Post reply

Warning - while you were reading 49 new replies have been posted. You may wish to review your post.

Note: this post will not display until it's been approved by a moderator.

Name:
Email:
Subject:
Message icon:

Verification:
What is the fourth planet?:

shortcuts: hit alt+s to submit/post or alt+p to preview

Please read the rules before you post!


Topic Summary

Posted by: Steve Walmsley
« on: December 11, 2018, 10:56:55 AM »

Field position sets the formation to front-line attack, front-line defence, support or rear-echelon. It affects how the formation participates in ground combat.
Posted by: snapto
« on: December 11, 2018, 10:02:01 AM »

Loved the recent screenies of the npr ai in action.  On the Ground Combat->Order of Battle tab, I was wondering what the "Field Position" button did?  Thanks for all the hard work Steve. Can't wait to play!!
Posted by: King-Salomon
« on: December 06, 2018, 11:43:21 AM »

will it be possible to capture ground-combat equipment and to check it for research or intelligence results?

also about ELINT:

will it be possible for ELINT (or any ship at all) to "see" what kind of weapon etc a hostile ship is using and list this correctly in the intelligence database and the "known technologies" of the race?

for example: the player is in battle with 3 enemy ships ... 1 ship is firing missiles, 1 ship is firing lasers and the third is not firering at all...

can the intelligence report add the information which ship class is using which kind of weapon - (6x size 8 missile launcher for class A, 4x 30cm laser for class B) ? in VB we have to manually add this kind of information's - would be great if it could be added automatically IF the use of the weaponsystem by this ship is clear...

also it would make sense to add something like "30cm laser tech" to the race-infos if known ships of this race are using this kind of weapon...

long words short.. would it be possible to ELINT to "detect" used technology/ship modules and update the information's by itself?
Posted by: Steve Walmsley
« on: December 06, 2018, 10:43:01 AM »

Although the AI will convert conventional factories if they exist, I haven't added full support for progressing from conventional to TN yet. I plan to to do that at some point though.

Out of curiosity, are there unique difficulties in doing this, or is it just low on your priority list? If there are unique difficulties, what are they?

The difficulty is that everything the AI wants to do or build or design has restrictions based on technology. For a TN start, the AI has a definite set of starting tech that can be relied upon (the same TN starting tech that a player receives). For a conventional start, the AI is going to have to check what portion of that potential technology base exists. At the moment, the code doesn't check for technology that is provided for a TN start, so I would have to add those checks in every place in the code where it is needed.
Posted by: Jovus
« on: December 06, 2018, 10:15:18 AM »

Although the AI will convert conventional factories if they exist, I haven't added full support for progressing from conventional to TN yet. I plan to to do that at some point though.

Out of curiosity, are there unique difficulties in doing this, or is it just low on your priority list? If there are unique difficulties, what are they?
Posted by: Steve Walmsley
« on: December 03, 2018, 03:06:30 AM »

I can't recall if this has been covered or not, but will C# Aurora treat TN and non-TN NPR's in the same way as the current Aurora? So will a Non-TN NPR ever be able to progress to a TN civilization?

Although the AI will convert conventional factories if they exist, I haven't added full support for progressing from conventional to TN yet. I plan to to do that at some point though.
Posted by: Rich.h
« on: December 03, 2018, 01:52:45 AM »

I can't recall if this has been covered or not, but will C# Aurora treat TN and non-TN NPR's in the same way as the current Aurora? So will a Non-TN NPR ever be able to progress to a TN civilization?
Posted by: Barkhorn
« on: December 01, 2018, 01:07:31 PM »

I can't decide if I'm more excited about the AI or the performance improvements.  I mostly don't do much roleplay when I play; I'm interested more in playing well.  Better AI and better performance will make this much more rewarding.
Posted by: Jorgen_CAB
« on: December 01, 2018, 12:54:26 PM »

I must say I enjoy it all... a better AI will mean that I can add some interesting AI NPR in my multi-faction Earth games and get an even better experience.

Must say I'm very excited for what this new version have in store for us, the new AI seem interesting and I hope Steve have the stamina and will to implement it the way he envision it.
Posted by: Kaiser
« on: November 30, 2018, 01:18:09 PM »

How far are you going to go with the new AI routines? One thing that (in general) annoys me with AI is the "All or Nothing" approach. Once an AI has set itself the goal to conquer an object or destroy a certain target, it will not deviate from that if initial plans fail. It just begins pumping more and more ships and resources into the set goal; rather than a human player who would at some point see, that he looses his long term goal, when he continues that way.
Also in ship-to-ship combat. At some point human players realize that it would be better to withdraw from a certain position and take the loss of a space station or a planet in order to preserve the fleet, which could be very usefull having it in a different combat. Meaning, it would be worse if there is no fleet but still have that one planet or space station.

Will the C# AI be able to alter its goals depending on such circumstances?

The AI will assess the situation and determine if an attack is a good idea. If the AI does not believe it can win a fight, it will retreat or not engage. It will start withdrawing unarmed ships from systems near any threat and if hostile forces are in the system it will choose the least threatened route to run (one AI function is actually called PlanEscapeRoute). If no good way out of the system exists, the AI will move to the cover of a defended population if possible.

If the AI chooses to fight, it will deploy accordingly. A single hostile ship might only attract the attention of a destroyer squadron while any NPR battle fleet remains at a strategic location. The AI will stage forces before launching a major attack and will not approach piecemeal. It will also decide if certain locations need protection and if that is more important than attacking. The major difference for C# is that there are Race, System, Fleet, Ship and Population AIs that exchange information and make decisions at different levels. In VB6, each fleet has its own directives and these are often not coordinated (although escorts will attempt to protect other ships for example).

I haven't coded this portion yet but the AI will also learn about your ships through observation and estimate their threat level, strength and defences based on that observation. Those estimates will influence its actions.

Although it won't be a good as a human, the C# AI should be significantly better than in VB6.

That's what I'm expecting more from C Aurora, not plenty of new weapons/race/options which are useless with a boring AI, but a better, tactical, strategic and dynamic AI as possible which makes a game funny and unpredictable. 
Posted by: Steve Walmsley
« on: November 30, 2018, 06:59:14 AM »

How far are you going to go with the new AI routines? One thing that (in general) annoys me with AI is the "All or Nothing" approach. Once an AI has set itself the goal to conquer an object or destroy a certain target, it will not deviate from that if initial plans fail. It just begins pumping more and more ships and resources into the set goal; rather than a human player who would at some point see, that he looses his long term goal, when he continues that way.
Also in ship-to-ship combat. At some point human players realize that it would be better to withdraw from a certain position and take the loss of a space station or a planet in order to preserve the fleet, which could be very usefull having it in a different combat. Meaning, it would be worse if there is no fleet but still have that one planet or space station.

Will the C# AI be able to alter its goals depending on such circumstances?

The AI will assess the situation and determine if an attack is a good idea. If the AI does not believe it can win a fight, it will retreat or not engage. It will start withdrawing unarmed ships from systems near any threat and if hostile forces are in the system it will choose the least threatened route to run (one AI function is actually called PlanEscapeRoute). If no good way out of the system exists, the AI will move to the cover of a defended population if possible.

If the AI chooses to fight, it will deploy accordingly. A single hostile ship might only attract the attention of a destroyer squadron while any NPR battle fleet remains at a strategic location. The AI will stage forces before launching a major attack and will not approach piecemeal. It will also decide if certain locations need protection and if that is more important than attacking. The major difference for C# is that there are Race, System, Fleet, Ship and Population AIs that exchange information and make decisions at different levels. In VB6, each fleet has its own directives and these are often not coordinated (although escorts will attempt to protect other ships for example).

I haven't coded this portion yet but the AI will also learn about your ships through observation and estimate their threat level, strength and defences based on that observation. Those estimates will influence its actions.

Although it won't be a good as a human, the C# AI should be significantly better than in VB6.

Posted by: TMaekler
« on: November 30, 2018, 04:11:01 AM »

How far are you going to go with the new AI routines? One thing that (in general) annoys me with AI is the "All or Nothing" approach. Once an AI has set itself the goal to conquer an object or destroy a certain target, it will not deviate from that if initial plans fail. It just begins pumping more and more ships and resources into the set goal; rather than a human player who would at some point see, that he looses his long term goal, when he continues that way.
Also in ship-to-ship combat. At some point human players realize that it would be better to withdraw from a certain position and take the loss of a space station or a planet in order to preserve the fleet, which could be very usefull having it in a different combat. Meaning, it would be worse if there is no fleet but still have that one planet or space station.

Will the C# AI be able to alter its goals depending on such circumstances?
Posted by: Steve Walmsley
« on: November 27, 2018, 01:37:54 PM »

Has loading / unloading of ground units and the implementation of drop pods already been discussed and fletched out (and I can't find it) or is it just in the works still?

There was a discussion and I have coded it. I haven't got around to writing the changes post yet. I haven't coded boarding though.
Posted by: Vivalas
« on: November 27, 2018, 11:22:29 AM »

Has loading / unloading of ground units and the implementation of drop pods already been discussed and fletched out (and I can't find it) or is it just in the works still?
Posted by: Steve Walmsley
« on: November 26, 2018, 04:43:27 PM »

Not sure if it was raised when posted, but for the changes to the abandon overhaul system, shouldn't there be a delay in the penalty being applied?

What I mean is: it doesn't make sense that a ship pulls into the drydock for an overhaul, then suffers 100% of the penalties when it has to move away 30 seconds later to respond to an incident.

I've never had a problem with the 30-day abandon timeframe (except maybe the lack of feedback that it takes 30 days), but that you have to wait 30 days even if you started the overhaul 5 seconds ago.

You could equally argue the penalty should be less if it was close to the end. However, I don't want to have to start tracking all that for the rare occasions when it will be relevant, plus going into overhaul should have consequences.
Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54