- AI: easy-to-medium difficulty - even on manical. Probably personality distribution; it's cool that there's a random chance for a hard game, but when it's too easy... ya feel a little cheated.
- Can't tell difficulty of AI opponents.
- What difficulty do the AI homeworlds name map to? aka: Haven = what? Ivory Tower = ?
- AI seems to pass up habitible planets
- Invadable planets were not invaded
- An exobot ship was hoving over an asteroid belt (where the AI had an outpost)
- Sensor ghost made another appearance (though it eventually cleared up)
- Omni-scanner can't see entire universe (intended or no?)
- Turn lag at turn 150 (minor), noticeable (but playable) at 200 (universe: 99 stars, 3 AI)
- Planet Suitability report: Annoying when the best species can't actually produce colony ships
- Psychic Dominance seems to work (previously never saw it proc) in the last few builds (RC1 included)
- Possibly needs a SitRep(?)
- Organic Line -> Quantum ship (not surprisingly) still OP.
- Sometimes the production queue loses the count and repeat settings I give it (not sure how to repro tho)
- Slight audio hang/repeat when dragging a fleet onto another fleet when there is UI lag.
- Fleet's need kill count!
- Occasionally seeing this illegal turn error:
Code: Select all
2013-02-04 20:40:07,817 ERROR Client : HumanClientFSM : A TurnProgress event was passed to the HumanClientFSM. This event is illegal in the FSM's current state. It is being ignored. 2013-02-04 20:40:07,834 ERROR Client : HumanClientFSM : A TurnProgress event was passed to the HumanClientFSM. This event is illegal in the FSM's current state. It is being ignored. 2013-02-04 20:40:07,846 ERROR Client : HumanClientFSM : A TurnProgress event was passed to the HumanClientFSM. This event is illegal in the FSM's current state. It is being ignored. 2013-02-04 20:40:07,863 ERROR Client : HumanClientFSM : A TurnProgress event was passed to the HumanClientFSM. This event is illegal in the FSM's current state. It is being ignored. 2013-02-04 20:40:07,875 ERROR Client : HumanClientFSM : A TurnPartialUpdate event was passed to the HumanClientFSM. This event is illegal in the FSM's current state. It is being ignored. 2013-02-04 20:40:07,892 ERROR Client : HumanClientFSM : A TurnProgress event was passed to the HumanClientFSM. This event is illegal in the FSM's current state. It is being ignored. 2013-02-04 20:40:07,904 ERROR Client : HumanClientFSM : A TurnProgress event was passed to the HumanClientFSM. This event is illegal in the FSM's current state. It is being ignored. 2013-02-04 20:40:07,921 ERROR Client : HumanClientFSM : A TurnProgress event was passed to the HumanClientFSM. This event is illegal in the FSM's current state. It is being ignored. 2013-02-04 20:40:07,933 ERROR Client : HumanClientFSM : A TurnUpdate event was passed to the HumanClientFSM. This event is illegal in the FSM's current state. It is being ignored.
- A touch of performance metrics:
Code: Select all
2013-02-04 20:55:00,157 DEBUG Client : MapWnd::InitTurn getting known starlanes and visible systems time: 929 2013-02-04 20:55:00,158 DEBUG Client : Tech Tree Layout Preparing Tech Data 2013-02-04 20:55:00,274 DEBUG Client : Tech Tree Layout Doing Graph Layout 2013-02-04 20:55:00,276 DEBUG Client : Tech Tree Layout Creating Panels 2013-02-04 20:55:00,278 DEBUG Client : Tech Tree Layout Done 2013-02-04 20:55:00,281 DEBUG Client : MapWnd::InitTurn research wnd refresh time: 124 2013-02-04 20:55:01,782 DEBUG Client : MapWnd::InitTurn sidepanel refresh time: 1501 2013-02-04 20:55:01,783 DEBUG Client : ProductionWnd::UpdateQueue() 2013-02-04 20:55:02,288 DEBUG Client : BuildDesignatorWnd::BuildSelector::Refresh() 2013-02-04 20:55:02,349 DEBUG Client : MapWnd::InitTurn m_production_wnd refresh time: 567
- When combining or splitting a fleet, note that it is spending time Universe::UpdateMeterEstimatesImpl-ing zero (0) elements.
Code: Select all
2013-02-04 20:56:28,200 DEBUG Client : Type times: planet species: 9 ship species: 54 specials: 2 techs: 15 buildings: 7 hulls/parts: 30 fields: 1 2013-02-04 20:56:28,200 DEBUG Client : Universe::UpdateMeterEstimatesImpl on 0 objects time: 118 2013-02-04 20:56:30,716 DEBUG Client : SelectedShipsChanged starting... 2013-02-04 20:56:30,716 DEBUG Client : MapWnd::UpdateMeterEstimates 2013-02-04 20:56:30,716 DEBUG Client : Universe::GetEffectsAndTargets 2013-02-04 20:56:30,834 DEBUG Client : Type times: planet species: 9 ship species: 55 specials: 2 techs: 14 buildings: 7 hulls/parts: 31 fields: 0 2013-02-04 20:56:30,834 DEBUG Client : Universe::UpdateMeterEstimatesImpl on 0 objects time: 118 2013-02-04 20:56:30,856 DEBUG Client : SelectedShipsChanged starting... 2013-02-04 20:56:30,857 DEBUG Client : MapWnd::UpdateMeterEstimates 2013-02-04 20:56:30,857 DEBUG Client : Universe::GetEffectsAndTargets 2013-02-04 20:56:30,976 DEBUG Client : Type times: planet species: 9 ship species: 55 specials: 2 techs: 14 buildings: 8 hulls/parts: 30 fields: 1 2013-02-04 20:56:30,976 DEBUG Client : Universe::UpdateMeterEstimatesImpl on 0 objects time: 119 2013-02-04 20:56:55,057 DEBUG Client : SelectedShipsChanged starting... 2013-02-04 20:56:55,057 DEBUG Client : MapWnd::UpdateMeterEstimates 2013-02-04 20:56:55,057 DEBUG Client : Universe::GetEffectsAndTargets 2013-02-04 20:56:55,176 DEBUG Client : Type times: planet species: 9 ship species: 55 specials: 2 techs: 15 buildings: 7 hulls/parts: 31 fields: 0 2013-02-04 20:56:55,176 DEBUG Client : Universe::UpdateMeterEstimatesImpl on 0 objects time: 119 2013-02-04 20:56:55,218 DEBUG Client : SelectedShipsChanged starting... 2013-02-04 20:56:55,218 DEBUG Client : MapWnd::UpdateMeterEstimates 2013-02-04 20:56:55,218 DEBUG Client : Universe::GetEffectsAndTargets 2013-02-04 20:56:55,335 DEBUG Client : Type times: planet species: 9 ship species: 54 specials: 2 techs: 14 buildings: 7 hulls/parts: 31 fields: 0 2013-02-04 20:56:55,335 DEBUG Client : Universe::UpdateMeterEstimatesImpl on 0 objects time: 117 2013-02-04 20:56:55,391 DEBUG Client : SelectedShipsChanged starting... 2013-02-04 20:56:55,391 DEBUG Client : MapWnd::UpdateMeterEstimates 2013-02-04 20:56:55,391 DEBUG Client : Universe::GetEffectsAndTargets 2013-02-04 20:56:55,508 DEBUG Client : Type times: planet species: 9 ship species: 54 specials: 1 techs: 15 buildings: 8 hulls/parts: 30 fields: 0 2013-02-04 20:56:55,508 DEBUG Client : Universe::UpdateMeterEstimatesImpl on 0 objects time: 117 2013-02-04 20:56:55,527 DEBUG Client : MapWnd::UniverseObjectDeleted
- Another block of perf metrics (this one around the suitability report): Basically, ExecuteEffects takes not much time:
Code: Select all
2013-02-04 21:06:21,781 DEBUG Client : Type times: planet species: 28 ship species: 84 specials: 22 techs: 21 buildings: 10 hulls/parts: 55 fields: 0 2013-02-04 21:06:21,783 DEBUG Client : Universe::ExecuteEffects time: 2 2013-02-04 21:06:21,783 DEBUG Client : Universe::UpdateMeterEstimatesImpl on 1 objects time: 222
v0.4.2 RC1 Discussion
Moderator: Oberlus
Forum rules
Always mention the exact version of FreeOrion you are testing.
When reporting an issue regarding the AI, if possible provide the relevant AI log file and a save game file that demonstrates the issue.
Always mention the exact version of FreeOrion you are testing.
When reporting an issue regarding the AI, if possible provide the relevant AI log file and a save game file that demonstrates the issue.
v0.4.2 RC1 Discussion
(note sure if reply to announcement is proper, feel free to move to "Notes on build X" post if needed)
Code released under GPL 2.0. Content released under GPL 2.0 and Creative Commons Attribution-ShareAlike 3.0.
- Geoff the Medio
- Programming, Design, Admin
- Posts: 13587
- Joined: Wed Oct 08, 2003 1:33 am
- Location: Munich
Re: v0.4.2 Release Candidate #1
The scanlined monster fleet at the nameless no-star system isn't necessarily a ghost, depending on your scanner range. Note that ships on starlanes don't give detection.yandonman wrote:Sensor ghost made another appearance (though it eventually cleared up)
I can try adding an early exit to the meter updating functions. Shouldn't cause any problems, but I'm not entirely sure. It's probably currently spending time looping over the possible source objects and getting their info before checking if any of the zero potential targets are matched by their scopes...yandonman wrote:...spending time Universe::UpdateMeterEstimatesImpl-ing zero (0) elements.
...
Would it break anything to add a condition to the MONSTER_MOVE_ALWAYS macro so that they don't move before turn 10? I don't like how Floaters end up in homeworld systems so quickly, causing confusing battle messages. An extra 10 turns gives some time to research active radar, which I might also drop to 50 RP to make it more readily accessible.
- eleazar
- Design & Graphics Lead Emeritus
- Posts: 3858
- Joined: Sat Sep 23, 2006 7:09 pm
- Location: USA — midwest
Re: v0.4.2 RC1 Discussion
Intended. It's ultimate-ness is that it can see through any stealth. Is the description confusing?yandonman wrote:]Omni-scanner can't see entire universe (intended or no?)
Sounds fine.Geoff the Medio wrote:Would it break anything to add a condition to the MONSTER_MOVE_ALWAYS macro so that they don't move before turn 10? I don't like how Floaters end up in homeworld systems so quickly, causing confusing battle messages.
- eleazar
- Design & Graphics Lead Emeritus
- Posts: 3858
- Joined: Sat Sep 23, 2006 7:09 pm
- Location: USA — midwest
Re: v0.4.2 RC1 Discussion
Playtesting Thoughts
1st game with RC1, at turn 74.
* The Industrial Center line of buildings is too cheap, especially the middle one. You can knock them all out in the first 100 turns, and relative to the most of the other global production boosters, there's no special location condition, and the building is very cheap.
Instead of 20 -> 40 -> 200, i'd like to see something like 30 -> 150 -> 500
* Organic Hull line seems very badly paced. I'll usually have discovered several different (but pretty similar) hulls before i get the necessary buildings for any of them built.
* It is really hard to see the techs on the extreme right or lower edges of the tree. Even with the "Display" window moved out of the way, techs refuse to let you pan them into view, unless you zoom out to the point where text starts taking a legibility hit.
1st game with RC1, at turn 74.
* The Industrial Center line of buildings is too cheap, especially the middle one. You can knock them all out in the first 100 turns, and relative to the most of the other global production boosters, there's no special location condition, and the building is very cheap.
Instead of 20 -> 40 -> 200, i'd like to see something like 30 -> 150 -> 500
* Organic Hull line seems very badly paced. I'll usually have discovered several different (but pretty similar) hulls before i get the necessary buildings for any of them built.
* It is really hard to see the techs on the extreme right or lower edges of the tree. Even with the "Display" window moved out of the way, techs refuse to let you pan them into view, unless you zoom out to the point where text starts taking a legibility hit.
Re: v0.4.2 RC1 Discussion
Would it be possible to have a mouse-over window that shows AI difficulty in the empire window, or have a "show difficulty" option in the right click drop list for AI empires?Can't tell difficulty of AI opponents.
The AI wants to colonize the asteroid belt with the exobots. You probably caught it in action, unless the exobot ship was hovering there for many turns...An exobot ship was hoving over an asteroid belt (where the AI had an outpost)
I have a patch that roughly balances the production cost of all hulls, maybe you might want to give it a try: viewtopic.php?f=15&t=7183Organic Line -> Quantum ship (not surprisingly) still OP.
This would be very helpful .Fleet's need kill count!
Seconded.The Industrial Center line of buildings is too cheap, especially the middle one. You can knock them all out in the first 100 turns, and relative to the most of the other global production boosters, there's no special location condition, and the building is very cheap.
Instead of 20 -> 40 -> 200, i'd like to see something like 30 -> 150 -> 500
Would it be better to adjust the RP and/or turn count of the techs or to make the hulls even more differentiated?Organic Hull line seems very badly paced. I'll usually have discovered several different (but pretty similar) hulls before i get the necessary buildings for any of them built.
- eleazar
- Design & Graphics Lead Emeritus
- Posts: 3858
- Joined: Sat Sep 23, 2006 7:09 pm
- Location: USA — midwest
Re: v0.4.2 RC1 Discussion
"Need"? "Helpful?" I can see how it might be fun, but not useful.unjashfan wrote:This would be very helpful .Fleet's need kill count!
The scope of the changes made as we move towards 0.4.2 should be strongly influenced by what the AI can easily be made to accommodate. I don't know exactly where that line is.unjashfan wrote:I have a patch that roughly balances the production cost of all hulls, maybe you might want to give it a try: viewtopic.php?f=15&t=7183Organic Line -> Quantum ship (not surprisingly) still OP.
Seconded.The Industrial Center line of buildings is too cheap, especially the middle one. You can knock them all out in the first 100 turns, and relative to the most of the other global production boosters, there's no special location condition, and the building is very cheap.
Instead of 20 -> 40 -> 200, i'd like to see something like 30 -> 150 -> 500Would it be better to adjust the RP and/or turn count of the techs or to make the hulls even more differentiated?Organic Hull line seems very badly paced. I'll usually have discovered several different (but pretty similar) hulls before i get the necessary buildings for any of them built.
There doesn't seem to me to be much sense in having a bunch of very similar hulls even if they are spaced out by research. I don't focus on the combat side as much as some others, so i may miss some nuances, but it doesn't seem to me that the "regenerates structure and fuel" is enough practical differentiation between the two organic lines.
Re: v0.4.2 RC1 Discussion
I agree. The distinction is nice in principle, but I've never been able to notice it in practice, and it hardly seems worth the ensuing bluriness of choices.eleazar wrote:There doesn't seem to me to be much sense in having a bunch of very similar hulls even if they are spaced out by research... it doesn't seem to me that the "regenerates structure and fuel" is enough practical differentiation between the two organic lines.
(Taken out of order)
Here's my thoughts: first and foremost, I would suggest avoiding doing more than minor restructurings with Release Candidates. That said, from the AI persepctive paring down the Organic line could probably be done as still a fairly minor restructuring (again, from the AIs standpoint). If we pare down the Organic hulls, it would help if, at least for now, we left the line with the hulls the AI is using; the major ones it relies on are - Organic, Endomorphic, Endosymbiotic, Sentient. There's a couple others in that line that it uses some but could be dropped without leaving big gaps in AI capability. Increasing research costs is fine so long as , at least for now, it's still viable relative to the other hull lines -- I can't get into dynamically evaluating hull lines, and probably even simply planning for a significantly different hull line, anytime soon, but I can put a little time into at least partly adjusting tech research priorities to deal with RP cost changes.The scope of the changes made as we move towards 0.4.2 should be strongly influenced by what the AI can easily be made to accommodate. I don't know exactly where that line is.
As far as other stuff, I can adjust the AI for colonization costs within the ranges we've been discussing without making it any worse than it is now
Is there anything else significant we're thinking of changing?
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0
Re: v0.4.2 RC1 Discussion
Another (minor, but still annoying) issue turned up at my last paytest: the Solid Core special has now a stealth rating of 20, thus it's not detected with the inital empire detection strength. Which lead to a nasty surprise: The first planet I colonized, a medium sized planet with adequate EP (after having researched Subterranean Habitation), which was displayed giving me a pop of 3, in fact had the pop slowly die out after colonization.
Especially with how expensive colonization is now, this is very confusing and un-fun for the player (at first I thought I've run into a bug). Even if the probability for that to happen isn't particulary high, it still can happen (as my playtest proved), and will likely knock you out. An AI player most certainly will not be able to handle that. So I recommend to set the stealth rating of the Solid Core special back to 0.
If there are no objections, I'll commit the change myself.
Especially with how expensive colonization is now, this is very confusing and un-fun for the player (at first I thought I've run into a bug). Even if the probability for that to happen isn't particulary high, it still can happen (as my playtest proved), and will likely knock you out. An AI player most certainly will not be able to handle that. So I recommend to set the stealth rating of the Solid Core special back to 0.
If there are no objections, I'll commit the change myself.
Re: v0.4.2 RC1 Discussion
For me, it would be helpful because I'd want to know how well my fleets are doing, especially after I reload a saved game. It's hard to keep track of every fleet that I might have, and it can help me plan out what I want to produce each turn instead of making extra ships just in case. But that's just me, maybe I try to do too many things at once."Need"? "Helpful?" I can see how it might be fun, but not useful.
Right. When it comes to just research cost changes though, I don't see anything that might significantly impact the AI; it's just the matter of how many more turns it would take for an AI to research something and this applies to all players too.The scope of the changes made as we move towards 0.4.2 should be strongly influenced by what the AI can easily be made to accommodate. I don't know exactly where that line is.
The performance of ship hulls predominately depends on what ship parts are installed. I haven't really deeply explored the possibilties yet, mainly because the shield parts are still not differentiated from armour parts as of now; I want to experiment with the new shields and see how they impact the hulls before I suggest any major changes. Regarding the two organic lines, it's hard to differentiate them since there's so many of them. I also agree that the difference between the two lines is minimal, especially with the fuel factor. So far, with the exception of the early game scouts, none of my ships in recent playtests have run out of fuel since they were almost always within my supply range. Scouting has become less practical as well with the introduction of the scanning facility, so fuel tends to become forgotten.There doesn't seem to me to be much sense in having a bunch of very similar hulls even if they are spaced out by research. I don't focus on the combat side as much as some others, so i may miss some nuances, but it doesn't seem to me that the "regenerates structure and fuel" is enough practical differentiation between the two organic lines.
Shields. Changing them from extra HP to damage reduction. This will affect the ships the AI will make. As I recall, the idea was supported but the discussion somehow died out and this ended up going nowhere. Here's the thread: viewtopic.php?f=6&t=6913&start=15Is there anything else significant we're thinking of changing?
Re: v0.4.2 RC1 Discussion
Yes, please set the stealth rating to 0. In fact i think we should set most of the special's stealth to 0, because a player will not see them even when he has colonized the planet (i.e. your citizens are using the natural tunnels, but you can't see them ) .Vezzra wrote:Another (minor, but still annoying) issue turned up at my last paytest: the Solid Core special has now a stealth rating of 20, thus it's not detected with the inital empire detection strength. Which lead to a nasty surprise: The first planet I colonized, a medium sized planet with adequate EP (after having researched Subterranean Habitation), which was displayed giving me a pop of 3, in fact had the pop slowly die out after colonization.
Especially with how expensive colonization is now, this is very confusing and un-fun for the player (at first I thought I've run into a bug). Even if the probability for that to happen isn't particulary high, it still can happen (as my playtest proved), and will likely knock you out. An AI player most certainly will not be able to handle that. So I recommend to set the stealth rating of the Solid Core special back to 0.
If there are no objections, I'll commit the change myself.
All released under the GNU GPL 2.0 and Creative Commons Attribution-ShareAlike 3.0 licences.
- Geoff the Medio
- Programming, Design, Admin
- Posts: 13587
- Joined: Wed Oct 08, 2003 1:33 am
- Location: Munich
Re: v0.4.2 RC1 Discussion
That is not happening before v0.4.2.unjashfan wrote:Shields. Changing them from extra HP to damage reduction. [...]the idea was supported but the discussion somehow died out and this ended up going nowhere.
- eleazar
- Design & Graphics Lead Emeritus
- Posts: 3858
- Joined: Sat Sep 23, 2006 7:09 pm
- Location: USA — midwest
Re: v0.4.2 RC1 Discussion
revision 5715Vezzra wrote:Another (minor, but still annoying) issue turned up at my last paytest: the Solid Core special has now a stealth rating of 20, thus it's not detected with the inital empire detection strength. Which lead to a nasty surprise: The first planet I colonized, a medium sized planet with adequate EP (after having researched Subterranean Habitation), which was displayed giving me a pop of 3, in fact had the pop slowly die out after colonization.
Code: Select all
"removed stealth from specials not specifically stealth related, and removed last vestige of LRN_RUIN_MOD, and commented out Super Tester Takeover in preunlocked to avoid error messages."
So perhaps we comment out Symbiotic, Protoplasmic, and Bio-Adaptive, and double or triple the cost of the remaining organic hulls?Whatever the minimum change is to make it not feel broken.Dilvish wrote:Here's my thoughts: first and foremost, I would suggest avoiding doing more than minor restructurings with Release Candidates. That said, from the AI persepctive paring down the Organic line could probably be done as still a fairly minor restructuring (again, from the AIs standpoint). If we pare down the Organic hulls, it would help if, at least for now, we left the line with the hulls the AI is using; the major ones it relies on are - Organic, Endomorphic, Endosymbiotic, Sentient. There's a couple others in that line that it uses some but could be dropped without leaving big gaps in AI capability. Increasing research costs is fine so long as , at least for now, it's still viable relative to the other hull lines -- I can't get into dynamically evaluating hull lines, and probably even simply planning for a significantly different hull line, anytime soon, but I can put a little time into at least partly adjusting tech research priorities to deal with RP cost changes.
And changing the cost of the industrial center series?Dilvish wrote:As far as other stuff, I can adjust the AI for colonization costs within the ranges we've been discussing without making it any worse than it is now
What's the deal with Death Spores and Bio-Terminators?
Bio-T requres Death S. as a prereq, but only costs 75 RP to Death S.'s 450 RP.
Re: v0.4.2 RC1 Discussion
sure, it will be easy enough to at least come up with a passable adjustment to the schedule for the the two refinements, delaying them to where they won't (be expected to) unreasonably tie up research for a long while.eleazar wrote:And changing the cost of the industrial center series?Dilvish wrote:As far as other stuff, I can adjust the AI for colonization costs within the ranges we've been discussing without making it any worse than it is now
If I provided any code, scripts or other content here, it's released under GPL 2.0 and CC-BY-SA 3.0
- eleazar
- Design & Graphics Lead Emeritus
- Posts: 3858
- Joined: Sat Sep 23, 2006 7:09 pm
- Location: USA — midwest
Re: v0.4.2 RC1 Discussion
Also noticed-- it looks like the repetition thing for ship building isn't kind to changes.
If i have for instance 5x ships being built, and then after some progress, set it to 10x ships, the progress bar goes back by what looks like a reasonable amount. If i then change my mind and set it back to 5x or 1x, the progress bar does not move forward-- it looks like you loose RP.
Fiddling with your setting during a single turn should never change the outcome -- only the setting present when you hit "TURN" should matter.
Also, the Planet Suitability report is very handy, but it includes in the list species that cannot colonize. I think it would be best if it excluded from the list non-colonizing species unless they are already living on the planet in question.
If i have for instance 5x ships being built, and then after some progress, set it to 10x ships, the progress bar goes back by what looks like a reasonable amount. If i then change my mind and set it back to 5x or 1x, the progress bar does not move forward-- it looks like you loose RP.
Fiddling with your setting during a single turn should never change the outcome -- only the setting present when you hit "TURN" should matter.
Also, the Planet Suitability report is very handy, but it includes in the list species that cannot colonize. I think it would be best if it excluded from the list non-colonizing species unless they are already living on the planet in question.
Re: v0.4.2 RC1 Discussion
The new nacent art. intelligence feels quite OP. It essentially gives all planets an automatic research focus, and also is worth the equivalent of algorithmic elegance and distributed thought computing combined. It's possible to survive with all planets focused to industry, which sort of takes some flavour out of the gameplay.