Arrakis.py map script discussion

[Ahriman]

The idea is that Graben will only ever appear at Sink level and Rock and Rugged with only ever appear at Flatland level. Bear this in mind when testing.

Ok. In this case then we should probably tone the sink-graben back down to 1h1c.

So:
saltpan: 0, terraforms to graben
graben(sink) = 1h1c
rock(flat) = 1h
rugged(flat) = 1h, 25% defense, 125% build cost.
mesa = 1f1h, 25% defense.

Can this be installed over the previous one, or should it go over the base 1.5?

 

[davidlallen]

Can this be installed over the previous one, or should it go over the base 1.5?

There are xml files in the first download which are not present in the second, so my assumption is that you *must* install the first download and then the second. The second by itself, probably won't work.

 

[Deliverator]

Sorry not to be clearer - yes these files are to be placed over the ones in the first terrain patch.

I can put it together into a new patch so that it is easier to understand.

 

[davidlallen]

Sorry not to be clearer - yes these files are to be placed over the ones in the first terrain patch.

I can put it together into a new patch so that it is easier to understand.

I believe I extracted all your files. The second patch is a little confusing because there is no directory structure, or at least not enough of one. It is obvious enough what should happen, for modders. I put the xml file into assets/xml/art and the textures directory into assets/art/terrain. After I do this, I get one undefined art file: art/terrain/textures/blend/iceoceangrids.dds, referred from civ4artdefines_terrain.xml at line 36. There is a file art/terrain/textures/iceoceangrids.dds (notice the missing "blends" directory level) in the 1.5 fpk file, so I have copied that in. However, this could lead to that crash we used to have, when switching on grid line display.

If anybody has this crash, please start up some *other* mod, turn off grid line display, then go back to DW and it should work.

Please note I used civcheck to find this problem and the missing sound defines, which I highly recommend before releasing stuff.

 

[koma13]

I hope Cephalo's changes will be included too.

 

[Deliverator]

Thanks for sorting through the mess david, I'll do a proper patch next time.

The question is are you any happier with the textures? I'm going to do some further work - I think Rugged needs improving.

 

[davidlallen]

The question is are you any happier with the textures? I'm going to do some further work - I think Rugged needs improving.

I still find it too monochromatic. In particular there is a huge play difference beween graben and rock, and I find I need to "hunt the shadows" to trace where graben is. See attached screenshot. The arrow at the right points to a transition from graben to rock, which is very subtle. There is a shade difference of brown, but that is not enough to quickly make a visual distinction. I feel like I need to mouseover to be sure, and it makes it too hard to plan where my cities should go.

Maybe if you used more of a yellow for desert/coast, then you could use a paler brown for graben?

Also, I am curious about the one "coastal" mesa at the bottom, with the arrow. It seems to have a whole layer of sand around it, which the other "coastal" mesa at the left does not have. It is very distinctive, but why is it the only one like that?

I hope Cephalo's changes will be included too.

Do you mean the AI changes, or the mapscript change, or both? This requires a whole discussion of its own. On archipelago, the AI changes are nice but may not matter too much. I have no strong objection to including it, it's just merging work. (I have some other local sdk fixes for other problems which have been reported.)

However, his current mapscript is a pangaea. This will play totally differently and we have no experience in balancing it. What do you think we should do with this? If I make a 1.5.1 patch this coming weekend, my inclination is to leave out the mapscript. We need a lot more discussion and playtesting before switching to pangaea.

 

[cephalo]

However, his current mapscript is a pangaea. This will play totally differently and we have no experience in balancing it. What do you think we should do with this? If I make a 1.5.1 patch this coming weekend, my inclination is to leave out the mapscript. We need a lot more discussion and playtesting before switching to pangaea.

I will be making improvements to the map script very soon, but I would like to have other people besides myself test it out on pangaea to see what comes up. I'm thinking that if the mod works on pangaea and archipelago, it should work on everything in between as well.

 

[Deliverator]

I'll make rock and graben more distinctive for round 3.

I realised the Layer Orders were muddled - that was causing some weird issues.

I don't know the cause of the strange Mesa-on-Ocean/Coast tiles, perhaps it is a consequence of my mapscript hacking. In your screenshot the far left arrow is pointing Mesa on Coast, and the middle one to Mesa on Ocean I believe hence the lighter colour. I may be able to fix this with the Layer Orders.

 

[davidlallen]

In your screenshot the far left arrow is pointing Mesa on Coast, and the middle one to Mesa on Ocean I believe hence the lighter colour. I may be able to fix this with the Layer Orders.

I didn't save the game. But, how can there be mesa next to ocean, without coast in the way?

Do you think it is possible to make ocean more yellow, and then use the current ocean pale brown as graben?

 

[Deliverator]

I was wrong - it was Mesa on top of Dunes. I'm trying to make it so that all elevated terrain uses the Mesa texture, but I haven't quite got it working yet.

We can try and make the sand a bit more yellowy - might look bad though. I have some other ideas to try first.

 

[koma13]

Originally Posted by koma13:
I hope Cephalo's changes will be included too.

Do you mean the AI changes, or the mapscript change, or both?

I would say both. I played a game with cephalo changes yesterday and it had much more action going on. Duke Leto "The Righteous" attacked me, Baron Harkonnen, with a SoD! Imagine that.

 

[davidlallen]

This is getting out of control. There are suddenly 16 different play experiences; pick one AI, one mapscript and one terrain set from the following:

AI: original or cephalo

mapscript: original archipelago or original arrakis or deliverator archipelago or cephalo arrakis

terrain: original or deliverator

We need to focus on one or two and get them working.

 

[Deliverator]

Don't panic. I don't think it's out of control. Merging in cephalo's AI seems the way to go. Both he and koma are reporting positive experiences.

The purpose of my terrain changes is to give cephalo the toolkit he needs to make a more 'realistic' Arrakis.py. I'll be putting up another version within the hour. I don't mind if these don't make the patch, because we need to iron out any issues with bonuses, improvements, promotions, etc first.

 

[koma13]

I think we can ignore the orignal play experiences, leaving us with two possibilities (deliverator archipelago or cephalo arrakis).

 

[davidlallen]

I think we can ignore the orignal play experiences, leaving us with two possibilities (deliverator archipelago or cephalo arrakis).

Cephalo arrakis doesn't work correctly at all with deliverator terrain. That can be fixed. But the pangaea play balance will be quite different. This can turn into a release discussion. We have a small stack of graphical and play balance improvements, which could go into a "production" 1.5.1. The new terrain by itself needs tuning for bonuses and improvements, and adding a new style of mapscript on top of that means we would definitely want some "experimental" patch.

It is already a little inconvenient to merge deliverator's improvement/unit changes into my local changes, and it will get worse fast. We need something to serve as a sync point; but it will be too unbalanced for a general release.

Shall I just cut a 1.5.1 with all the changes made so far? Most of the map combinations will be unplayable, or at least unplaytested. It would have to be experimental, so that our 500+ recent downloaders (!) will not all get stuck into unplayable combinations.

 

[Deliverator]

Right I'm done tweaking so I'm going to assemble another terrain patch. Hopefully, graben is distinct from rock now.

An experimental patch sounds good to me - gives us a baseline to work off.

 

[davidlallen]

(redirect to incremental patch thread)

 

[Deliverator]

Terrain Patch V3

Spoiler :

I have assembled it as a full patch over 1.5, although all that has changed since the first patch is the contents of the art/terrain/textures subfolders and CIV4ArtDefines_Terrain.xml. If you are merging this in, you can just clear out the contents of art/terrain/textures and put the new folders in.

The textures have been further refined. All the issues with funny edges, blending, etc have been fixed between the XML and the graphics apart from a tiny one at edge of polar coast.

I've included my amended mapscript renamed Dunipelago.py. This is intended to be a stop gap to playtest the new terrain until Cephalo's has finished the next Arrakis.py. Dunipelago should allow us to weed out issues - both gameplay and cosmetic - with the new terrain system.

Download Link

Known issue - You'll noticed some of the bonuses morph to the shape of the heightmap and some don't. This is fixable.

 

[cephalo]

I have a question. Graben and salt are the only terrains that are supposed to occur in the pans correct? Are all the pans supposed to start as salt or do we want graben also?