CMO Air Combat Philosophy (from .37 thread) (Full Version)

All Forums >> [New Releases from Matrix Games] >> Command: Modern Operations series >> Tech Support



Message

boogabooga -> CMO Air Combat Philosophy (from .37 thread) (12/1/2021 12:44:25 PM)


quote:

ORIGINAL: boogabooga

quote:

ORIGINAL: michaelm75au

The IPC has always been changed to 'yes' on patrol missions in order for it to investigate contacts. Changing it to 'no' would presumably make it ignore contacts.
If there is a specific issue here, please raise a defect in Support for us to look at.


I'm not trying to sound aggressive, but there is a general issue here, in that the community does not always understand the developer's philosophy on such things (and vice versa) and mismatched expectations end up being micromanaged in tech support. For example, is it in the manual that the game engine changes the IPC setting on it's own? Also, I don't necessarily agree with the statement that " "Ignore plotted course" really only has meaning to Patrol mission." I suspect others don't as well, which is why my being nervous got a few +1s. Different expectations in the community.

What happens with a manual plotted course then, when the A/C is not assigned to ANY mission? Should IPC be obeyed?


quote:

ORIGINAL: Dimitris

I think this is better discussed on a dedicated thread on the tech Support forum.


Here you go.

Please elaborate more on how you expect air combat missions to work in CMO and the purpose and limitations of the IPC doctrine setting, etc.



Dimitris -> RE: CMO Air Combat Philosophy (from .37 thread) (12/1/2021 12:50:44 PM)

This point originated here: http://www.matrixgames.com/forums/fb.asp?m=5108026



RoryAndersonCDT -> RE: CMO Air Combat Philosophy (from .37 thread) (12/1/2021 4:13:08 PM)

Perhaps the best way to look at this issue would be to look at some example scenarios or some illustrative examples? Examples of how this change would be used 'in anger' so to speak.

We're definitely not trying to break existing scenarios and behavior or anything like that, but we don't have an all-seeing-eye to be able to see and understand every logical (and unintended) result from our changes.

In short: we have good intentions when we make changes, but don't know everything and could be miscommunicating or under communicating.


My understanding of the full extent of this change is that IPC doctrine setting is ignored for Patrols so a unit on a patrol ignores its randomly assigned course on a patrol to investigate contacts?

I haven't checked it out though in Command as I'm working on a different branch and I don't want to potentially screw with my current task



DWReese -> RE: CMO Air Combat Philosophy (from .37 thread) (12/1/2021 4:51:09 PM)

Rory,

I believe what the others are stating is that the CMO dev team has made the announcement that in Version 37 some changes pertaining to Patrols and the IPC option were made, and the gamers would like to know what they might expect from those changes, and how they may potentially disrupt any of their previous efforts with scenario design. Not really knowing what was changed is what is causing the concern, not the fact that anything was changed. It's really boils down to the question of what can be expected now because of the changes? What's different than before?



boogabooga -> RE: CMO Air Combat Philosophy (from .37 thread) (12/2/2021 12:09:04 PM)

quote:

ORIGINAL: RoryAndersonCDT

Perhaps the best way to look at this issue would be to look at some example scenarios or some illustrative examples? Examples of how this change would be used 'in anger' so to speak.

We're definitely not trying to break existing scenarios and behavior or anything like that, but we don't have an all-seeing-eye to be able to see and understand every logical (and unintended) result from our changes.

In short: we have good intentions when we make changes, but don't know everything and could be miscommunicating or under communicating.


My understanding of the full extent of this change is that IPC doctrine setting is ignored for Patrols so a unit on a patrol ignores its randomly assigned course on a patrol to investigate contacts?

I haven't checked it out though in Command as I'm working on a different branch and I don't want to potentially screw with my current task


Hello Rory,

Really, the reason that I started this thread is the opposite; to be a top-down conversation rather than a bottom-up. FWIW, I (and others) have a few specific bug reports in the que regarding A/C course keeping issues.

As DWReese alluded to, my concern is not that I have a specific problem but that I don't know if I do. The fact that a little bombshell like "IPC only makes sense on a Patrol mission" was dropped in one line as the third bullet point on the 37th patch makes me think that we are on really different philosophy pages. Specifically, I feel very, very strongly that any intentional human intervention needs to be able to completely and reliably over-ride anything that the AI is doing at any time. Under that assumption, IPC is a critical concern, not a minor detail of a certain mission type. Whatever has changed, was it meat to make the entire simulation more hands-off or more hands-on? That's the kind of question I am talking about when I say top-down "philosophy."

Here is another "philosophy" question...what exactly do you mean by "plotted course" in the first place? Or more specifically, how is the plotted course intended to interact with the altitude and throttle settings? In the UI, these are two distinct entities, with entirely different menus (F2 and F3), etc. But in practice, they seem not so distinct. Ignoring the "plotted course", will get units to automatically change their altitude and throttle as well. Since a few patches ago, even the presence of a plotted course in the latitude/longitude sense might affect whether a unit changes its altitude/throttle. None of that is necessarily a problem, but do you understand how this might seem inconsistent with the UI and lead to confusion? If you ignore the plotted course, will you ignore the set speed and altitude as well? Whatever the answer is right now, is that what you intended?



RoryAndersonCDT -> RE: CMO Air Combat Philosophy (from .37 thread) (12/2/2021 7:36:49 PM)

The change in the code that generated that line in the changelog is a modification to 2 lines of code, it is truly a minor minor tweak.

Edit: We will discuss the change internally.



KnightHawk75 -> RE: CMO Air Combat Philosophy (from .37 thread) (12/7/2021 9:36:26 PM)

I'm late to chime in, but ditto #4 and #5 and few cents more.

quote:

Not really knowing what was changed is what is causing the concern, not the fact that anything was changed. It's really boils down to the question of what can be expected now because of the changes? What's different than before?

This.   Though depending on what the case might be maybe there is concern, but we gotta know to evaluate it. For instance when reading the release notes I recalled having worked on a scene and related code that made use of an overall patrol mission to host some aircraft, at some point it assigned a specific course table upon a patrolling unit directing it toward a target on the path I designated\generated to eventually manually target\attack the desired target without removing said aircraft from it's over all assigned patrol mission and to my memory it was troublesome till I flicked off IPC. It only worked right consistently in the scene because IPC=no setting was being obeyed, such that no matter what else it was being instructed to do next (target someone else, investigate, whatever) it first was going to follow my explicit navigation instructions. Is that affected by the change, it was unclear from the initial release notes.

Adding at least to my confusion was that "it's always been this way for patrol missions". Which is not what I recall for a scene (granted scene I'm was thinking about I last touched like a year+ ago) or at least not in _all_ cases, the way I recall it in the past was the aforementioned recollection. Now I can work around either behavior so long as I know I need too. I can also understand that it's possible that 'it's been this way for awhile' and that in the distant past some scene only worked they way it did in my recollection due to some other condition\bug\whatever in the build at that time, and that somewhere along the way was 'fixed' such that it would appear now to always worked that way.


Anyway I'm still wondering if IPC on patrol mission/members has effectively no meaning at this point in time (be it new or preexisting). As I understand it 'post-fix' it's supposed to still have meaning for NON-patrol missions.

quote:

Specifically, I feel very, very strongly that any intentional human intervention needs to be able to completely and reliably over-ride anything that the AI is doing at any time.

I very, very strongly concur with this opinion [:)]

With that in mind I question the value of patrol missions changing the setting on the player (or player programed ai-side) so that it can go investigate or engage (no matter if that behavior is actually new or existing for awhile). If I've assigned a manually-plotted course AND IPC=false, my expectation is it's going to obey my setting. To the detriment of responding to contacts\it's mission if needed, at least from a navigation stand point (which given boogabooga's comment I consider to include speed+alt since it's settable for each waypoint). Now I can imagine that might mean people are going to forget they have both enabled, and wonder why there is what seems at a glance no response or a delayed response, but that's on the player for forgetting what they set. I'm not immediately seeing (and I accept I could just be blind to other issues, complications, or maybe missing something obvious) what makes patrol missions so special that compels the users setting to be ignored in their case.



boogabooga -> RE: CMO Air Combat Philosophy (from .37 thread) (12/8/2021 1:59:03 AM)

My impression is that the developers generally intended for the missions to be simply set up and left to run autonomously, while direct human player control works best on unassigned units that need to be micromanaged in all respects from start to finish.

Actually, there is a third play style involving heavy human intervention in the pre-planned missions to try to make them more intelligent. That seems to be where many of the issues are, and I am not sure of the philosophy as to how plotted course contradictions between the loadout, the mission, and the player get resolved in all cases.



SeaQueen -> RE: CMO Air Combat Philosophy (from .37 thread) (12/8/2021 11:40:15 AM)

quote:

ORIGINAL: boogabooga
My impression is that the developers generally intended for the missions to be simply set up and left to run autonomously, while direct human player control works best on unassigned units that need to be micromanaged in all respects from start to finish.


I think you need to be a little bit careful about that. A lot of people interpret that kind of statement to mean, "If I set up a mission, it will run autonomously exactly the way I expect it to be run autonomously." People get hung up on that.

You need to be very careful about the specific settings of the mission, and even then you need to check on it. I wouldn't expect a mission to always do exactly what I want it to do. I expect it to be a rough solution that I can count on when things aren't too complicated or fast moving, which allows me to devote my attention elsewhere, but I really can't go completely "hands off." People sometimes marvel at how meticulously and consistently I set up my missions. There's a reason I do that. I'm trying to minimize unexpected behaviors and get my intended results (roughly). I don't think people ought to expect missions to consistently produce optimal, correct, or "best practices" solutions. That's for a player to decide. They ought to expect them to do what they're told to do, and if they don't check all the settings to make sure they're telling it to do the right things, then there's no guarantee that they'll do what you intend, because you didn't check to make sure that it did what you intended.

In general, so long as I pay careful attention to the settings, everything works (mostly) great. Usually where things get complex is when there's lots of contacts, or I need them to fly close to the limits of their range, which is sort of an edge case, really.



boogabooga -> RE: CMO Air Combat Philosophy (from .37 thread) (12/12/2021 12:39:17 PM)

Thank you for the input, SeaQueen, but I'm talking about something different. I'm talking about when the game engine itself will not follow what I will call the "HI Over AI Doctrine": that any intentional human intervention needs to be able to completely and reliably over-ride anything that the AI is doing at any time. More specifically, that in some cases human control becomes more difficult/impossible when A/C are assigned to a mission than when they are unassigned.

I posted a specific example here:
https://www.matrixgames.com/forums/tm.asp?m=5114861

This is longstanding (back to CMANO) behavior in which wingmen ignore the human plotted course and keep going on their mission (Even with IPC=NO). That this behavior has persisted so long makes me think that it was never intended for the human player to intervene in the mission. It makes it impossible, for example, to vector the flight behind the threat without unassigning. But there are still benefits to being assigned to the mission: EMCON, Doctrine, WRA, Auto launching, the ability to retarget automatically, tanker settings, for Lua scripting reasons, etc. So, yes, it comes as a surprise when one says that the IPC only makes sense on a patrol, because I can see it being used elsewhere.

Again, it might just be a matter of managing expectations through the UI. If I see all of the interface for plotting courses speeds, and altitudes, then I expect them to work. If human plotted courses are not going to be supported during say, an air-intercept mission- then have a pop-up or something saying so and block the plot. Just like what happens with RTB-Bingo now. But these are the things that I am talking about when I say that the developer's "philospohy" on Air Combat needs to be communicated a bit more.






Dimitris -> RE: CMO Air Combat Philosophy (from .37 thread) (12/13/2021 2:13:18 PM)

Generally speaking, yes, human overrides should trump AI-made directives/decisions. In cases they don't, there needs to be a specific reason why that happens (e.g. fixing a bug that cannot be solved any other way).

On the example you linked to I already explained what is going on and why. Yes, it's a problem in that very specific example that you provided. Should we rush to fix that and break multi-target strike missions? Will you live with that?



morphin -> RE: CMO Air Combat Philosophy (from .37 thread) (12/14/2021 11:13:26 AM)

I can understand boogabooga very good. I too wish you can (temporary) override the mission to take control of the unit, see

https://www.matrixgames.com/forums/tm.asp?m=5023756&mpage=1&key=unassign%2Cmission�

That is actually one of my biggest wishes (beside path issues [:)]) that i have




boogabooga -> RE: CMO Air Combat Philosophy (from .37 thread) (12/17/2021 12:52:07 AM)

Thanks for the link, morphin. That post brings up many of the issues that I was talking about.



Dimitris -> RE: CMO Air Combat Philosophy (from .37 thread) (12/17/2021 5:09:14 AM)

That thread is marked as fixed. Wasn't the problem resolved?



boogabooga -> RE: CMO Air Combat Philosophy (from .37 thread) (12/17/2021 5:32:14 AM)

SteveMcClaire's response was yesterday, 12/16/2021, so it was marked "fixed" only very recently.



jannas34 -> RE: CMO Air Combat Philosophy (from .37 thread) (12/21/2021 6:58:20 AM)

Not likely anything will get done about this



Page: [1]

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
0.75