AI Discussion (dev branch)

[froggyluv 2130]

53 minutes ago, inlesco said:

How did you arrive to the conclusion that I have a degree?

 Probably something to do with your 

1 hour ago, inlesco said:

I'm speaking from my own experience in Software Engineering and Engineering Management.

 

 engineers usually have degrees unless your speaking as a layman programmer in which case -join the club.

53 minutes ago, inlesco said:

I understand you're disappointed by Arma 3 devs, but you know - you don't have to be an asshole to others because of that, unleashing your frustrations here on the forums

 

Actually no its not disappointment with Arma 3 -its there trajectory and willful drive to move away from AI development for a long time -predating Arma 3. As far as calling me an asshole correct me if im wrong but you just pulled a rather benign post of mine from last june and preceded to mini-lecture me on something ive been long invested it -but felt the need to explain things down to me like i dont get it. Ive never even seen you around here before.

53 minutes ago, inlesco said:

People make mistakes. Mistakes and mismanagements were possibly made. Now, it's probably too late to fix those mistakes for A3,

 

Are you aware of their intentional drive to move away from singleplayer and alongside that AI? Are you aware of them removing things like SP mission campaigns and even just basic AI squads from their games and DLC? How is this a mistake? Its a busines decision -not a mistake. mistakes can be easily cleared up one simple AI Dev Blog -whens the last time we've seen one -how about never. So please, its disingenuous to say it was all just a small mistake

53 minutes ago, inlesco said:

but it might not be too late for Arma Reforger.

 

no it may not be to late but again if your a regular on these forums ive stated this exact same hopeful thinking and well wishing to the devs multiple times upon release. So again , maybe you should think twice before just reposting a small post I made over half a year ago and think you understand my motivation or some lack of knowledge on the subject -thats arrogance

To finish - even tho again they are showing signs of neglecting AI (as per the dropped promised AI squad command for reforger) -i still have some hope as they have broken down the old Arma AI which contained to many inherent limitations and spaghetti code, but their lack of any communication always on this one issue alongside this should give cause for concern and we all have the right to verbalize this in the manner we think it warrants

[oldy41 61]

Sadly, I believe froggyluv has summarized the AI situation much too good. 😞

Personally I have only little hope to get a good AI for Reforger.

But I still hope BI knows or recognizes that a major part of their fanbase actually prefers SP or Coop over PvP, and consequently invests a lot of love and effort into AI for Arma4.

In any case I hope the new AI framework of Enfusion will provide all the flexibility to allow AI modders to fill in any gaps.

 

Since the first demo for OFP the series has been my absolute favourite game series. So I put a lot of hope and expectations on A4.

 

[mickeymen 322]

On 1/27/2023 at 10:30 AM, inlesco said:

Don't be too surprised - since Arma Reforger is already public, it is currently a top priority for AI engineers because a new AI framework has to be built, tested and released to the public to keep Arma Reforger fresh and constantly improving as a sandbox.

Don't be too surprised, but Arma Reforger against the background of Arma3 content, which the community and BIS has been creating for 10 years, is simply unusable material. It is not able to give its player even 10% of what Arma 3 can give. Basically, Arma Reforger - just a test.

 

On 1/27/2023 at 10:30 AM, inlesco said:

Forom the software engineering perspective, it does not make much sense to invest time of Senior or Lead AI Engineers / Architects into a 10-year old game because Enfusion Engine's AI modules need a lot of attention.

And who told you that you need to invest "senior engineers"?
I said oh 1-2 people and these people do not have to be "Senior or Lead AI Engineers"

 

On 1/27/2023 at 10:30 AM, inlesco said:

Enfusion's AI is now where all the action is - it's the right opportunity to make things right from the start.

to make things right from the start and wait at  decades? 

Personally, I don't want to wait decades, I'm satisfied with old Arma3, which will still be playable for at least 10 years, if this game is supported by the developers and by mod-makers.

As far as Enfusion is concerned, I'm not against it developing, I'm for it!

But while this is still being developed, I and other fans of the game would like to have a stable and supported Arma3.

 

On 1/27/2023 at 10:30 AM, inlesco said:

Unless Enfusion's and Real Virtuality's AI parts are highly reusable. Does anybody have insights on this - on code or engineering effort reusability between the 2 engines, specifically?

There are a lot of developers who wrote AI for Arma3, these people could not lose their abilities in one second, after the release of Reforger.

 

On 1/27/2023 at 10:30 AM, inlesco said:

I'm thinking - it is best for BIS to probably make a framework for AI control and configuration now. And include that into Enfusion Workbench (official dev tools) as a visual AI behavior editor.

Maybe, but Arma3's AI shouldn't be forgotten, at least if you as a developer promise to support it in 2023

 

On 1/27/2023 at 10:30 AM, inlesco said:

Similar in concept to Unreal Engine's Blueprint -> https://www.youtube.com/watch?v=bY6Nl-OEhSo

Has nothing to do with what I said. This is another topic

[Black__Knight 0]

Is it possible that Bohemia or anyone bothers the AIs can use step over or enhanced movement feature? Because AI if they know how to use step over feature, they can unstuck themselves in some situations or cross the fence easily without looking for a break hole 500m away

[Wombat72 0]

Put this in the ARMA3 forum as well but it's probably better here.

 

Good morning everyone.

I work at a weapons range that uses VBS3 as the basis for all our scenarios and I'm trying to re-author and update one or our All Arms Call for Fire scenarios to add more realism and I have struck a problem.

So the basic premise is that we use a 2 system network, one runs the maps and scenario and then a client logs in with Fires running to direct artillery from a bank of 155's, 105's and M80's that are already placed in the scenario.

So the issue I'm striking is that while I have all the objects on the field I cannot get the client artillery to actually cause any damage to the buildings that I have placed. Any vehicles or personnel hit are destroyed but not any of the structures. Even after using the "this setDamage 0.95" command to bring building health to practically 0, I cannot get them to be destroyed using the in scenario artillery.

If I manually call in an artillery strike using the command menus It will knock them over pretty quickly though.

So I guess the questions I have are:

How do I get the buildings to actually take damage from the client, and

How do I ensure they are destroyed when their health reaches 0.

 

Thanks for any help I can get, I'm really starting to pull my hair out over this.