KillerClowns
Emperor
- Joined
- Oct 6, 2007
- Messages
- 1,139
Ah, yes... the Chaos III spell, Wonder, has a chance of summoing permanent Treants (e.g. Ents) and Krakens. Tebryn starts with Chaos Mana. Some Archmage must have been having some fun.
FfH2 0.33 Bug Thread
- Thread starter Kael
- Start date
Ah, yes... the Chaos III spell, Wonder, has a chance of summoing permanent Treants (e.g. Ents) and Krakens. Tebryn starts with Chaos Mana. Some Archmage must have been having some fun.
Nor'easter
Deity
- Joined
- Feb 25, 2007
- Messages
- 2,022
Kael said that he's going to fix it in patch "e," you'll see that in the first post in this thread.
Frozen-Vomit
Prince
AI_unitUpdate
In AI_unitUpdate, the return value 1 is supposed to be used to indicate that the move is handled in python, however, this python function doesn't actually handle the move on that layer. It should do the equivalent of:
getGroup()->pushMission(MISSION_SKIP);
To indicate that a particular group should skip its move.
Since it doesn't, bool CvSelectionGroupAI::AI_update() fails its infinite loop check (thus looping 100 times instead of 0 - this may be a time resource drain).
I have the following patch in the C++ code in my next version of Wiser Orcs, but I think it would be better to fix this in python, in case you'd like to actually handle non-skip moves in the future in AI_unitUpdate:
CvUnit::healRate
You added the following code:
This will not work correctly if pPlot->getOwnerINLINE() == NO_PLAYER.
Suggested fix:
CvUnit::changeExtraEnemyHeal, CvUnit::changeExtraNeutralHeal, CvUnit::changeExtraFriendlyHeal
I'd suggest removing the asserts. They will trigger on any negative heal promotion.
CvPlayer::changeImprovementYieldChange
I'd suggest removing the last assert. Agriculture (as an example) will trigger it.
In AI_unitUpdate, the return value 1 is supposed to be used to indicate that the move is handled in python, however, this python function doesn't actually handle the move on that layer. It should do the equivalent of:
getGroup()->pushMission(MISSION_SKIP);
To indicate that a particular group should skip its move.
Since it doesn't, bool CvSelectionGroupAI::AI_update() fails its infinite loop check (thus looping 100 times instead of 0 - this may be a time resource drain).
I have the following patch in the C++ code in my next version of Wiser Orcs, but I think it would be better to fix this in python, in case you'd like to actually handle non-skip moves in the future in AI_unitUpdate:
Code:
bool CvUnitAI::AI_update()
{
PROFILE_FUNC();
CvUnit* pTransportUnit;
FAssertMsg(canMove(), "canMove is expected to be true");
FAssertMsg(isGroupHead(), "isGroupHead is expected to be true"); // XXX is this a good idea???
// allow python to handle it
CyUnit* pyUnit = new CyUnit(this);
CyArgsList argsList;
argsList.add(gDLL->getPythonIFace()->makePythonObject(pyUnit)); // pass in unit class
long lResult=0;
gDLL->getPythonIFace()->callFunction(PYGameModule, "AI_unitUpdate", argsList.makeFunctionArgs(), &lResult);
delete pyUnit; // python fxn must not hold on to this pointer
if (lResult == 1)
{
[B] // Wiser Orcs start - prevent infinite loop.
if (getGroup()->getLengthMissionQueue() < 1)
{
getGroup()->pushMission(MISSION_SKIP); //TODO: check if this prevents spellcasting.
}
// Wiser Orcs end
[/B] return false;
}CvUnit::healRate
You added the following code:
Code:
//FfH Mana Effects: Added by Kael 08/21/2007
iTotalHeal += GET_PLAYER(pPlot->getOwnerINLINE()).getHealChangeEnemy();
//FfH: End AddSuggested fix:
Code:
//FfH Mana Effects: Added by Kael 08/21/2007
[B] // Wiser Orce start - defect fix
if (pPlot->getOwnerINLINE() != NO_PLAYER)
{
[/B] iTotalHeal += GET_PLAYER(pPlot->getOwnerINLINE()).getHealChangeEnemy();
[B] }
// Wiser Orcs end[/B]
//FfH: End AddCvUnit::changeExtraEnemyHeal, CvUnit::changeExtraNeutralHeal, CvUnit::changeExtraFriendlyHeal
I'd suggest removing the asserts. They will trigger on any negative heal promotion.
CvPlayer::changeImprovementYieldChange
I'd suggest removing the last assert. Agriculture (as an example) will trigger it.
Nikis-Knight
Deity
Might not be bugs, but it's weird that centaurs can benefit from nightmares.
Also, Ride of the Nine Kings requires horses and will not work with nightmares.
(Might be intentional balance issues, but seemed odd to me.)
Also, Ride of the Nine Kings requires horses and will not work with nightmares.
(Might be intentional balance issues, but seemed odd to me.)
I know it, but next patch could break save games (and it could appeared at the end of the week), and I have some really fun game now (first time from ages), which I don`t want to break. So I just want to fix it by myself in my game.
tiberion02
Chieftain
- Joined
- Dec 30, 2007
- Messages
- 73
Any idea on why, when winning the race to infernal pact, I get the option to play as Hyborem/Infernal, and when I hit yes, Dis is already founded and has no demonic citizens?
patch c
patch c
Love
Deity
He must have casted wonder every single turn.
xienwolf
Deity
Sorry, I wanted to test the code a bit longer and didn't think Kael's patch was coming up so soon. I am pretty sure he used the code I supplied to "fix" assimilation, but I learned in FF it broke UpgradeCivilization. Attached is the full (so far) working edit.
Spoiler :
Milaga
Prince
- Joined
- Jan 19, 2008
- Messages
- 470
I've attached a save from patch D. In it you can see Ljosalfar Balor (!), Sheiam Wolfriders and Centaur as well as Basium when Hyborem hasn't entered the game yet.
The problem I had with immortal units being reborn when upgraded or sacrificing themselves to found religion was in patch B but I didn't get a chance to test it this game.
I was going for a cultural victory with Cardith Lorda. I noticed a quirk. When a city flips to Kuriotates, you are prompted if you want a settlement or city but the camera doesn't center on the city in question. After you choose you are told a city flips to you. Not a big deal since you can upgrade settlements now.
The problem I had with immortal units being reborn when upgraded or sacrificing themselves to found religion was in patch B but I didn't get a chance to test it this game.
I was going for a cultural victory with Cardith Lorda. I noticed a quirk. When a city flips to Kuriotates, you are prompted if you want a settlement or city but the camera doesn't center on the city in question. After you choose you are told a city flips to you. Not a big deal since you can upgrade settlements now.
0.33D:
just end the turn:
just end the turn:
Code:
Traceback (most recent call last):
File "CvSpellInterface", line 21, in canCast
File "<string>", line 0, in ?
File "CvSpellInterface", line 399, in reqCommanderJoin
NameError: global name 'pPlot' is not defined
ERR: Python function canCast failed, module CvSpellInterfacefire immune units get collateral damage from fireballs and if there are other units in the stack, which are not immune to fire, however, they are picked as defender, in most of the time.
think that is also true for all other non-physical damage.
think that is also true for all other non-physical damage.
westamastaflash
Lord Commander
- Joined
- Nov 1, 2007
- Messages
- 933
It appears that wolf riders and centaurs have no upgrade requirement...
.33 Patch D.
Settings:
Standard Size
Erebus Map
19 Civs
Monarch
Raging Barbarians
Aggressive AI
Permanent Alliances
Require Complete Kills
Barbarian World
Blessing of Amathon
Living World
Conquest, Religious, Tower, and Altar Victories on.
My scout shows that he can upgrade to a Centaur and a Wolf rider right off the bat (once you get horseback riding or start on classical start)
I will try a clean install to ensure there are no changes...
.33 Patch D.
Settings:
Standard Size
Erebus Map
19 Civs
Monarch
Raging Barbarians
Aggressive AI
Permanent Alliances
Require Complete Kills
Barbarian World
Blessing of Amathon
Living World
Conquest, Religious, Tower, and Altar Victories on.
My scout shows that he can upgrade to a Centaur and a Wolf rider right off the bat (once you get horseback riding or start on classical start)
I will try a clean install to ensure there are no changes...
xienwolf
Deity
I haven't tested it yet, but I am uncertain if the fix to Hidden Nationality being forceably removed is accurate.
Rather than:
I think it ought to be
This is because one of the first IF statements in updatePlunder checks "ifEnemy" which the unit will not be if it has already lost Hidden Nationality. So the blockade must be cleared before the Hidden Nationality is adjusted at all.
Also in CvPlayer::killUnits it might be appropriate to use a number much larger than -1 for changeImmortal(-1), as this would leave any unit which is both naturally Immortal AND has the Immortal Promotion quite alive. Negative numbers are not an issue since the changeImmortal function will never allow a full negative, so any larger number ought to suffice quite nicely (I think there were some other places where this was a minor issue, but I never remembered to look into the details for if it was actually an issue).
Rather than:
Code:
m_iHiddenNationality += iNewValue;
if (!isHiddenNationality())
{
updatePlunder(-1, false);
joinGroup(NULL, true);
updatePlunder(1, false);
}I think it ought to be
Code:
updatePlunder(-1, false);
m_iHiddenNationality += iNewValue;
joinGroup(NULL, true);
updatePlunder(1, false);This is because one of the first IF statements in updatePlunder checks "ifEnemy" which the unit will not be if it has already lost Hidden Nationality. So the blockade must be cleared before the Hidden Nationality is adjusted at all.
Also in CvPlayer::killUnits it might be appropriate to use a number much larger than -1 for changeImmortal(-1), as this would leave any unit which is both naturally Immortal AND has the Immortal Promotion quite alive. Negative numbers are not an issue since the changeImmortal function will never allow a full negative, so any larger number ought to suffice quite nicely (I think there were some other places where this was a minor issue, but I never remembered to look into the details for if it was actually an issue).
