Jump to content

elexis

WFG Retired
  • Posts

    3.644
  • Joined

  • Days Won

    59

Everything posted by elexis

  1. Perhaps we can buff rams if turtling is such an issue :P
  2. Above is added to the XML file, taken from Gallic Fields (3).xml. Below is the triggerscript JS file:
  3. Yes, all players need to have the same bugfix, otherwise it will be out-of-sync. But we can fix it for the next release. wiki/GettingStartedProgrammers
  4. This garrisons all towers to 60% capacity with kush Citizen Soldier Infantry units for player 5. Taken from elephantine_trigger_scripts.json. One can do it without a trigger script too, for example Sicilia Nomad has this: But you can see it's quite uncomfortable - one has to place the entities manually and find the entity IDs of these units. See Gallic Fields for example to see how triggerscript files can be loaded.
  5. I thought I fixed this one - it's the Archipelago map? It's a good starter task anyhow.
  6. The first error occurs if there are multiple soundcarts or muted soundcarts IIRC. The second one is reported and can happen on any map.
  7. We had not reduced it further than archer range (72m) because we didn't want people to surround it with 80 archers and get 0 damage while taking out the thing 1HP at a time. Next you're gonna say walls could get 100% pierce armor then, maybe, yes. To be honest I don't see the problem with the current balancing (other than spamming wall turrets without walls as seen on the Dizaka screenshot, which can be addressed by having non-arrow turrets depending on the density, but even that can be torn down with siege engines easily if the enemy doesnt have an sustained economy). That's already the case that they have to be manned before they shoot arrows.
  8. At least all user accounts that logged in within the last 8 months should be restored. IPv6 gamelist issue was solved (hosting without STUN). Ratings were fixed (Dunedans gamereport patch was applied) So it seems all jumpscares of the server upgrade passed.
  9. It's a misrepresentation. We should be guided by good arguments for things being true or false, not our feelings. It's not that there is a choice between optimizing the game for the top 1% or the 99%, it's that the best players have the most experience, seen the most things to get a clear view of the game mechanics and balance to understand the needs of every player, experienced and inexperienced. It's in everyones interest that games don't end up with one unit or structure being spammed and the rest being useless. If there is a thing where experienced players don't complain about but inexperienced ones do, it seems to be a matter of missing information, such as was intended to be provided by tutorials. That. In that case I might call players that what you mentioned above if they request us to change the game without having read the current attack and armor stats of the units they use. If one had a problem with destroying rams, one looks at the armor statistics of rams and notices they have 99% pierce armor or such but only very few hack and pierce armor. Next, one looks at ones own units and checks which units have more hack and crush attack. One should learn by mistakes, not request the game to change if one does a mistake. The counters to units is not a fixed list and can change with every tempalte chane. Swordsmen are one example, elephants are also good, even spear cavalry or outnumbering catapults. While one can add a number of tooltips to these units for rams, one runs into the same problem with the next best unit that one doesn't know how to counter. The missing piece of information might or might not be how our attack model actually works (pierce / hack / crush damage and armor model). It can be easy to lose a match against an AI or a human who wins by outnumbering one and then instead blames the unit that was outnumbering one. It really depends on the individual match why a player has lost. I have seen a lot of players have a terrible economy, don't train 5 units at a time but 1 by 1, don't research economic upgrades or phases, build storehouses far away from the resources and thus might not be able to reach age 3 as quick as the human or AI opponent to counter the siege warfare effectively. So just because one is bombarded with lots of different balancing requests doesn't mean they are all right (in particular if some of the requests are mutually exclusive). I didn't say that the game balance shouldn't be changed, I was merely answering to andy5995 why I don't want to review a patch that changes the balancing while we don't have unambiguous evidence of an 'OP' unit. While movement speed might not change balancing that much, attack and armor stats are very volatile. I don't feel capable to judge a balancing patch without investing much time with other players to have something that works for 100% of the players. I don't need someone to change a number, I need someone credible that knows that this number is the best one and actually exhausted the evidence for that (at least if you're working on a unit that is central to the gameplay, such as siege engines). Why one lost can best be found by watching the replay. If the opponent outnumbers one and one didn't have the hard-counters to that unit, the opponent clearly should win. While not everyone knows about the snowballing effect of Lanchesters Law, I assume everyone knows that one can't beat someone with 50 units if one has only 30 units of the same type. If the numbers were equal but the unit types differed, it should also be clear that the units have properties that make them stronger or weaker. Where would they be seen other than the tooltip? I'm losing many games, but it's because I'm not good enough at it, not because the game is imbalanced. Consequence? I continue to play with the better players because I want to be challenged. If I want to increase my chance to win, I have to spend more time on practise and theory or I can play with more players closer to my level or the AI. Anyway, the discussion is getting more broad with every sentence rather than converging towards an implementation. In case I didn't mention it already in this post, ok for me to restrict rams attacking other units (maybe except siege engines).
  10. Also I agree that movement speed of rams should be slightly slower than that of melee infantry units, just don't mess with the effectiveness of the thing once it attacks.
  11. 1v1s are over before it comes to the time of walls. 4v4s (with bots or not), it takes a lot more time until everyone is defeated if the players are similarly capable. It's important to note that walls can accumulate successively while the unit population is limited and the area where progressively is obstructed in lategame. From #3811 and https://wildfiregames.com/forum/index.php?/topic/20621-good-players-usually-say-no-walls-in-the-multiplayer-lobby/&do=findComment&comment=314820
  12. New players on the forums yes, on the lobby Ive never seen a single complaint. If it's player error, it should be addressed by informing the player better. For defending against rams, we may also consider giving more civs easier access to sword units (or new upgrades or whatever) It's refering to the historic reason why they were buffed (it's a big part of the announcement and even in the trailer...). Siege engines were inherently incapable of breaking well fortified players and then games could not conclude. Now it's the better player (or AI) being able to destroy the opponent, fortified or not. One requirement of siege balancing is to not have walls become useless nor overpowered, but just right. Walls are quite effective if they are used correctly. It's just that a big army is more useful first and then the walls come after that, economically. What do you need a wall for if the player can just walk around it. If you make walling too cheap, then it will be too easy to fortify and possibly result in a stalemate. If players want to have walls before they have an economy to afford them, they usually do contracts to first build a base for some time (if they didn't discover the ceasefire setting yet). Possibly. (Note gates are incentivized already to be attacked, as they have less HP.) I wonder if it isn't the AI spamming rams that people complain about, rather than multiplayer. The AI does economic cheating, so it might be that this is one source of the issue.
  13. Preventing 100 lines of complaint each day for half a year is worth a lot of additional testing time. So I agree that the question is whether there is gross imbalance. It takes a number of matches, not minutes to test the balance, i.e. more hours. Sometimes balancing issues are unambiguous, but I have seen many balancing discussions about concrete balancing patches on code.wildfiregames.com (not the forums) where one gets 15 opinions by 10 players, so increasing the pool to 100 players might give you 200 opinions. So unless it's unambiguous, one might have to work with forensic evidence or trust. Players who are the most competitive ones in the lobby have the best qualification to judge, it was them fixing the grossly imbalanced units in the last alphas. I'm not even trusting my own judgement in most cases. Being able to download and install mods easily is extremely powerful and can help if there is a grossly imbalanced unit. Unless everyone is using the same balancing mod however, one doesn't have the freedom to chose the lobby games, not the freedom to chose with whom to play (which is one of the most decisive factors as to why people play currently). My impression was that it was just right, maybe 15% off in the attack speed, but not so much more. You have to consider that they have to be able to break multiple layers of walls, not only some houses and a CC. In particular their attack speed is comparable to elephants no? So isn't it rather a problem of missing swordsmen of the defenders? They melt within few strikes if one hasn't neglected that the enemy will have to train rams. (What does seem overpowered are massed ranged siege engines - one can't get even get close to them with a huge army before everything died off. Rams seem like the best unit to counter in my experience. But the problem of massed units is universal, so not sure what one could do there.) Anyhow, I put my imaginary money on the players better than me to judge that
  14. Balancing is very delicate, I don't want to burn my fingers with that unless there is powerful evidence for the balancing not resulting in 10m high waves of complaints. It's easy to type a different number, but it takes the best players of the lobby to determine if it should be 10% greater or smaller.
  15. Sure it was a rated game? Check the commands.txt file of the replay for "RatingEnabled".
  16. On nomad, The most decisive factor is the CC placement. Resources: I can't tell you how many games I lost because the enemy could place the CC immediately, had all four resources and could boom quickly while other players had to run around for a minute to get 3 of the resources and then is stuck for 5 minutes somewhere because of the missing fourth resource. Then you're rushed and it's over. One often places the CC in very close proximity to the best opponent. This first match intereaction can already have settled defeat, it's not even funny. (Don't get me started on possible solutions to this, in particular since it differs per map) Startung units placement is entirely random as (-_-) mentioned. Consequently the most important playerdecision after the CC placement is scouting the player distribution and developing and communicating a matchstrategy with the team as soon as possible to address the particularities of the player distribution. For example by having the safer player do economy boom and a "sling" (sending over resources to the player on the frontline). Unusual player placement can be interesting to the ones who expect it: On Unknown Nomad, there was/is a river map too. It also yielded 5v3 distributions. Sometimes there is one player over the other side of the rest of the team. It is clear that this player is going to be attacked 3v1 soon. But this can be turned to an advantage. This player must receive additional resources from the other side of the river, should send a woman via a fishing ship possibly and have his alllies build a CC on the other side as soon as possible. If this player receives a lot of resources and allies support from ships, his mere territory presence will impede the enemy economy (gathering and trade). It can be turned into a wedge splitting the enemy in half. On River Archipelago without nomad, it's 50% chance that it's going to be "left vs. right" and 50% chance of "top vs bottom" when there are two teams - but the islands are always in north-south direction. So sometimes it ends up as a 2 * 2v2 on land, sometimes it's a 4v4 naval. We also like to play a Corsica vs. Sardinia 2v2v2v2 sometimes. It means there is a 2v2 on each of the two islands and the winners then fight against each other, like a little tourney. It's planned to add a player placement setting to the gamesetup however. So one could chose whether it's grouped by team or not, or specifically opposing.
  17. In addition to what Vladislav said, there had been a number of catches to get the simulation deterministic. For example the Math.random function was replaced in ScriptInterface::ReplaceNondeterministicRNG by a deterministic boost RNG that is initialized with a seed that the host choses at gamestart. Some other Math prototype functions were replaced in globalscripts/Math.js to be platform-consistent. https://trac.wildfiregames.com/wiki/SimulationRequirements#Determinism When you create a new simulation component, you should be more worried that it is serialization-safe, otherwise there will be an out-of-sync on rejoin or savegame load. The most important part is to have the simulation not change it's code through GUI calls as you mentioned. Sometimes one does have to do that (optional range visualization or diplomacy colors). Then one has to pay extra attention that the GUI dependent variables don't change the simulation state.
  18. The player positions on random maps are sorted per team and I cannot reproduce this issue. Please provide a replay where they aren't. The other issues, other than the ram balance, are confirmed to be well known issues that will most likely remain unsolved for the next years.
  19. STUN working now after Jan reloaded the firewall
  20. and with that information I could reproduce the error, thanks! I've reported it here in the hope that our AI developers return: #5282. Should also mention that rejoining with AI players is currently not supported.
  21. Not all user accounts made it to the other side yet The STUN port(s?) need forwarding
  22. Congratulations on the release! The announcement is both informing and entertaining and makes one interested in playing this new version. The nomad screenshot is flawless. Some feedback on the screenshotst: and for the feature presentation: (not that these are important defects, rather ways to spend time on the computer)
×
×
  • Create New...