C2C SVN Changelog

[Koshling]

Today's changes:

• Moved the AI pathing option from player options to a BUG option (in the general tab)
• Added another option (also in the BUG options general tab) to auto-end turns in which there are no human decisions to be made

The first of these changes will remove the problem of incompatibility when switching between different mods. However, it does mean there is a ONE-OFF options incompatibility vs recent SVN versions, so SVN users will lose their options once more (only) as they switch to this.

The second change is to make the phenomenum of 'just press ENTER' turns a little easier to skip over, especially in early pre-historic. Basically with this option set, the turn will auto-end (without you needing to press enter) on any turn in which none of the following occured:

• You had a non-automated unit to move
• A unit of yours was attacked
• You had to interact with a popup (e.g. - choose new research)
• Any diplomacy occured in which you were involved
• Any 'the enemy has been spotted' messages occured

Pressing ESCAPE (which brings up the main game menu as always) will force it to register a popup interaction and allow you to make any chnages you need to, should you need to interrupt the automated continuing.

 

[Sgtslick]

A unit of yours was attacked

Does this mean only when you need to move it to heal etc, if the ai has chosen to heal it itself, which it often does when you auto-exploring or whatever, will this still cancel this auto end turn function? Coz your units are attacked literally every turn usually

 

[Dancing Hoskuld]

Update
- Adjusted return from Wembley Stadium to one tenth of what it was.

 

[Koshling]

A unit of yours was attacked

Does this mean only when you need to move it to heal etc, if the ai has chosen to heal it itself, which it often does when you auto-exploring or whatever, will this still cancel this auto end turn function? Coz your units are attacked literally every turn usually

It does at the moment, because for me at least, I like to take more intelligent healing decisions than the automation would necessarily make. I could ad a global define (say) that would define flags for which conditions it considers to be requirign of interaction (then you could just turn off the flag for attacks if thats what you want)...?

 

[strategyonly]

Changed Hero Canute from Bonus Dutch to Viking

 

[Koshling]

XML issue with latest SVN:

• TECH_CANAL_SYSTEMS is a prereq for TECH_ENGINEERING but is not defined!

Edit - also CIVILIZATION_MAORI has:

<DerivativeCiv>CIVILIZATION_POLYNESIA</DerivativeCiv>

but CIVILIZATION_POLYNESIA is not defined

 

[strategyonly]

XML issue with latest SVN:

• TECH_CANAL_SYSTEMS is a prereq for TECH_ENGINEERING but is not defined!

Edit - also CIVILIZATION_MAORI has:

<DerivativeCiv>CIVILIZATION_POLYNESIA</DerivativeCiv>

but CIVILIZATION_POLYNESIA is not defined

My bad, thats what i get for not checking closely. Hopefully fixed.
Caste System and None.

 

[Koshling]

I started playing a new GEM game on my new (well, new to me - it's actually a hand-me-down) laptop yesterday, and was immediately struck by how slow the turns seemed right at the start when there is little to do (of course this IS a gigantic map with 40ish civs).

Anyway, as a result I took a look at the eraly game profile today, and have started addressing points that are signficant to early game performance.

These are quite different to the things that tend to contribute the most to more mature games (where its large numbers of cities and units that dominate). In the early game it's the constant overheads that are per-turn that take up the most time. They may only be half a second here and half a second there, but in a 6 second turn those are pretty significant. I had an initial crack at optimizing some of them today (lopped off maybe 25% so far of the turn time on my GEM game), and plan to continue for the next couple of days since there looks to be fairly low-hanging fruit to be picked.

Pushed to SVN today:

• Improved performance of civic evaluation in the early game when few choices are available
• Improved performance of Python canBuild callbacks (basically this is Great Farmer stuff which has a quite high peruturn overhead regardless of whether there are any Great farmers in the game)
• Improved performance of city founding location evaluation
• Fixed a small memory leak and a few holes in memory leak detection

 

[strategyonly]

I started playing a new GEM game on my new (well, new to me - it's actually a hand-me-down) laptop yesterday, and was immediately struck by how slow the turns seemed right at the start when there is little to do (of course this IS a gigantic map with 40ish civs).

Glad i wasn't the only one that noticed this, that is the MAIN reason, i stopped playing on the GEM map, as you stated, right from the git go, turn lag, actually bad turn lag. . . ...... ......... .. . . . .

 

[DRJ]

The interesting thing is that the GEM starts slow but gets actually faster once the animals are less and the empires larger (in my recent game even hunting south america nearly animal free helped quite a bit!). But because of the early slowness I have requested/suggested several times that the sea critters should only start spawning when the first civ has reached ancient era, not earlier (as they can't be hunted before that, anyway). My calls have not been heard so far.

Once we have herbivore/carnivore barbs the problem will solve itself but as long as this intelligent barb design is not implemented yet it is a bit slow indeed...
but with recent changes it has been getting better, to my feeling turntimes are just a third than they were during last phase of v21 SVN GEM.

 

[Koshling]

The interesting thing is that the GEM starts slow but gets actually faster once the animals are less and the empires larger (in my recent game even hunting south america nearly animal free helped quite a bit!). But because of the early slowness I have requested/suggested several times that the sea critters should only start spawning when the first civ has reached ancient era, not earlier (as they can't be hunted before that, anyway). My calls have not been heard so far.

Once we have herbivore/carnivore barbs the problem will solve itself but as long as this intelligent barb design is not implemented yet it is a bit slow indeed...
but with recent changes it has been getting better, to my feeling turntimes are just a third than they were during last phase of v21 SVN GEM.

Animal movement is about 0.5 seconds of the 6 or so turn time on my early GEM game. It's really not the most significant factor (forth or so I'd say). However, I am going to optimize it a bit by having animals nowhere near occupied territory just skip turns most of the time I think. Having said that, it's other factors (the biggest by far is Python callback overhead, which I'm working on - the great farmer onBuild callback I mentioned above was a chunk of it)

 

[Koshling]

In the latest SVN apples are enabled by naturopathy, but not revealed until orchards! Suit they should be revealed with othe fruit (which I think is gathering?)

 

[Dancing Hoskuld]

The interesting thing is that the GEM starts slow but gets actually faster once the animals are less and the empires larger (in my recent game even hunting south america nearly animal free helped quite a bit!). But because of the early slowness I have requested/suggested several times that the sea critters should only start spawning when the first civ has reached ancient era, not earlier (as they can't be hunted before that, anyway). My calls have not been heard so far.

Once we have herbivore/carnivore barbs the problem will solve itself but as long as this intelligent barb design is not implemented yet it is a bit slow indeed...
but with recent changes it has been getting better, to my feeling turntimes are just a third than they were during last phase of v21 SVN GEM.

And as I replied we can delay spawning to a specified "date" but not to a specified tech. So what date should they start spawning after?

 

[DRJ]

Let's say 4500 BC? Has anyone seen earlier progress into ancient? I think it depends on difficulty level though. If I play on settler I can reach rafting maybe earlier, but on deity I need much longer...


So, sry if I don't understand it yet:

if I have rifling, after a while barbs get rifles too - so you say that's triggered by date?

I'd say it wents by "who's how far in the game" era/tech-wise, since vanilla?

Would barbs also get rifles if all Civs still stuck in medieval at a given date or would it be other factors weighing in for that to happen as well?

Thinking about something like 'if {any} civXY researches {any of the first ancient techs}, sea animals are triggered to spawn'

Doesn't it work that you get events being triggered by tech? I believe there are those!
For instance if you build a scout you can trigger the 'all scouts promoted to combat I to survive the age of the bear'.

What if any civ builds a raft you trigger 100% a not-visible one time event/a modified "hidden quest", so the player just sees them spawn once that has been executed?


I really think there are ways to tie that to technology, albeit making some detours?

 

[Hanny]

Animal movement is about 0.5 seconds of the 6 or so turn time on my early GEM game. It's really not the most significant factor (forth or so I'd say). However, I am going to optimize it a bit by having animals nowhere near occupied territory just skip turns most of the time I think. Having said that, it's other factors (the biggest by far is Python callback overhead, which I'm working on - the great farmer onBuild callback I mentioned above was a chunk of it)

If you going to fiddle with the spawn points in the DL why not make use of Stephi idea in MoM, and have the least dangerouspawn the closest to borders, and the most dangerous furthest away from borders. Held in cache this means no recalculation if im right. This simulalates the increaased lethality of wilderness, compared to the borders
of civilzation and wilderness.

 

[Dancing Hoskuld]

Let's say 4500 BC? Has anyone seen earlier progress into ancient? I think it depends on difficulty level though. If I play on settler I can reach rafting maybe earlier, but on deity I need much longer...


So, sry if I don't understand it yet:

if I have rifling, after a while barbs get rifles too - so you say that's triggered by date?

I'd say it wents by "who's how far in the game" era/tech-wise, since vanilla?

Would barbs also get rifles if all Civs still stuck in medieval at a given date or would it be other factors weighing in for that to happen as well?

Thinking about something like 'if {any} civXY researches {any of the first ancient techs}, sea animals are triggered to spawn'

Doesn't it work that you get events being triggered by tech? I believe there are those!
For instance if you build a scout you can trigger the 'all scouts promoted to combat I to survive the age of the bear'.

What if any civ builds a raft you trigger 100% a not-visible one time event/a modified "hidden quest", so the player just sees them spawn once that has been executed?


I really think there are ways to tie that to technology, albeit making some detours?

We moved spawning of animals out of the barbarian spawn code because they would not spawn after a certain turn in the game. The type of barbarians spawned is dependent on the game era which is based on each player (human and AI) era. Animal spawns are just on year now. I'll go for -5000 to start with so the seas are not empty and we will adjust it from there. It will take me a couple of days to get to it as I need to recover from my weekend which was mentally exhausting.

If you going to fiddle with the spawn points in the DL why not make use of Stephi idea in MoM, and have the least dangerouspawn the closest to borders, and the most dangerous furthest away from borders. Held in cache this means no recalculation if im right. This simulalates the increaased lethality of wilderness, compared to the borders
of civilzation and wilderness.

There is nothing in the spawn XML to support this at the moment. Animals spawn based on terrain not distance from populated lands.

 

[Hydromancerx]

Updates
- Updating my FPK.
- Tweaked Paved Roads building.
- Added Trails and Roads buildings.

 

[DRJ]

We moved spawning of animals out of the barbarian spawn code because they would not spawn after a certain turn in the game. The type of barbarians spawned is dependent on the game era which is based on each player (human and AI) era. Animal spawns are just on year now. I'll go for -5000 to start with so the seas are not empty and we will adjust it from there.

Thx for the clarification. I nearly thought I was hallucinating about the barbs being tied to tech/era in vanilla^^ So there can't be a modification of the original animal code in regard to set a 'no value' of the turn when they are not to be spawned anymore by original code?

And what about a triggering quest/event tech for animal spawn is that an option for later maybe? But I should be happy - 5000 BC is better than 12000 BC and will surely save some playtime on GEM or other huge maps, thx!!!

It will take me a couple of days to get to it as I need to recover from my weekend which was mentally exhausting.

Get well soon!

But remember: he who is partying hard has also be able to work hard right after (at least that's what old people here in Germany tended to tell me when I was young and wild and not yet 30ish and totally adepted to capitalistic societies rules

Ok who am I kidding^^

 

[Hanny]

There is nothing in the spawn XML to support this at the moment. Animals spawn based on terrain not distance from populated lands.

I refered to the DLL spawn, which can/does allow distance from cultural borders, vis of units etc, and would bea nice refinment of the spawn location points.

 

[strategyonly]

I refered to the DLL spawn, which can/does allow distance from cultural borders, vis of units etc, and would bea nice refinment of the spawn location points.

With all this info your providing, is there anyway you know how to mod??