Jump to content

FeXoR

WFG Retired
  • Posts

    1.426
  • Joined

  • Last visited

  • Days Won

    28

Posts posted by FeXoR

  1. IMO the cost of technologies has to be balanced against the cost of units and buildings.

    If technologies are to expensive and/or take to long games will lean towards "reach the unit cap ASAP and overrun the enemy".

    (Instead of upgrading to City Phase you could alternatively build 40 skirmishers for example)

    The cost of upgrading everything should cost about as much as building 2x unit cap cheap combat units (as a rough guideline so about 2*250*100 ~= 50000 resources)

    The main thing driving research is the satisfaction of basic needs so Food should be the resource needed most in the research tree.

    (That's why the neolithic revolution had such a big impact on social and technological development)

    • Like 1
  2. I really think those suggestions are best put into the game (if wanted) as techs.

    In the end "gods" are psychological/sociological ideas that got widely spread and so influenced people.

    Since this is still the case adding specific gods will likely generate controversy about this matter so I would avoid them.

    Since we dropped "exclusive or techs" (AFAIK entirely, so you can upgrade all techs, not have to decide between two, I like this very much) re-adding that by choosing a "god" would IMO also go against the path of the general game design we seam to lean towards.

    I congratulate anyone seeing no problems with specific gods but I believe many would consider this blasphemy.

  3. Hi juanjo, Welcome.

    This game has the aim of being historically accurate.

    Since gods are only mentioned in stories rather than physically appearing they had an impact on society (and that's in the game in form of temples and priests) but never physically appeared and/or directly intervened.

    So they will not get part of the original game.

    However, you are free to build a mod including gods.

    • Like 2
  4. EDIT: I was at r16253 when this all happened (so maybe fixed)!

    Not sure if those are related.

    I'm at r16256 and build a storehouse on uneven terrain. It seamed to be completely build and the unit AI send the builders to harvest wood that in fact could be delivered to the storehouse in question.

    However, the storehouse had 1 hitpoint and no upgrades showed up when selecting it.

    I attacked an enemy CC that was not finished yet but my javlin units dealt no damage to it. Same when attacking a defense tower.

    I placed a CC with 10 workers and they started constructing. They where attacked and killed so I send ~50 javelin infantry to kill the enemy units around and send them to finish the CC. However, even after several insane move actions of them they never started to actually build it any further.

    I'm not sure what is causing all this but IMO ATM the game is in an unplayable state.

    The unit AI is entirely broken:

    - Collectors don't divide to collect from other resources of the same type right next to the resource targeted and, if many unit are selected, block each other so nothing gets done.

    - Collectors stop at storehouses while resources of the same type as send to collect are just 2 meters away (so for me for no reason at all). This happened with the storehouse above.

    - If many units are selected and send to an incomplete building they never start building it (see the CC above).

    - If many units are send to build several buildings in a queue (e.g. houses) some go to the next building (likely because blocked by other builders) while some just seam to drop the entire queue (maybe because the couldn't reach any foundation).

    - The unit AI doesn't recognize if a target isn't reachable in all cases (like harvesting from a tree within a forest - it just should send the unit to the reachable resource of that type closest to the targeted resource IMO).

    • Like 2
  5. Sorry my fault. opening atlas i recognized that the female workers and the other units are placed on the map by default. So opening up a random map worked.

    But thank you for your time.

    Yes, scenarios and skirmish maps use starting entities placed by the map designer.

    Random maps (in most cases) use the starting entities as defined in the civ .jason.

  6. :pardon: Strange, I was suspecting A17 would be the problem as that's when we switched to SDL2 on Windows. It sounds like an SDL bug, but I can't think of anything relevant we've changed since A17. Did you have a Windows or driver update or something? Maybe try A17 again to confirm.

    Likely you are right and I just noticed it later (most of the time I'm working with Atlas - playing is quite lagy on my laptop).

    Sorry for the late answer and the confusion :sorry:

  7. hi, that's not implemented. The difficulty with walls is in the positioning and checking of obstructions.

    But if you want to look at it, you can take inspiration from the function findDefensiveLocation (also in headquarter.js) which find the best location for towers and fortresses. It is called from the constructionPlan (in queueplan-building.js) when asked to build a tower.

    ...yet another argument for shared libraries for simulation and RMGen IMO.

    Make walls buildable through trees. This can remove 75% of obstructions. my opinion.

    I agree. An RMGen patch for this is in progress - maybe that could help for the in-game wall placement as well as for AI wall placement, not sure.

  8. Have played in other maps with them I dont know if it random generated or what but I had the walls already closed with gates and everything, is better this way I think, in other map got just the tower of the walls and had the same problem couldnt close, had to delete and rebuild

    On random maps you either get closed walls or defense towers (not wall towers)

    Defense towers have higher range and grand 5 population limit for Iberians (So they are not really weaker than walls).

    Defense towers can't be connected by walls (at least not in-game, so that's not a bug).

    If you get walls or towers (or something else) is the decision of the random map designer.

    If you get defense towers "only" there might be not enough space for walls on the map (e.g. tiny/small maps) or you start on some kind of island (e.g. the "Snowflake Searocks" map).

    As far as I can see there is always a good reason if you don't get walls as Iberian civ bonus on random maps.

    Not sure about skirmish or scenario maps (I mentioned that the walls might cause some trouble back when skirmish maps where added but didn't look into that lately).

    An issue could be the compatibility of random map wall placement and in-game wall placement.

    If a wall part is destroyed you might not be able to rebuild it in-game (not sure).

    (That is one reason - besides code duplication - that I asked for shared functions)

  9. Yes, I'm quite sure the way most archers before the time heavy armors got widely spread (not sure when exactly) used this method.

    I would assume though that more organized groups like (like roman legions, not sure if they had archers) and civilizations used to heavier armors may have switched to a more "static" archer behavior and instead of focusing on fast/agile archery focused more on long range/full draw archery and covering the archers by melee troops if the enemy gets to close.

    This is just what I'd assume though.

    • Like 1
  10. Very nice job (y)

    This might also get to be a part of the (AFAIK planned) detailed civilization description.

    If clicking a unit/building a more depth stat frame would show up e.g. below the techtree.

    (Though I'd prefer a table with all units/buildings)

    That'd be a bit more involved. Mostly because we don't even support scroll bars for objects, but just for lists (and some similar elements). I'll see if I can take a closer look sometime.

    IMO scrollbars should be available for all GUI objects that could make use of it, yes. :acute:

    That could enable us to get rid of several (though likely not all) cases of "things not fitting on the screen" and would be much more sane then several other proposals I've seen so far.

    • Like 1
  11. IMO it's not up to a game to change society and we should focus on a sane system do determine the players gameplay capabilities rather than his social skills.

    To get rid of occurrences where PPL feel insulted by others a mute function (don't show a specific players messages including map pings) should be enough (this could be per game or permanently) but should not include any change of which games are shown (in the end it's a disadvantage for the player that feels insulted because he has less games to choose from while it should be a disadvantage for the insulting player).

    Additionally what happens if an unwanted player enters the game the insulted person is in?

    (The insulted person could be kicked but that's a disadvantage to the wrong person again.

    or the unwanted person could not see any game a person is in that used the function but this can/will be "misused" so that's even worse.)

    Insulting, however, should IMO not have any impact on the gameplay.

    Voting systems don't work since PPL playing one style tend to blame PPL playing another style independent of the play styles efficiency (similar for talking style).

    I'm not sure if banning is the way to go though for now (with sane moderators) it seams like a valid solution though this will get time consuming the more PPL play this game (and likely the sanity of moderators will vanish with more moderators required to handle this).

    What PPL like or don't like is in general not objective so we should not focus to much on this IMO.

    PPL should learn to get along with other PPL and just masking them will not get the PPL together but divide them further IMO.

    (And if a AIs start taunting I'd like to have the mute function also work for AIs!)

  12. At Alpha 17 it did work fine, Could you give me any hint which revisions I should look for (between A17 and r15998) and which revision A17 was?

    It might be something with the resolution. My screen resolution is 1280x800 (and maybe the game resolution is set to 1280x768 for some reason and the monitor doesn't switch correctly or something).

    So how can I manually set the resolution and how to tell what resolution the game is running in?

    EDIT:

    Adding to local.cfg:

    xres = "1280"

    yres = "768"

    fixes the issue (ofc. it's still not my monitors resolution!).

    So it might be a specific problem with my monitors resolution (1280x800 at least that's what my drivers tell me as well as several full screen applications).

    Minimal resolution also works fine (1024x768).

    Sometimes with 1280x800 the bottom part of the mainscreen is "wrapped" to the otherwise unused space at the top of the screen (e.g. the report buttons) and the "wrapped" part then flickers.

    In the other cases with the same resolution, however, I can't see any missing parts (the report buttons are present at the bottom of the screen).

    I am not able to deterministically enforce one of those two behaviors.

  13. In the latest SVN the actual point of interaction of the cursor is shifted to the bottom by a fixed space compared to the visual curser. I also have black, unused space at the top of the screen that may be the same space. This happens in the main menus as well as in the game (where also moving the cursor to the bottom of the screen does not make the view scroll downwards as intended).

    Sadly the screenshots neither contain the black space nor the cursor to show the issue.

    Information:

    Revision: 15998

    Operating System: Windows Vista 64 Home (Longhorn) fully updated

    System Info: system_info.txt

  14. I have an unprotected version of this I could give you. The only issue might be that I asked for edit permission for some models specifically for this mod so thehiveworkshop's rules might be violated unless the creators are asked again for general permission and grant it. And this will be rather time-consuming since the models are quite many and I can't remember what was by whom.

    OK, thanks anyways.

    What about the terrain and the tree models? AFAIK they are from 0 A.D. mainly, not sure.

  15. If this would be the case then he couldn't host I guess.

    Well, you could host and your mate could join and you could start a game but at the moment you build a wall (and the difference in calculation on both systems happen) an out of sync (OOS) error would occur. I don't think it's this, though, since I think an OOS error would be logged and it isn't.

  16. Prodigal Son: I understood that.

    Since this is a balancing branch thread (with the aim of maximizing playability for the current implementation - if I got that right) your solution is OK here.

    What I wanted to say is that IMO it is much more productive to improve the basic game system instead of fiddling around with the current, incomplete implementation that should be replaced in the end anyways (and with it your suggestion).

    But within this thread you're entirely right (and I was stupid to post here in the first place), so go on ;)

    • Like 1
  17. Though I don't really like attack priorities due to the wide range of armor values (if units should be able to attack buildings at all) I have to say in this case it would help.

    (However, attack priority should come late in the unit AI's decision queue which target to attack at least after "target in range" (for ranged units))

    AFAIK in the end units should not be able to attack buildings at all but capture them.

    So this will not be an issue at all in the end (again, if I'm informed correctly).

    Is there any place that sums up the targeted gameplay (http://trac.wildfiregames.com/wiki/GameplayFeatureStatus does not provide much information about this)?

  18. [...]

    Ranged Cavalry/Chariot automicro could be gone, and unlocked with a mid/late game tech adding back their minimum range as an intermediate solution.

    Minimum range will never be a solution to anything IMO.

    As stated several times before:

    - The minimum range of a ranged weapon is quite small (1-1.5 m) so historical accuracy is not a valid argument for min. range

    - Min. range causes problems like the chase forever scenario (with equally fast units)

    (Combined with formations and attack priorities several other issues arise and still will with them redone)

    So IMO only siege units should have minimum range (from the historical point of view due to their fixed projectile launching angle).

    The so called "automicro" is a byproduct of the simple unit AI and the minimum range. If the Unit AI would be better it would be an advantage for the melee units (as it should be).

×
×
  • Create New...