain,
   snow, etc., there is only a single layer of this type


 * wheatergenerators, generators of particles

 * background layers, everything that paints something background
   like, be it a solid color or a background-bitmap

 * entrances, exits, traps, worldobjects etc.

The last three are actually just classes with inherit from WorldObj
and do something in there draw() and update() methods. They don't
provide an interface with which one could access them.
 
> Unfying the collision code could prove a bit hard. A climber for
> example won't collide at places where other pingus do. But I suppose
> that most of the code will be common for each, so it could prove
> usefull to have a common collision implementation which can be
> either overriden for a special action or wrapped with some specific
> code.

Yep, unifing it for all actions is impossible, after all some actions
are there to prevent collision (slider). But a walker and a climber
have very similar collision points.

>> A generic unstuck-method for pingus would also be interesting, so that
>> we don't need to avoid each and every stuckiness, but simply wrap to
>> pingu to the next free pixel.

> This sounds like a very good idea. But I guess really implementing
> those changes is a bit harder then talking about them. ;-)

Implementing something like that is not that difficult, but making
sure that it works in most situations and doesn't cause unpredictable
behauviour is actually the hard part.

-- 
Advent:     http://www.freesoftware.fsf.org/advent/
Feuerkraft: http://www.freesoftware.fsf.org/feuerkraft/
Pingus:     http://pingus.seul.org/           || ICQ:      59461927
Home:       http://pingus.seul.org/~grumbel/  || JabberID: address@hidden