A New Dawn Bug Reports and Feedback

Status
Not open for further replies.
This is just a balance issue, and I would guess that Afforess would already know this, but the AI has an abnormally large number of spearmen from bronze working up until gunpowder. This makes mounted units almost useless except as advanced scouts.

Well, from a RL POV spearmen were nearly useless against horse unless they were very well disciplined and stayed in ranks. Granted, any military unit of the pre-rifle era that formed and could keep ranks was formidable on the battlefield, but the catch has always been that forming and keeping ranks is ninety percent training and ten percent experience. Same with pike and phalanx.

But Civ IV is horrible at dealing with these exceptions. In fact, I can think of dozens if not hundreds of other exceptions that should be but aren't in place. Horse archers, for instance, should be practically immune to everything but archers, faster horse mounted units, and their own hubris and excitement for battle. That's the reason why the Mongols were so effective. Of course, just having a killer unit isn't enough, as the fall of the Mongol empire showed, and keeping that high level of training and ability around is just as hard in its own way as keeping the training to form and hold ranks.

Only running up against infantry with strong enough shields to stop their arrows, which the infantry are well-enough trained to use (not an easy thing to accomplish) would mean either getting in close (and getting hurt in the process) or having almost no affect.
 
Game crashing bug 1.73E, using default install (I chose custom but didn't actually change any checkboxes) Installed AND, then patch D, then patch E on top.

I did use a modified bonusinfo and featureinfo xml but that's all.

IF you change Civs to the Native American's You wont crash but if you change civ's to the Persians (the next civ's turn after NA) you will crash. Obviously selecting any other civ after NA will also cause a crash.

EDIT: I suspect it's your landmark code. If you play the NA's turn, if you move one of their scouts north the game will crash. They are revealing a location of the map I edited in WB, which probably unbeknownst to me had a landmark tied to it. There isn't anything else around that I could think of as the issue. (Why did I edit the map in AI territory? does it matter? - actually I wanted some of the AI to survive the raging barbarians so created some defendable chokepoints which a city could get settled on. In this case it was Japan just to the north of the NA's scout. Would also let me see how things would turn out if some of the AI had some XP choke point GG farms instead of just me knowing how to take advantage of that)

Are landmarks locations pre-determined at map generation? Are their terrain types pre-determined then as well? (such values might change via WB) When a landmark site is revealed do you just show the precreated sign or is it created randomly on the spot based on the terrain that is active at the time?

EDIT: I was right. I entered WB as the NA civ, switch to reveal tile mod, and revealed a 5x5 area which created a forest landmark on top of a (now) mountain tile that I had altered. I did not crash via revealing from WB, only when revealing the landmark of an altered tile from unit movement.
 

Attachments

This is just a balance issue, and I would guess that Afforess would already know this, but the AI has an abnormally large number of spearmen from bronze working up until gunpowder. This makes mounted units almost useless except as advanced scouts.

actually light cavalry should make mincemeat out of spearmen

pikemen on the other hand shut down any horse army except maybe heavy cavalry (get those shock 1-3 too) but by then you should be on your way to gunpowder anyway :)
 
Game crashing bug 1.73E, using default install (I chose custom but didn't actually change any checkboxes) Installed AND, then patch D, then patch E on top.

Spoiler :

I did use a modified bonusinfo and featureinfo xml but that's all.

IF you change Civs to the Native American's You wont crash but if you change civ's to the Persians (the next civ's turn after NA) you will crash. Obviously selecting any other civ after NA will also cause a crash.

EDIT: I suspect it's your landmark code. If you play the NA's turn, if you move one of their scouts north the game will crash. They are revealing a location of the map I edited in WB, which probably unbeknownst to me had a landmark tied to it. There isn't anything else around that I could think of as the issue. (Why did I edit the map in AI territory? does it matter? - actually I wanted some of the AI to survive the raging barbarians so created some defendable chokepoints which a city could get settled on. In this case it was Japan just to the north of the NA's scout. Would also let me see how things would turn out if some of the AI had some XP choke point GG farms instead of just me knowing how to take advantage of that)

Are landmarks locations pre-determined at map generation? Are their terrain types pre-determined then as well? (such values might change via WB) When a landmark site is revealed do you just show the precreated sign or is it created randomly on the spot based on the terrain that is active at the time?

EDIT: I was right. I entered WB as the NA civ, switch to reveal tile mod, and revealed a 5x5 area which created a forest landmark on top of a (now) mountain tile that I had altered. I did not crash via revealing from WB, only when revealing the landmark of an altered tile from unit movement.

Glad you could find the source. I've always thought there should be a localized landmark regen when changing tiles, but I doubt it's really doable. Maybe landmark rebuild after leaving WB? Don't know.

So the thing I wanted to mention is that unless Afforess fixed it since the 1.73 Betas, the Default as seen on the install choices page is not the same as the default checks in custom installation.
 
EDIT: I suspect it's your landmark code. If you play the NA's turn, if you move one of their scouts north the game will crash. They are revealing a location of the map I edited in WB, which probably unbeknownst to me had a landmark tied to it. There isn't anything else around that I could think of as the issue. (Why did I edit the map in AI territory? does it matter? - actually I wanted some of the AI to survive the raging barbarians so created some defendable chokepoints which a city could get settled on. In this case it was Japan just to the north of the NA's scout. Would also let me see how things would turn out if some of the AI had some XP choke point GG farms instead of just me knowing how to take advantage of that)

Are landmarks locations pre-determined at map generation? Are their terrain types pre-determined then as well? (such values might change via WB) When a landmark site is revealed do you just show the precreated sign or is it created randomly on the spot based on the terrain that is active at the time?

EDIT: I was right. I entered WB as the NA civ, switch to reveal tile mod, and revealed a 5x5 area which created a forest landmark on top of a (now) mountain tile that I had altered. I did not crash via revealing from WB, only when revealing the landmark of an altered tile from unit movement.

Yes, landmarks are pre-determined at startup. (It would be very very slow otherwise, it's a lot of calculations). Changing the tiles that contain landmarks will cause CTD's.
I don't even need to look at your saves... I know from my code. Don't cheat and use Personalized Maps.

So the thing I wanted to mention is that unless Afforess fixed it since the 1.73 Betas, the Default as seen on the install choices page is not the same as the default checks in custom installation.
No, it is.
 
Playing now 1.73D game on Carter giant earth map and having a blast :)

Most latest changes are great and game is again very challenging. I would, however, tweak promotion value balancing a little more toward reducing promotions versus number of units ratio.

Only thing i would plead about - name generator for landmarks ... while having unique landmarks is fun seeing 90% of cities named like this - Hnesueti Nvabaue makes my eyes bleed and my brain explode ... only possible solution right now is turning it off altogether :(

And small problem with name generator - if there's landmark and 2 civs place cities near it - both cities would have same name.
That is what I dislike about the name generator , I don't like it naming my cities.
 
Yes, landmarks are pre-determined at startup. (It would be very very slow otherwise, it's a lot of calculations). Changing the tiles that contain landmarks will cause CTD's.
I don't even need to look at your saves... I know from my code. Don't cheat and use Personalized Maps.

What's ironic is that I personalized the map to give the AI advantages in order to make it more challenging for me and evaluate different situations. Different games for different purposes.

This still does not explain why it should crash on the triggering event of a landmark reveal. Obviously there is code that has to execute on a landmark reveal to make the sign visible and/or look up its name. A hash table lookup perhaps based on the tile plot coordinates? Why is tile type a requirement there? I understand tile type at the begging of the process used to construct the name, "such and such Forest", or "Mount Tallrock" etc. But once its a string who cares what the tile type is later. If you are using the tile type as a key in your landmark look up then the exception is probably a "did not find key" type of error for the cases when the tile type changed. Remove that as part of the key and just look up by tile coordinates and the tile type wont matter (even if it label is mismatched due to the tile change, but that the price for altering a tile.) These are very easily handled, if the key doesn't exist don't put up a sign.

What's going to happen to a landmark tile that was terraformed in the late game and then finally "discovered" by some other civ post tile alteration? This is a legitimate means of altering the tile w/out using the WB.
 
The game saves the landmark type as an int, the name, and the adjective as strings. So, for a forest, 3 values are saved:

LANDMARK_FOREST
Hyboria
Forest of Hyboria

When the tile is revealed for your civilization, it looks for the forest icon. If you have removed the forest, a NULL icon is found, and the game crashes.
 
I don't know if this is a ROM bug or AND bug but since I've never encountered this before I start here.

A section of my map contains a sea blocked trade greyed area but there is no enemy ships there, I opened the editor and theres definetly nothing, just my ships and my cities.

Those coastal cities used to be of another civ and from the fog I did see that greyed area much earlier in the game, however, once the fog was cleared and the cities conquered the blocked area failed to disappear.

Anybody knows a way I could manually edit it out? (for now and future events)

Also noticed somebody reported an invisible pirate corvette in the bugs reports section, could this be related?

Thanks
 
Also noticed somebody reported an invisible pirate corvette in the bugs reports section, could this be related?

Must be it. It's a nasty random event. You will need to sweep every tile with your ships to hit it.
 
The Final Five game option is broken on higher difficulty levels. It kicks in too soon such that the human player will be the first to be eliminated as the AI always expands faster at first, thus having more points.

Either the

1) Turns for this game option need to be modifiable OR
2) Final Five should be represented in the new BUG interface in game and can be enabled and disabled at will.

I support the latter option as it would be easier to implement; simply give the user the ability tick an option to enable / disable it, in game.
 
The quoted posting here:

I'm getting a CTD whenever I attempt to host or join a multiplayer game. Only direct IP tested.

Specs:
Win7 Pro x64
q9450
4GB RAM

RoM 2.81 full
AND 1.53
Fresh installs on both tested, as well as reinstall of Civ4 completely.

Game crashes upon MP loading, with errors related to custom leaderheads, starting with wilhelmina, and moving in reverse alphabetical order in the stack.

Note: Single player loads without issue.

AND install was default options (all but sea monsters and deep red interface.)

Any information I neglected to provide?

was made about 3 months ago

and Afforess answered:

"AND 1.53 doesn't work in Multiplayer.

I believe I've corrected the multiplayer issues in my latest beta's, but the official releases are way behind. "

Well: I wanted to host a multiplayer game today and got exactly the same error. We both have a fresh install of RoM 2.91 and AND 1.73E. What can we do about it?
 
The quoted posting here:



was made about 3 months ago

and Afforess answered:

"AND 1.53 doesn't work in Multiplayer.

I believe I've corrected the multiplayer issues in my latest beta's, but the official releases are way behind. "

Well: I wanted to host a multiplayer game today and got exactly the same error. We both have a fresh install of RoM 2.91 and AND 1.73E. What can we do about it?

should not happen.

Did you try to unload the mod and start an MP to see if it's working without ROM ?
Make sure both of your PC are good enough for the map size and number of civ you are planing to play.
 
It happens ... and we played a normal session of BtS just 3 days ago. We didn't even start a big session. We just wanted to know if it works at all. RoM Terra as map, large size and just about 5 AIs ... we really only wanted to test it as we had quite a few problems with mods and modmods. Wild Mana (Fall from Heaven 2) is almost identical, with the difference that it crashes without any error message ... but it happens at exactly the same point.
And like I said: We played BTS just a few days ago and for about 5 hours ... no problems or crashes at all.
 
AngelGabriel,

I did find the cause of this error, it seems to be related to files in your CustomAssets loading. Even if you never added anything to CustomAssets, some files seem to be generated there by the game. Go to your "My Games/Beyond The Sword" folder and delete the CustomAssets folder on all computers. Then try Multiplayer Again.
 
A couple of problems I have encountered with 1.73E - caravels don't get many xp when being attacked by the dread pirates (max I have seen is 0.5). Ships set on auto explore seem to avoid all unexplored areas unless there is a bit of land exposed.
 
Status
Not open for further replies.
Back
Top Bottom