Buildable Bridges

[Stan`]

21 minutes ago, wowgetoffyourcellphone said:

Pathfinder has to update the path grid all the time for new buildings {blocked paths} and depleted resources {new paths}.

Yep but mountain path don't change and make it bug anyways

12 minutes ago, Juli51 said:

Freedom to build is the key of the game, but limit that freedom with reasonable concepts like expensive to build, deep or mud waters unavailability or tie them to paths from one city center to another.

I agree, my concern is also that you need some kind of modulable art structure to make modular bridges. So looking at the current messy wall placement code, I can't help but wonder how the one for bridges would be. That's why I propose three size, big medium and small bridges.

[wowgetoffyourcellphone]

4 hours ago, stanislas69 said:

Yep but mountain path don't change and make it bug anyways

4 hours ago, Juli51 said:

I would say bridge can only start and end on a shore.

[Stan`]

Which makes it a tiny path on both sides of it. 

[wowgetoffyourcellphone]

Just now, stanislas69 said:

Which makes it a tiny path on both sides of it. 

I don't know what you mean.

[Stan`]

Ah missed the point of your last comment.

[wowgetoffyourcellphone]

Just now, stanislas69 said:

Ah missed the point of your last comment.

I can imagination see it in my head how it all work, and it is quite simple and intuitive, but it's hard to put it into words.

Like a dock, the end of a bridge can only be place on a shoreline. As you stretch the length of it, like place a wall, it will preview what you are doing. Then the other end under your cursor will "snap" to the next shoreline, like how walls "snap" to existing wall tower, but you can wave it back and forth along the shore to place it where you want to. I think there would be a max "angle" of maybe something like 20 degrees across the water course. Maybe test that out if implemented for best max angle. I think bridge would have a max length and a min length. When the wall placed, the foundations show up in the watter and the units start building the bridge from foundation to foundation from the starting end of the bridge until completed. Maybe 2 bridge, 1 stone and 1 wooden, wooden is destructible if you right click or alt-click one of the pylons of the bridge; clicking on the span surface does nothing but move units to that spot.

[Stan`]

I totally understand how you guys meant it. The issue is the code here, the closest thing we have to that in the wall placement, and it's bad. Plus imagine the issue of having scalable foundations, since units can't walk on water AFAIK, that's why I said the best would be to have only three sides, and fixed points, for simplicity sake.

[wowgetoffyourcellphone]

7 minutes ago, stanislas69 said:

I totally understand how you guys meant it. The issue is the code here, the closest thing we have to that in the wall placement, and it's bad. Plus imagine the issue of having scalable foundations, since units can't walk on water AFAIK, that's why I said the best would be to have only three sides, and fixed points, for simplicity sake.

How is the wall place code bad? Can it be fixed?

[elexis]

Should be possible to implement your approach, with or without prior wall placement refactoring, mostly lacks manpower though (or womanpower for that matter).

[Stan`]

My point was simply to have something basic at the beginning, not some kind of steam machine with mesh blocks getting together to make a bridge to see if it works, and then see what we could improve from that.

[Libervurto]

I don't know how the code works but it is probably easier to make modular bridges look good than walls.  Bridges can be made of many lengths by having a choice of different length end-pieces, the middle consisting always of multiples of the same piece.

/MM\
/TMT\
/MMM\
/TMMT\
/MMMM\
/TMMMT\

Edited December 28, 2016 by Libervurto

[Stan`]

The problem I have with that is that it doesn't allow for more than two variation of the bridge. Also it makes the artist make a number of pieces, and those pieces have to be exactly at the same height.

[Stan`]

Todays IRC conversation

 

11:55 < elexis> FeXoR: you wrote the wall placement code right? do you think we can feasibly extend to work with bridges?
11:58 < FeXoR> elexis: I wrote it for rmgen, yes. And yous it can be extended for bridges. To do this propperly in rmgen we would need terrain analysis there (for slope, passability and water coverage). This should not be a new implementation/duplication of e.g. the ingame dock placement so those functions should be shared.
11:59 < elexis> no rgmen, ingame wall-placement
12:00 < FeXoR> The actual state of the rmgen wall builder is...well, messy, since the templates don't contain the information we'd need (or rather the concepts of ingame wall placement and rmgen are different so they are incompatible including the values in the templates).
12:02 < FeXoR> The ingame wall placement was written by voetsjoeba, at least he calls himself that on youtube
12:03 < elexis> perhaps we could just claim its a simple ticket if we tell that its only one file to change (for ingame bridge placement) dx
12:04 < FeXoR> However, if you plan to add this cosider shared functions for placement checks that can than also be used by random maps to avoid code duplication and further incompatibilities
12:04 < FeXoR> elexis: It is not simpe at all...
12:05 < FeXoR> Tough one could take the dock placement for inspiration ;)
12:07 < FeXoR> However it's also related to wall placement to make bridges have arbitrary length. And since we don't have walkable actors (AFAIK) it's also related to terrain deformation which we clearly did not want to have ingame so it's not possible without also implementing walkable actors. So no, it's not simple ;)
12:07 < FeXoR> (Also includes a pathfinder rewrite if done correctly
12:08 -!- wraitii [~Adium@foe37-1-88-183-130-79.fbx.proxad.net] has joined #0ad-dev
12:08 < FeXoR> (For multiple planes - other than the terrain surface - to be checked for paths)
12:09 < FeXoR> I'd like briges both to be easier to place in Atlas as well as in random maps and ingame. However, if we do it we should do it sane ;)
12:15 < elexis> it would have the same mechanic as wall placement
12:16 < elexis> start a bridge at the shoreline, extend it like walls, finish it to the shoreline
12:16 < elexis> *at
12:16 < wraitii> I feel like bridges would be better if they were between set "can build abridge here" points to avoid lame-ing
12:17 < elexis> hf letting rmgen decide which points are lame
12:20 < wraitii> eh, doesn't seem that hard to me, but whatev'
12:20 < wraitii> whoever implements the feature will decide as usual
12:23 < elexis> shoreline to shoreline, max-length, min-length, max-angle
12:24 < elexis> an issue might be the terrain changing while placing a preview, or units (f.e. ships) moving in the way while in the placement preview
12:25 < elexis> or we allow placing partial bridges, just like partial walls
12:25  * Philip recommends not doing anything that makes pathfinding harder than it already is :-)
12:27 -!- sbirmi [~sbirmi@117.212.91.11] has quit [Quit: Leaving.]
12:34 < elexis> (I guess thats about the unit motion rewrite)
12:35 -!- sbirmi [~sbirmi@117.212.91.11] has joined #0ad-dev
12:35 < wraitii> nah, the unitmotion rewrite is making pathfinding easier :P
12:37 < Philip> I meant bridges, since multiple movement planes makes everything hard, and dynamically changing passability is annoying
12:37 -!- sbirmi [~sbirmi@117.212.91.11] has quit [Quit: Leaving.]
12:37 < wraitii> btw re-unitmotion rewrite I'm taking all reviews gladly, I'm kind of not going to do much until it's accepted (or scrapped entirely)
12:38 < wraitii> in the meantime I'll probably focus on reviewing other stuff
12:39 < Imarok> So nobody should review your unitmotion, until you're through the rq ;P
12:40 < wraitii> er, the phabricator RQ :P
12:41 < Imarok> ^^
12:41 < elexis> why multilayers? the pathfinding grid could just be updated to the position of the bridge
12:41 < Imarok> Shall I port the trac rq to phabricator for you? ;P
12:45 < Yves`> elexis: if something can pass below the bridge, you have multiple planes. If the bridge just changes water to land, then you only have one plane.
12:46 < elexis> nonono, nothing can pass there
12:46 < elexis> only units on top of the bridge
12:50 < elexis> wraitii: going through your campaign patch, looks like a good start
12:50 < wraitii> there's not necessariliy multiple planes if it's another passability class that can go below the bridge
  
  15:17 < Stan`> Why does everyone want complex bridges, while pre defined neutral structures with two angles 0 and 90° would be perfect... Also I also believe that when you make a map you put things here and there for a reason
15:17 < Stan`> not for anyone to hack in there by putting a dumb 37.5° bridge in some random places
15:19 < wraitii> that's the FOSS effect
15:19 < wraitii> everybody wants everything in every way
15:19 < wraitii> and it never makes any sense after the second post
15:19 < wraitii> Stan`: performance wise it should be the same, it's just another way to do mostly the same thing
15:20 < Stan`> I also do believe in forbidding docks in those kind of maps. That would make the land unit fishing
15:20 < Stan`> wraitii, Yep, but gameplay wise it's dumb
15:20 < Stan`> also, artistically speaking it's lame
15:20 < Stan`> I'd rather have thre our four really nice bridges
15:20 < wraitii> making games is difficult, as it turns out
15:20 < Stan`> than one ugly one, that can adapt
15:21 < Stan`> no need for another wallPlacement.js clone
15:21 < Stan`> for bridges either
15:22 < Stan`> FeXoR is having enough trouble on that
15:22 < Stan`> Land fishing ticket #1437
15:22 < WildfireBot> #1437 (Decide on land unit fishing) – http://trac.wildfiregames.com/ticket/1437

 

[wowgetoffyourcellphone]

On 12/28/2016 at 5:47 AM, stanislas69 said:

Also it makes the artist make a number of pieces, and those pieces have to be exactly at the same height.

Pieces height can be calculate relative to water plane height.