FeXoR

Sometimes the map still contains unreachable trees:

THX! I think I will start writing the bot next week. As far as I get it and questions: - First each turn the data changed in the engine the turn before are applied to the BasiAI's data - BaseAI.OnUpdate() is the call for the actual custom AI script that should be overwritten in the constructor. - In opposite to the data update at the beginning of the turn (engine -> BaseAI) setting orders is send individually? Or did I miss a function sending a diff of orders from the BaseAI to the engine? And if it's like think doesn't it defeat the ability of the AI to be threaded (because of many interactions with the engine)? - What does 'settings' in the constructor contain?

Testmap for Spahbod: fortress_test.zip

I'm at it though I have to agree it's not that easy...

Agreed. But keep in mind even if the civil soldiers have arms equipped they can die and the resources they are carrying will be lost. So returning them before fighting would be good and after fighting send them back to whatever they where doing without having to figure out what they where doing would be even better. I guess that's a part of what Quantumstate meant.

Confirmed

Since there have been no objections I'll start editing http://trac.wildfiregames.com/wiki/Rmgen_Library

Those units might get overpowered, but we'll see. IMO both are already overpowered and if they are the only non-siege units able to attack buildings they'll get even better in comparison to other units...

There are some units that can raze walls well like war elephants. Most heroes with hack attack only do fine as well.

- Rotating buildings in-game: Fixed. After selecting a bunch of units that can build and clicked a build button the building is attached to the cursor. When holding down the left mouse button the buildings location is fixed and and it can be rotated with the mouse. If you right-click in any part of this action the placement is canceled. - Wall placement: It's on it's way: - Textures: Yes, I think some textures are not fitting well. Don't know what the art department thinks about that though. BTW: Real ground/stones can be quite shiny white: http://mkeamy.typepa...6327c970c-800wi http://www.ithakabou...Cliff-Beach.jpg http://images.volcan...milos_g6685.jpg http://static.design...fs-of-dover.jpg http://images.world6..._do_galleryfull - Water shore wave animation/Ship idle animation: AFAIK both in discussion (not sure though) - Civilizations: AFAIK 6 civilizations are planed for 0AD Part 1, those already in the game. For part 2 others are planned. - Waste land: There are many stuff available for decoration. The usage depends mainly on the designer map designer. Feel free to design maps with tons of decorative stuff - Neutral faction: There is one: Gaia. She attacks everyone else (wild animals are an example). If you start a scenario with a player unassigned all it's units/structures are given to her and they'll attack everyone coming near. I don't know how far and when an AI is in discussion for her. - Town bell: Hot discussions about this are burning here. - Farms: Auto rebuild is in discussion. No idea concerning the model. Further info in the next post by wraitii. - Fish: See the next post by wraitii. - Treasures: Treasure are available too. The units as treasure are not planed as far as I know but 'capturing'/converting units and/or buildings is planned AFAIK. - Outposts: There are some buildings that can be placed anywhere on the map (not on cliffs/water ofc.). The outpost/watchtower (mainly to get visual on the surroundings) and the roman army camp for defensive/siege purpose. Further info in the next post by wraitii. - Carrying stone: Visuals, visuals, visuals... I don't get it. Well, I'd say the artists/animators are heavily on it doing a great job. Adding different animations for everything is time consuming. But if you can inspire them... - Siege units: I totally agree. Don't know if it's planned though. - Day/night cycle: If it's to fast it's annoying, if it's to slow it's a bit unreal as you said. I don't think it's worth it.

Thx already! Though I don't know what "settings" is for right now but it works without errors. It seams that communication without importing the API is not that easy (because I already need it to avoid the error). I will read more code to get used to the common-api-v2 and look forward to some explanation of yours. Do I get it right that in the AI API rotation is again like in the engine but height isn't really needed? ^^ However, I look forward for some explanation and thx for your patience with me.

I don't think a wall without any attack has much of a worth in game. It might not be realistic but the costs of a wall should be low to match the actual advantage in game. The time to build a structure and the time to repair a structure should indeed be the same I think. In fact it should depend on the materials needed to build an entity IMO so the build time would be something like (0.7*foodCost + 0.9*woodCost + 1.2*stoneCost + 1.5*mettalCost)/2 seconds. Same with the maximum hitpoints (structural points/life). It could be something like (1.3*foodCost + 0.7*woodCost + 1.5*stoneCost + 0.5*metalCost)/2. Same with the armor. It could be (0.1*foodCost + 0.5*woodCost + 0.9*stoneCost + 1.9*metalCost)/10 and distributed with another function as seen fit for this unit type. This would allow to balance things easily but I think I run into great opposition with that... In this case palisades would indeed be build faster. However, the advantage of walls in the age series didn't match the worth. I never build one because a fortress had a much better price/effect ratio and stone was the most limited resource (at least in AoE II - AoK/AoQ).

What do I need to write a new AI from scratch? I have: data.json { "name": "FeXoR Bot", "description": "A simple AI by FeXoR.", "constructor": "fbotAI" } fbot.js function fbotAI() { } When starting a game in Atlas it claims: "ERROR: AI script Serialize call failed" "ERROR: AI script Deserialize call failed" It's the same when I add: _init.js Engine.IncludeModule("common-api-v2"); So what do I need in addition to avoid errors? In addition I would like to know how the connection to the engine works. It would be nice if someone would explain it to me first without any API usage (so just what the engine does) and then how the common-api-v2 achieves communication with the engine. I know this is very basic and so might not be that easy to explain but I'm not very good at reading and understanding code especially when something appears like "Engine.xyz" and "Engine" is not defined in the API, sorry for that Thx in advance

Doesn't the AI get all entities it owns at the start including all it's props (like position, placement angle etc.)?

I have a professional microphone but that has to be amplified before plugging into a PC. I only got a guitar amplifier and the frequencies to adjust doesn't fit voice frequencies. I bought an amplifier construction kit that was really cheap (~15 $/12 €) but has very good stats (indeed the price of a ready to use amplifier greatly depends on the casing and its style, not the electronics). A microphone of cause is not that cheep (Shure offers decent microphones for 100 € /130 $ for example). The according environment is an other issue. Since I don't think my English pronunciation wouldn't pass any kind of quality test I'm out for English tutorials ...but later in game development I could do the German ones.

You could do that with an abstract alignment saved in the AI. I did something similar to this in the rmgen wall_builder lib in getWallAlignment. In my case I needed it to determine the center (center of mass) of a given fortress to place it right. It simply does the same as a place function would do but instead of placing each wall element it puts everything needed for placement into a data structure and returns it. If you'd save such an alignment you could check if a part of the wall is missing and rebuild it correctly. Of cause the AI API have to support rotation for building placement...

Added the additional warning and it's indeed the problem: interestinglog.html mainlog.html Still in atlas there are no warnings but no unit is moved by qbot and he don't seam to produce anything as well.

I did with SVN revision 11547. Still the no resource warnings tested on new_rms_test (game and atlas): ...and the Sicilia errors: In Atlas the Sicilia scenario seams to run fine though. There where no custom content accessible to the game while I tested and both where tested with default settings (but reveal map and small size for new_rms_test). I have Alpha 9 installed as well but I don't think it interacts with the SVN version and I made sure I user the SVN version ofc .

What a button like this should do really depends on many other things suggested here. I like this idea: ...but only for citizen soldiers actually gathering. The radius that would make sense however is harder to determine. I should include all gathering citizen soldiers nearer to the civil center the button was clicked than to any other gathering building able to garrison. When they have returned there resources they carried they should guard the civil center in a current distance depending on the number of units reacting to the alarm so other units can still pass through them but they are still tight enough packed to have decent fighting efficiency. Non-attack citizens should garrison in the nearest building that can be garrisoned and has an attack at the same time so you don't need to click 2 buttons. Some things might collide/contradict with this behavior though like stances since citizen soldiers when returning to secure there carried resources might break the command when attacked. I already said in other topics that this is a very bad idea in general but it might be OK if an order priority system is implemented.

While were at it there's another issue. I don't think I need to add another topic for it though. If there's no stone left on the map QBot or the common-api-v2 warns that no stone is found on the map. Since that is not unnatural to appear I think a bot should be able to deal with that even if there's no stone or other resource present at the beginning. Another thing is that QBot spams warnings on the scenario sicilia: http://fexor.dyndns....a_walrnings.jpg

I want to add the wall_builder.js rmgen lib documentation to the wiki. While reading it again I found some things that could be improved but I don't want to change things before others agree especially the original authors. I think that are mainly Spahbod and historic_bruno is that right? Map Coordinate System I think it would be better to tell it uses 2D placement (x/y plane including the default rotation) and the third coordinate is set automatically (cannot be set). When given to the game engine it is converted to 3D with the x/z plane (with opposite rotation) as the map's surface. The later thing doesn't matter to random map scripting at all but is just an information for rmgen script designers that may also work on the game engine with different conventions for example. Placing terrain and units Before explaining area placement and unit group placement it would be good to note that there are simple placement methods for terrain and units (setTexture and placeObject). There might be other things but I think I'll ask before I change things. When I add the wall_builder.js documentation I'll put it after the unit group placement description I think. Any thoughts about that?

If the starting walls are a civ bonus then all the random maps would need that to to be consistent. I could add the walls to the default starting entity placement function createStartingPlayerEntities for iberians if that is wanted. I think some other changes like cut some trees around the start locations would be needed then too though. BTW Mythos_Ruler, I send you a map in a PM that had this functionality. Guess it's outdated now though.

Thx much for explaining I think those issues should be fixed soon. I already send a false patch to be reviewed because I couldn't see the false behavior in Atlas. I will try to remember to restart Atlas ofc. but it's not my work practice to close and reopen programs over and over again. At least it's an annoyance. Since i had problems with broken patches too I think I will work as follows: - Modify files in cache (if possible for all my desired changes) to be able to see the results in Atlas without reopening it until I'm done. - SVN update - Copy modified files to the SVN tree - Rename cache folder - Test - Generate patch - Revert SVN and delete all added files - Apply patch to see if everything works fine - Test - Send patch to be reviewed Will that likely lead to problems or will it work fine (until the real custom content folders are implemented ofc.)?

What you are saying is similar to what Philip said in Ticket #631 and your previous answer agreeing with Ykkrosh in this post. Atlas Tricks (Page seams a little hidden for me) says %appdata%\0ad\cache\mods\public\maps\scenarios would be the default directory for saving maps in Atlas but it changed to %localappdata% as I stated in this post. The above is true for Alpha 9 (not sure about %localappdata% or %appdata%) but for the SVN version the default save path is \binaries\data\mods\public\map\scenarios in the SVN directory. Atlas and the game itself search for files in %localappdata% as well though (SVN version). Ticket #1145 sais it would be fixed but I cannot find out where in My Documents the files have to be placed (and so doubt that it works at all). It provides a link to this list and some other useful related stuff. Now I'm quite confused and don't understand: - Why you say %appdata%/0ad\cache\mods\public\maps is not the custom map folder though it behaves exactly like I would think a custom map folder would? - Why was it changed from %localappdata% to %appdata% (indicating it can be changed) but not to My Documents/0ad/maps as suggest in this list? - In witch way the virtual file system is broken and how you 'lost control' of the paths like said in this post? - Why is Ticket #1145 set to fixed though it doesn't appear to be fixed? BTW: It works for AI as well: %appdata%\0ad\cache\mods\public\simulation\ai

Random maps are loaded from different paths in Atlas and the game itself. - The sub path '\binaries\data\mods\public\maps\random' in the svn directory - The custom content path '%localappdata%/0ad\cache\mods\public\maps\random' (for windows I'm working on) If I have atlas opened and edit a file in the custom map folder and then generate the map (again) the changes are applied. If I do the same in the svn path the changes are NOT applied. In some cases it raises an javascript error stating a syntax error in the last line of the file (no matter if something near the last line was modified). Atlas needs to be reloaded to see the changes.That should be the same for the game itself I think. I don't know if this is a known issue or how it works with Alpha 9. I'm using Vista (longhorn, home premium), tortoiseSVN and the current SVN version.