Ships and movement/boarding

[LDiCesare]

My problem is that units can only be loaded onto carriers in ADJACENT squares. My order system cannot handle that. Nor do I think it is a good idea. The implication is of a 100km gangplank.
On the other hand, if we reject the adjacent square approach, we have to deal with the concept of ships in land squares. This gets back to my original arguments for rivers and coastlines to be in land squares (and beaches and ports also). This is the wrong thread to extend this discussion - but at least I have explained the reason for the delay.

Well, Gary, here is the Good thread
I can understand that adjacent ship and armies is poor, but is it possible to say that when army is given orders to go to a water square, it follows the path needed to get there, then realises it cannot cross water and stops, unless there are ships in the square, in which case it boards the ships instead of moving? It is kind of if (terrain not allowed) then (try to board something there). I am not sure it is user friendly, but it can allow for rendez-vous between land army and transport easily.

Also, a point in allowing an army to move even if water is somewhere along the path is when you move through black tiles. (Actually this has very little in common but I think as I type or the other way round): I'd like to be able to give orders to go through hidden terrain, then all the terrain should be considered neutral from a pathfinding point of view. I don't think it is now, which
1) prevents me from moving 10 squares west into the unknown if that square is a water square
2) allows me to artificially know the land/sea map by trial and error on the pathfinding algo.

[Gary Thomas]

Also, a point in allowing an army to move even if water is somewhere along the path is when you move through black tiles. (Actually this has very little in common but I think as I type or the other way round): I'd like to be able to give orders to go through hidden terrain, then all the terrain should be considered neutral from a pathfinding point of view. I don't think it is now, which
1) prevents me from moving 10 squares west into the unknown if that square is a water square
2) allows me to artificially know the land/sea map by trial and error on the pathfinding algo.

I am aware of this problem. The reason it remains is that at the moment hidden map data is kept only for the human player. The path calculation you suggest should also apply to the game AI. So it is waiting on me devising some way of non-player civilizations knowing about unknown squares.

Cheers

[Gary Thomas]

I really do not like the adjacent square idea. What I am working toward is having the MovementData object that you introduced to units also record the last square the unit was in.

Then a couple of things become possible:

1. A ship can enter a land square provided the last square it was in was a legal sea square. From that square it can move to the previous square, or to certain adjacent ones. This rule can be worked out to allow movement along edge rivers. However, when a ship is in a land square, it is viewed as being on the edge of the square, not the middle.

2. It would make retreat easy to calculate - I personally do not think a unit should ever retreat more than one square in a tick - even that is 100km in three days.

Cheers

Gary

[LDiCesare]

That seems a good approach. I actually had a last square in the beginnings of the code (early d5) so I can't say it's a bad idea.
It makes for one glitch however, that is if a ship wants to follow a coast, they have to back in-sea before coming back in-land. You may have trouble pathfinding around that. We could allow ships to move to land squares that are coast squares contiguous to the current coast. Configs like this can be messy:
SSS
SLL
SLL
If you come from the middle S square at the left, then go in the middle L, you should be able to go to any square but the lower-right one. This requires finding Sea squares around adjacent land squares, and seeing if these are adjacent to the original sea square...

[Gary Thomas]

It shouldn't be too difficult to work out all the possible combinations and produce a method to say whether a particular move is legal or not. We will (I think) also need to record the side of the square a ship is on, and the side it was previously on. For example:

+-----+-----+wwwww+
| 1   |  2  w  3  w
|     |     w     w
+wwwww+wwwww+wwwww+
|  4  |  5  |  6  |
|     |     |     |
+-----+-----+-----+

A ship is in the sea square 3, and wants to move up the river between the land squares 2 and 5. On its first move, it moves onto the coast of 2. Its last square is 3, and its last edge is null, its new edge is East. Next move it stays in 2. Its last square is 2 and its last edge is East. Its new edge is South. On its next move it moves up the river to 1, edge South. It could equally have crossed the river to 4, edge North.

I don't think sorting these out will be too difficult.

Cheers

[LDiCesare]

I can see you are subreptitiously putting rivers between squares here. A question: How do you know a riverine unit is in square 2 rather than 5 in your drawing? Geographically, it is in-between. Do we say 2 has a river on the edge with 5, so that it owns the relevant art and units, or can units be in both squares?
Anyway,
I agree that with such a model I need a last edge.
I am not sure about the rivers-between-square stuff however. Did anyone agree with that? I would tend to give in if the river belongs to a square and borders another one, thus being on the edge of the square. It would make it all the easier since I'd have to record either a last-square or last-river in the MovementData.

[Richard Bruns]

Originally posted by Gary Thomas
A ship is in the sea square 3, and wants to move up the river between the land squares 2 and 5. On its first move, it moves onto the coast of 2. Its last square is 3, and its last edge is null, its new edge is East. Next move it stays in 2. Its last square is 2 and its last edge is East. Its new edge is South. On its next move it moves up the river to 1, edge South. It could equally have crossed the river to 4, edge North.

I don't think sorting these out will be too difficult.

I think that sorting these out will be very difficult, at least from the player's point of view. I'm already confused.

How will a player know which edge something is on? What will they think when they order movement and the ship stays in the same square.

I vote for staying with the ship boarding that Civ 2 used. It was easy to deal with.

[Mark_Everson]

Hi Everyone:

AFAIK we never did reach agreement on whether rivers are in-square or between-squares in Design decision - Rivers - In square or on edge of square?. There are points for each approach, as you can read in the thread. If we get to the point where either art or coding needs to go forward, and its still a tie, then whoever has to do the most work to implement it gets the deciding vote in my book. I think by a slim margin that would be the artist.

[Gary Thomas]

For the sake of getting something going, I will accept the Civ way of doing things. The ship cannot move along rivers. The ship stays at sea (normally) and the land unit moves onto it. If there is a city in a land square, the ship can move onto the land square and is considered to be in port. The land units then get an extra option on the orders menu, allowing them to board the ship. If a ship attempts to move onto a non-port land square, the player is asked if they want to unlod, and which units to unload.

In passing, Civ 3 has rivers on the edges of squares.

Cheers

[Mark_Everson]

No ships on rivers? I really like ships on rivers, especially for pre-modern small-draft shipd. Without ships on rivers, modeling the Vikings getting down to the Byzantine empire is just about impossible to do. I'm sure there's an even better example but I'm drawing a blank right now.

For in-square rivers you just handle ships like land units, but that can only move along rivers. Boarding would be by special order just like when the ship's in port. My proposal for on-edge rivers would be that both banks of the river get a pointer to the ship. You of course draw the ship on the river. For boarding you can board from either bank as just proposed for a ship on the coast, but with a special order to distinguish from units that want to cross the river without boarding the ship. When the ship moves you update both pointers. Ships are only allowed on edge centers, not on 'corners'. Seems simple to me... why ditch it?

[LDiCesare]

I suggest we allow ships to go from a sea square to a land square and let boarding happen there. It has to do with coastal squares, and makes some sense: It will make it easier to have naval bombardments in the same square. Until recently, you could not bombard from 100km away. I'd rather let the rivers thing sleep until we know where those damn rivers should go (please nobody say "down to the sea"). Getting units board ships and cross seas already has a great potential for fun and bugs in itself.
Although art may be a problem, can't rivers be seen as another layer on top of the squares? Or additional polygons? They would hold their own unit info, they would be rendered after the terrain as an overlay, and have pointers to the square(s) they border, and the rivers and seas they connect to. Although visually you still have to decide if they are in the middle or on the edge of squares, the code and rendering would be pretty independent. I know that it may be hard from a graphics point of view (because of trees and such), but if we paint 1-terrain 2-rivers 3-trees/cities, it may work?

[Gary Thomas]

As I have said elsewhere, the only real argument for in-square rivers is that you can treat them like roads. The big disadvantages are that they then interfere with roads, they cannot be used as boundaries, and they have no effect on combat.

I have never believed that there is any difficulty in placing rivers, and the ships on them, including their picture, on the edge of squares. How this is handled internally is a coding issue and not relevant to the forum discussion. My main point here is that it can be done. If the code is more complicated, that is not an issue for general discussion. The ship will be deemed to be where it appears to be - in the middle of a sea square, or between two land squares (on the river). Deciding where the ship can go next is a problem for the path algorithm (me, in effect) and I have no problem with it.

Laurent's point about moving ships onto a land square is one that I have no problem with either, as long as the ship knows where it came from. Otherwise ships will be able to cross an isthmus (who would then need the Panama Canal?) or shortcut across a cape.

However, if ships can move onto coast squares it immediately leads to the question, if a ship is on a land square, can it move directly to an adjacent land square off the same sea, as opposed to moving to sea, moving at sea, then moving back to land? If the answer is that it can move directly, then, in effect, we have arrived at the situation I have been discussing in previous posts (up to my most recent, when I tried to give up). Everyone seemed to think that that approach was too complicated. Or something.

Cheers

[Mark_Everson]

Originally posted by Gary Thomas
However, if ships can move onto coast squares it immediately leads to the question, if a ship is on a land square, can it move directly to an adjacent land square off the same sea, as opposed to moving to sea, moving at sea, then moving back to land? If the answer is that it can move directly, then, in effect, we have arrived at the situation I have been discussing in previous posts (up to my most recent, when I tried to give up). Everyone seemed to think that that approach was too complicated. Or something.

Everyone? AFAIK Richard was the only one that expressed the view that it was complicated. And if I may be so cheeky, its just because your discussion was overly complicated. Discussing boats being either on the north edge of this or the south edge of that. All players need to know is that if you could sail a ship in that way in real life, then you can do it in Clash. Troops can board from any square that shares a common edge with the river segment the boat is in.

[HugoHillbilly]

I have a question about rivers, is the current plan to allow units to cross rivers ala Civ-series? I never liked this idea, as rivers are good geographical borders in ancient/middle eras. I believe that rivers shouldn't be crossable by units unless there is a boat to ferry them across or a bridge has been built across it.

BTW I just found out about this project now. Im at school so as soon as I get home im going to download the demo and boot it up. From what ive been reading this looks like an amazing project. I really look forward to trying it out.

[Mark_Everson]

Hi HugoHillbilly, yeah I'd like to see rivers have more effects too. Some rivers though are easily crossable. We'll just have to decide how much detail we want in that area.

Looking forward to hearing your opinions!

-Mark