Single Player bugs and crashes v37 plus (SVN) - After the 24th of December 2016

[Noriad2]

Not a bug - distance attacks don't contribute to GP pts.

Yes they do. But not immediately as the video proves. You get the GG points from distance attacks a bit later.

 

[Thunderbrd]

Yes they do. But not immediately as the video proves. You get the GG points from distance attacks a bit later.

hmm... not supposed to be. Are you sure the melee attack didn't actually generate that much xp itself?

 

[Noriad2]

hmm... not supposed to be. Are you sure the melee attack didn't actually generate that much xp itself?

I believe xp and gg points are separate subjects?
Anyway, the first example I pointed at (with timestamps on the vid) has no battle occurring after the ranged attacks, but a minute later, the GG points counter suddenly updates anyway.

 

[Thunderbrd]

I believe xp and gg points are separate subjects?

Sorta. GG Pts are based on XP earned, with exceptions and modifiers. You can get a LOT of GG pts from wonder modifiers, particularly for combat within your borders. Not all sources of XP gain will give GG pts either.

Anyway, the first example I pointed at (with timestamps on the vid) has no battle occurring after the ranged attacks, but a minute later, the GG points counter suddenly updates anyway.

hmm... then it's a caching factor, again, that would not be a bug but one of the things keeping the game from processing too slow. And if it IS assigning GG pts for ranged assault, that should be changed or the amount modified by greater/lesser group volumes - at least on SM games. By far the easiest code solution is to not have Ranged Assault XP add to GG gains.

Could the jump in GG pts there be a result of XP gains from defenses made over the course of the end of turn processing for the AI?

 

[Noriad2]

Could the jump in GG pts there be a result of XP gains from defenses made over the course of the end of turn processing for the AI?

I have NO idea. I just noticed it in IdioticUlt1mara's video, and provided a link with timestamps as evidence of something odd. I did not do tests as it was not my game. I also had the illusion that 1 won battle equals 1 gg point (modified by wonders). Apparently that is untrue.

 

[strategyonly]

Its still happening, i am about in the middle of the Classical Era, and one civ is all the way to the Industrial era already, i tried to beeline but i guess to no avail??

 

[raxo2222]

Its still happening, i am about in the middle of the Classical Era, and one civ is all the way to the Industrial era already, i tried to beeline but i guess to no avail??

Similar thing happened, when I was play testing on my spacemap for calendar overflow on Normal speed with Gigantic map and Deity/Nightmare difficulty.
I was far into AD years or so, when my civilization on AI autoplay reached classical era.
I guess such combination made tech progression waay too slow.
I guess you are playing on Normal/Epic speed and Large/Huge map with Deity+Nightmare.
Seems like tech costs are scaled disproportionally.

Edit: My endgame savegame is perfect for testing gold -> greater gold conversion.
That conversion can't keep up with my income of 10 000 000 gold per turn
On unrelated note game takes only 1.5 GB of memory after loading this savegame

 

[JosEPh_II]

Its still happening, i am about in the middle of the Classical Era, and one civ is all the way to the Industrial era already, i tried to beeline but i guess to no avail??

Can't wait till you met Hadrian.

My current Normal Test game is doing the same thing, in Classical era and it's 1572AD while researching Mounted Archer.

Another 23 techs till Medieval Lifestyle. But I have not met any neighbors yet either.

Have not pinpointed how the Research rate changed. Only have a guess as to why, New Option.

 

[Thunderbrd]

Can't wait till you met Hadrian.

My current Normal Test game is doing the same thing, in Classical era and it's 1572AD while researching Mounted Archer.

Another 23 techs till Medieval Lifestyle. But I have not met any neighbors yet either.

Have not pinpointed how the Research rate changed. Only have a guess as to why, New Option.

And that's not it. I verified that in the code the other day.

 

[raxo2222]

So basically Normal speed doesn't downscale tech costs enough

 

[JosEPh_II]

So basically Normal speed doesn't downscale tech costs enough

It was balanced and did not have this problem just a few months ago. Something was changed in the code and Now current games are acting like this. So you tell me what happened as I made no changes to Normal GS modifiers. Dates, eras, and research rates were hitting their "target" settings. Now it's not.

But, there was a Greater Gold change to the code......another possibility. The Tech recosting did have some effect as well. Imho this change is a code change somewhere.

So until and if it's found I will need to redo the research rates all over again it would seem.....

 

[raxo2222]

It was balanced and did not have this problem just a few months ago. Something was changed in the code and Now current games are acting like this. So you tell me what happened as I made no changes to Normal GS modifiers. Dates, eras, and research rates were hitting their "target" settings. Now it's not.

But, there was a Greater Gold change to the code......another possibility. The Tech recosting did have some effect as well. Imho this change is a code change somewhere.

Maybe building rebalance increased cost of buildings, and it resulted in less research/commerce producing buildings being built?

 

[JosEPh_II]

And that's not it. I verified that in the code the other day.

So what did your Greater Gold make changes to? Gold and Research go hand in hand..........

And If this is a manifestation of the Tech Recosting project, then it sure did take awhile to show itself....

SO's and my game were started pretty close together in relation to svn started from. My prior games did not show this problem. And as I stated this game of mine on Normal was started with SVN 9731.

 

[JosEPh_II]

Maybe building rebalance increased cost of buildings, and it resulted in less research/commerce producing buildings being built?

It definitely changed the order of When buildings got built.

Just looked at a 9704 Normal game and it is showing the same research degradation. It was started back on Sept. 29th, and only played on that day up to 27xxBC. Was still in Ancient Era, researching Oratory, but Classical was still 20 tech away if bee lined.

I then stopped playing C2C for almost a month till I started the 9731 game. Gotta have breaks now and then.

 

[Thunderbrd]

So what did your Greater Gold make changes to? Gold and Research go hand in hand..........

All that did was make it so that if you get over 1mil gold it converts to Greater Gold and reduces the normal gold amount by 1mil. And then the opposite - go beneath 0 gold and you exchange in a Greater Gold to cover it if necessary. It's a currency exchange system basically. 1 Greater Gold = 1 million gold. It keeps things from overflowing in gold totals. Nothing that would influence research rates.

But the building recosting certainly would make it tougher to get the research buildings as quickly as we once did so that would cause some longer research times.

I thought you HAD made an adjustment to research times lately that would account for what you're talking about. Maybe review other recent changes? Coding adjustments have not been made that would be adjusting research rates lately.

 

[JosEPh_II]

I thought you HAD made an adjustment to research times lately that would account for what you're talking about. Maybe review other recent changes? Coding adjustments have not been made that would be adjusting research rates lately.

10% increase for Snail and longer GS. But nothing to Normal and Epic. Maybe it was marathon and longer, anyway the 2 fastest GS were not changed. And SO's continued Bug postings correlate to what is happening in my own Normal game(s). Just trying to figure it out, exploring the possibilities. And you should agree coding can have far reaching effects wouldn't you?

 

[Thunderbrd]

And you should agree coding can have far reaching effects wouldn't you?

CAN, yes. If there is logical cause for it to. It's not like a mystic thing where anything can happen with the slightest change though. It is very possible to be completely sure of the inability for something to have an effect due to the manner in which it is applied.

For example, the option you are pointing to does nothing but eliminates the processing of WFL and TD modifiers for human players and only if those options are on. It does nothing else to the processing of the research and the modifier amount is added to the rest of the stream later only if there is a modifier to add, which this makes sure there wouldn't be.

Given that it shows up nowhere else in the code I can say with confidence that the ability for it to influence results on non WFL/TD games is absolutely impossible.

I do not recall anything else that could influence the research stream. Except, as noted, some XML changes that can have a more significant impact than perhaps imagined.

 

[tmv]

Potential if (a = b) error in the following places:

• Sources\CvGameTextMgr.cpp line 23186, 23221
• Sources\CvPathGenerator.cpp line 929
• Sources\CvPlot.cpp line 5674

 

[JosEPh_II]

CAN, yes. If there is logical cause for it to. It's not like a mystic thing where anything can happen with the slightest change though. It is very possible to be completely sure of the inability for something to have an effect due to the manner in which it is applied.

For example, the option you are pointing to does nothing but eliminates the processing of WFL and TD modifiers for human players and only if those options are on. It does nothing else to the processing of the research and the modifier amount is added to the rest of the stream later only if there is a modifier to add, which this makes sure there wouldn't be.

Given that it shows up nowhere else in the code I can say with confidence that the ability for it to influence results on non WFL/TD games is absolutely impossible.

I do not recall anything else that could influence the research stream. Except, as noted, some XML changes that can have a more significant impact than perhaps imagined.

if it's from the recosting projects then it's taken at least several months to manifest. A real possibility and I'm just trying to reduce those same possibilities.

I will need to start a Normal game w/o the New Option being used and with both TD and WFL Not being used. Then Compare results. I do know that with WFL On the research rate and date correlation is off as well, more tech researched in less time than target date range. The opposite of having the new Option On.

Then there is the 3rd combination of having New Option On and Both TD and WFL On, and it's results.

This is starting to look like a can of worms and they are escaping.

 

[Thunderbrd]

Potential if (a = b) error in the following places:

• Sources\CvGameTextMgr.cpp line 23186, 23221
• Sources\CvPathGenerator.cpp line 929
• Sources\CvPlot.cpp line 5674

I'll take a look... those are notorious issues. Thanks!