Single Player bugs and crashes v38 plus (SVN) - After the 20th of February 2018

[Shoobs]

My first three are usually nomad, organized and then either expansive or seafaring, then get the other of the two. Maintenance reduction I find is incredibly necessary while nomad makes prehistory much easier for hunting units. After that I evaluate my priorities, if I get there.

I just looked directly at creative and its giving me pause on whether I want to use it over expansive after the third trait or still get expansive ( I took seafaring on this map )

I play positive traits and no negatives on positive traits.

 

[amjh]

This save crashes for me, just open it and end turn.

 

[Thunderbrd]

This save crashes for me, just open it and end turn.

Latest SVN right?

I'll take a look when I can but I'm mid-move right now and probably won't have time till next week. Might need to start a new game til then.

 

[amjh]

Latest SVN right?

Only missing the single graphical fix to "Way of Shadows".

 

[raxo2222]

Only missing the single graphical fix to "Way of Shadows".

This may be reason, as there was fix to that fix.

Also I loaded your save and passed one turn, but autosave failed to compress data or something like that.

 

[Dancing Hoskuld]

I feel that it would be easier to balance developing leaders traits if there were many more "levels" on each trait and each level did only one thing. The first level would indeed be a poor version of the final level that trait could reach but each additional "level" only affected one thing. Perhaps, instead of it being a single line progression of levels it could be a branching tree where each branch allows you to progress in one aspect of the overall trait. You would not be able to reach the top/level level without completing all branches. Very similar to the tech tree in fact.

I would like to reserve one or two entries on most trees just to open up some "religious direction" buildings (a.k.a. Sevo's "Faces of God" mod stuff).

Take Productive, for example, and the fact that quarries are given so much more benefit over mines and lumbermills.

This might be a bit too "one thing" but Productive could have many lines one for quarries, one for mines, one for lumber mills and others for other production improvements eg work animal (eg horse) pastures.

Actually "work animal" improvements would be a very good one since you could choose to specialize the ones you have access to (if the tree branched into multiple OR branches).

 

[amjh]

I feel that it would be easier to balance developing leaders traits if there were many more "levels" on each trait and each level did only one thing. The first level would indeed be a poor version of the final level that trait could reach but each additional "level" only affected one thing. Perhaps, instead of it being a single line progression of levels it could be a branching tree where each branch allows you to progress in one aspect of the overall trait. You would not be able to reach the top/level level without completing all branches. Very similar to the tech tree in fact.

Would it then make sense to balance it so lower levels work well in the early game, while the higher levels become relevant in later eras? That way, you wouldn't need to fully balance everything for both.

For example, give flat bonuses early and percentage bonuses later.

 

[BlueGenie]

In Size Matters when merging and then splitting a group of 3 Stone Macemen only one of them retains Poison Tips, despite all of them having it when merging in the first place.
Rev 10041

 

[Thunderbrd]

I feel that it would be easier to balance developing leaders traits if there were many more "levels" on each trait and each level did only one thing. The first level would indeed be a poor version of the final level that trait could reach but each additional "level" only affected one thing. Perhaps, instead of it being a single line progression of levels it could be a branching tree where each branch allows you to progress in one aspect of the overall trait. You would not be able to reach the top/level level without completing all branches. Very similar to the tech tree in fact.

I would like to reserve one or two entries on most trees just to open up some "religious direction" buildings (a.k.a. Sevo's "Faces of God" mod stuff).

This might be a bit too "one thing" but Productive could have many lines one for quarries, one for mines, one for lumber mills and others for other production improvements eg work animal (eg horse) pastures.

Actually "work animal" improvements would be a very good one since you could choose to specialize the ones you have access to (if the tree branched into multiple OR branches).

The design has been done for over a year and waiting on me to find the time to put it to XML. It took nearly a year of side work to get that design done. And it's entirely a mathematically balanced out affair so it's past me working much further on it until testing shows need for tweaks.

If anyone else wants to make a trait set, we can always add another set option.

Some similar thinking has been included as to what you suggest there but I kept the progressions much more simple and straightforward. Some of Amjh's thinking in his last post was also applied.

 

[Thunderbrd]

In Size Matters when merging and then splitting a group of 3 Stone Macemen only one of them retains Poison Tips, despite all of them having it when merging in the first place.
Rev 10041

As designed for all free promotions. Take them back to a city that gives poison tips to get the other two updated. There are reasons for the way it is established. Alternatives to this run into problems.

 

[BlueGenie]

That's a shame. I like to run with a merged group to a position and then split them up for the actual fighting. Biggest reason being upkeep while moving and healing faster as split than grouped after the fighting. Not possible any more with the free promotions, which are often counted upon, being removed from 2 thirds.

 

[hau]

has always been an AI strategy since vanilla. Sometimes they just get all emotional about some diplomatic thing and declare a war without intending to actually even send an invasion force. Sometimes they're nowhere near being ready to launch an attack and being at war triggers them to get ready more but yeah, this isn't really a bug, just some of the funny behavior you get from the AI not being human.

1 and 2 are probably aspects of some things I just fixed. However, if it persists, I'm sure I'll be able to fix it soon. I will continue to do some testing soon and will be watching for feedback after today's commit. If you see it continuing, report it again please.

I have real life too, sorry for the delay.
SVN 1046
1) AI stacks now have lesser dogs but still too many.
1a) AI uses in stacks only siege weapons, atlatlists (javeliners/arsonists) and dogs. AI have macemans, axemans and spearmans, but they been used only for city defend. AI almost no produce swordsmans.
2) Enforcers still not in use in crime fightning. Town watchmens do. May be AI don't recognize enforcers as a crime fighters?
3) @TB. I disagree with you about strange wars in my game. I watched closely for a couple of that wars. AI have only one enemy. Attacking one have three times superior army globally and ten times on battle area. But attacking stack just stand in four tiles in front of city. 20 turns. That city have 100% defence. Attacking stack have enough battering rams but may be AI don`t understand what they may do. Next war don't became strange - attacking stack have few arsonists.
Now I trying to play in the classical era. May be I can find a some issues here?

 

[Thunderbrd]

2) Enforcers still not in use in crime fightning. Town watchmens do. May be AI don't recognize enforcers as a crime fighters?

Just recently fixed. It was a problem with the prehistoric era.

3) @TB. I disagree with you about strange wars in my game. I watched closely for a couple of that wars. AI have only one enemy. Attacking one have three times superior army globally and ten times on battle area. But attacking stack just stand in four tiles in front of city. 20 turns. That city have 100% defence. Attacking stack have enough battering rams but may be AI don`t understand what they may do. Next war don't became strange - attacking stack have few arsonists.
Now I trying to play in the classical era. May be I can find a some issues here?

If the stack hasn't triggered the 'ready to attack' status because it hasn't built to it's desired goals, it's difficult to get it to mobilize. I'm not going to pretend that city attack AI is refined at this point. There's a lot it still needs to be improved by. It is not the time for that since we're only looking at debugging major bugs right now. You're talking about a lot that will be addressed in an eventual overhaul. However, I'm happy enough with where it's currently at because I just watched an autoplayed AI get its entire nation captured by an incoming neighbor's invasion force. With this being possible, that's enough. For now.

It's possible that the AI needs to understand that the rams can ignore the barrier to attack... that could be something to take a look at. I would've assumed it only asks if the unit can attack and doesn't attempt to preempt the city being unattackable before making that inquiry.

That's a shame. I like to run with a merged group to a position and then split them up for the actual fighting. Biggest reason being upkeep while moving and healing faster as split than grouped after the fighting. Not possible any more with the free promotions, which are often counted upon, being removed from 2 thirds.

I'll try to improve on that soon, so as to provide some kind of promo tag that changes this behavior if the promo is free. Some promotions you don't want this behavior, others you do. I've now had it both ways and it appears it would be best to have an in-between solution.

 

[hau]

@TB
I agree, now battle AI is not critical, at least in prehistoric and ancient era. However, stack composition must be improved in future, swordsmans and axemans must be.
But I have one more issue - AI don't build spies at all.

 

[Thunderbrd]

@TB
I agree, now battle AI is not critical, at least in prehistoric and ancient era. However, stack composition must be improved in future, swordsmans and axemans must be.
But I have one more issue - AI don't build spies at all.

Agreed with many additional observations about stack comp improvements to add.

Spies? I can look into it. If complex then it should wait but if simple it could be immediate... relatively. Ill be without my computer for a day or two here.

 

[olduser]

I thought Alberts2 wanted us to not touch the XML files for awhile.

A full xml changes lock-out wasn't my intention.

I make my changes using a script i wrote for that so i won't have problems with any xml changes. But you guys will have a problem in case anyone works on a bigger set of xml changes because merging those changes with my changes could be problematic and time consuming.

It's best to commit all xml changes to the svn before i commit my changes so nobody has a problem.

 

[Dancing Hoskuld]

A full xml changes lock-out wasn't my intention.

I make my changes using a script i wrote for that so i won't have problems with any xml changes. But you guys will have a problem in case anyone works on a bigger set of xml changes because merging those changes with my changes could be problematic and time consuming.

It's best to commit all xml changes to the svn before i commit my changes so nobody has a problem.

OK I have a set of small changes that I'll commit. I do have some big ones to do but not until after the release so I wont start on them until you have done your thing.

 

[hau]

The

 

[hau]

I'm really sorry, but i can't delete previous post.
Something is wrong with production priorities. For example, in my game Mali is in war with half of the world. They have 65k money - but they build money in last city!
I want to attach the save but it's impossible.
I think it may be critical issue.

 

[strategyonly]

CTD, mini attached . .