Version 0.80 discussion thread

[jdog5000]

Exactly the idea ... changing every tech cost is time consuming, so I'm trying to create the right XML options to easily tweak costs globally.

 

[Roland Johansen]

Research costs might not have been improved linearly per era (with for instance 3% per era as in your example), so it would allow more flexibility if you could set a percentage for each of the 6 eras like: 2%, 5%, 6%, 10%, 12%, 20% (just an example). It will of course be nearly impossible to find the so called perfect percentages but allowing the flexibility of individual adjustable rates per era can't hurt. Maybe you were already planning it this way and your example with 3 percent per era was just a special case.

 

[Afforess]

Any chance of merging in the AI land transport modcomp? I believe that falls under the AI category and would make things more interesting for all the mods that use the Better AI mod.

 

[Maniac]

Yeah, but with the way things are setup that requires a loop over projects as the victory doesn't know what projects it requires ... this isn't bad, but delay turns alone works at least for the default victory types and most of the mod ones I've heard of (religion, mastery, etc).

I have a victory which requires building a building. But I always make the necessary change myself of course.

Any chance of merging in the AI land transport modcomp? I believe that falls under the AI category and would make things more interesting for all the mods that use the Better AI mod.

I would not recommend this at the moment. Judging from a quick look at the code, I have my doubts it's working properly.

 

[TheLadiesOgre]

I would not recommend this at the moment. Judging from a quick look at the code, I have my doubts it's working properly.

What would lead you to believe this? I really would like to know, as I have absolutely no formal training and exceedingly little prior experience coding C++. Instead of just casting doubt, perhaps maybe you'd like to enlighten an aspiring SDK modder as to what he is doing wrong. It has worked in the test games I've played (granted that has only been three), it would appear stable (no crashes or errors) and as it is being used by the AI, I would call it a success.

 

[jdog5000]

Research costs might not have been improved linearly per era (with for instance 3% per era as in your example), so it would allow more flexibility if you could set a percentage for each of the 6 eras like: 2%, 5%, 6%, 10%, 12%, 20% (just an example). It will of course be nearly impossible to find the so called perfect percentages but allowing the flexibility of individual adjustable rates per era can't hurt. Maybe you were already planning it this way and your example with 3 percent per era was just a special case.

Yeah, the correct way to do this is to add a new XML variable to eras to scale the costs of all techs in the era. Will do.

Any chance of merging in the AI land transport modcomp? I believe that falls under the AI category and would make things more interesting for all the mods that use the Better AI mod.

Starting to wonder if there's a conspiracy to get this included, I just got two PMs about it too

I think it's a good idea in principle and I'm not against adding it, it shouldn't slow down standard games at all and would allow modders who use BBAI to easily add a new class of units that the AI would use. However, I don't have time to double check the code and test it myself, so I'd prefer to wait a little while to let it mature so that TheLadiesOgre and the community can work out any issues that may exist in the first version.

 

[Maniac]

I think it's a good idea in principle and I'm not against adding it, it shouldn't slow down standard games at all and would allow modders who use BBAI to easily add a new class of units that the AI would use. However, I don't have time to double check the code and test it myself, so I'd prefer to wait a little while to let it mature so that TheLadiesOgre and the community can work out any issues that may exist in the first version.

Do you think the AI method I suggested here would work?

 

[jdog5000]

As you pointed out, there are two primary issues: getting units and transports to the same place at the same time, and deciding when to load into transports and when to walk.

The same issues exist in the naval transport scenario, it's gotten a bit better in BBAI but there still isn't a great solution to either question quite frankly. For the initial invasion, transports may leave from the wrong side of the island/continent, and if units can walk they usually will even if it's very inefficient.

For land transports of course the units and transports can more easily occupy the same tiles. This poses some challenges units are no longer inherently forced to use them to get to some places, but it also has benefits as you pointed out which can be used. Having some of the troops in a stack in transports, and other not, and all traveling together until enough transports show up seems reasonable.

Also, for certain mods it may be that land transports should always be used if the path enters a certain kind of dangerous terrain ... that's got to be an option somehow.

 

[Dom Pedro II]

The same issues exist in the naval transport scenario, it's gotten a bit better in BBAI but there still isn't a great solution to either question quite frankly. For the initial invasion, transports may leave from the wrong side of the island/continent, and if units can walk they usually will even if it's very inefficient.

Yes, we've talked about it before, but what we really need is for the AI to orchestrate unit movements and actions on the team or player level. I mean, unit-based decision making works well enough for periods without war, but when the AI is preparing for war or in the middle of one, the first step should be to allocate the best units to achieve certain objectives rather than having each unit find its own way. Once all the useful units have been put to work, it should cycle through all the units that haven't been assigned to a task.

Also, is there any chance you'll do work to get the AI to launch amphibious invasions on the same continent?

 

[razor436]

Does this mod affect how cities assign its citizens to work the land or become specialists in player cities?

I noticed that my biggest cities would go no higher than about 15 population. So I investigated and noticed that the city AI greatly favors hammer tiles (for example, choosing a 4hammer tile over a 2food tile). I could understand that, since we need production to build faster, but in the long run I'd rather let the city grow to get more trade for gold and research.

Also, I noticed that whenever a specialist is available to a city, for some reason the city AI allocates a citizen as a specialist rather than working a food tile. Now that seriously cripples the growth of a city.

 

[Dom Pedro II]

Ummm... has anybody tried to actually merge this code with anything? jdog, did you make any changes to this code before including it in this download? Because I'm seeing stuff commented out that really, really needs to be included to compile without errors.

 

[jdog5000]

Also, is there any chance you'll do work to get the AI to launch amphibious invasions on the same continent?

Won't happen for a few months at least ...

Does this mod affect how cities assign its citizens to work the land or become specialists in player cities?

Yes, there have been tweaks which could affect this, but there are also many factors which play a role in particular circumstance. If you upload a screenshot or a save that would be helpful.

Ummm... has anybody tried to actually merge this code with anything? jdog, did you make any changes to this code before including it in this download? Because I'm seeing stuff commented out that really, really needs to be included to compile without errors.

Which code? BBAI or the land transport thing?

 

[JimMac99]

Anyone experience a repeatable crash using 0.80?

Myself and three relatives meet every Sunday to play MP Civ over LAN. We have been using 0.80, as a loaded mod, in the game thats been running for the past three weeks. Last night following an 'End Turn' all four pcs crashed to Desk Top. (Not the out of sync we sometimes get with other mods.)

Loading in the most recent autosave produced a crash at the same point, as did creating and loading a save from a couple of turns prior to the crash.

This morning I loaded the save as a MP via hotseat game, and could see that the event that happened immediately after our crash occurred was AI Huayna Capac placed Stalin in charge of his oversea colonies. This may have no relevancy at all to the crash though!

Anyway, anyone else experience a crash in MP with 0.80? Or is our crash just "one of those things"? Three weeks play written off is a nuisance though, especially when I am in the lead.

 

[Cybah]

Some people were telling me this happened to them in multiplayer when playing my mod (which uses BBAI).

Also, I have a singleplayer savegame which is repeatable crashing on a turn. But also my mod, therefore no upload.

 

[Yakk]

JimMac99, posting saves of games that immediately crash upon loading is quite useful, as are posts of saves that crash after you load the game and hit 'end turn'.

 

[JimMac99]

Cybah and Yakk, thanks for your replies.

I didn't upload a save as I wasn't aware that anyone could examine a MP save in order to test it...

Save file attached. I believe this save is the turn when it crashes for the three remaining human players (I had forgotten one player has resigned a few turns previously).

 

[phungus420]

In my experience MP saves aren't useful because they have modified assets locked, and thus you can't use a debug dll to check it. Hopefully this save presents it's issue without needing a debug dll, but that's been my experience.

 

[Cybah]

A player of my mod just told me it could have something to do with spy AI. After a spy controlled by the AI got caught, the game did not crash.

...but I don't know.

 

[TheOnlyDJCat]

As JimMac99 and I share apparently the same bug in multiplayer, it seems the create colony feature is buggy (see also specific bug report thread)

 

[Mattygerst]

Any clue when a change this big (0.80, as well as the testing of 0.80j for speed increase) would be implemented into RevDCM. Playing standard 0.80 - the difference in late game AI is dramatic enough that it makes 0.78 feel very obsolete.