Robots should understand combat drops.

[Adamo]

We should make a definition for what is a combat drop, such as "is a request for weapons" and "has aliens in adjacent chunks" or "is near the player" -- something like that. Given the conditions, the robots know to coordinate such that they make drops simultaneously, rather than piecemeal.

[immibis]

What?

[Adamo]

What?

What isn't clear? I'm pretty sure this explanation will be clear to any technically minded people, most importantly anyone who develops the robot code.

[Ranakastrasz]

It is very much not clear.

After reading it four times, I have a theory on what you mean.

You are implying that Robots should not be willing to travel outside for supplying the player or similar if there are Biters around, to avoid them dying.

[Adamo]

You are implying that Robots should not be willing to travel outside for supplying the player or similar if there are Biters around, to avoid them dying.

Err, no. I mean exactly what I said. Set conditions for what a battle field is. If those conditions are observed by the bot network, then the bots should drop their cargo in groups, rather than one-at-a-time. Combat drops. I'm not sure how I can be more clear, but again, I am confident that any software engineer will get this.

[ssilk]

What should happen, if you order such a horde, they begin to gather and to start their drop, they need to wait for one bot, which 5 minutes away?

[DaveMcW]

Err, no. I mean exactly what I said. Set conditions for what a battle field is. If those conditions are observed by the bot network, then the bots should drop their cargo in groups, rather than one-at-a-time. Combat drops. I'm not sure how I can be more clear, but again, I am confident that any software engineer will get this.

I am a software engineer and I have no idea what this means. Combat logistics is not part of any course or job I have taken.

[Adamo]

I am a software engineer and I have no idea what this means. Combat logistics is not part of any course or job I have taken.

Hey. Let me make sure I have this straight. You are a software engineer, but you say you do not understand this statement:

We should make a definition for what is a combat drop, such as "is a request for weapons" and "has aliens in adjacent chunks" or "is near the player" -- something like that. Given the conditions, the robots know to coordinate such that they make drops simultaneously, rather than piecemeal.

That is concerning to me. Any trained software engineer should understand the basics of establishing conditions and making decisions based on those conditions. I have my doubts about your claim. It seems you are trying to object to my suggestion, or show support for others' objections, for erroneous reasons.

This all being said, I will try to explain more fully. We make a definition for what a battlefield is. I put my suggestions above, but this is open to debate. Bottom line, there is some set of conditions that tells a bot "this is a battlefield." Whenever a bot knows it's entering a battlefield, the bot knows to join with other nearby bots so that they drop their cargo simultaneously, in a group, rather than one at a time. This is the preferred method for making cargo drops in a combat zone, and is known to be much more effective than a steady stream.

[Adamo]

What should happen, if you order such a horde, they begin to gather and to start their drop, they need to wait for one bot, which 5 minutes away?

We should set a threshold so that they meet reasonable expectations for group formation. It wouldn't make sense for a bot to group with another bot that is on the other end of the base.

[keyboardhack]

Why is this needed. Is it not preferable to get the items when they are ready instead of in bulks?

[Adamo]

Why is this needed. Is it not preferable to get the items when they are ready instead of in bulks?

A steady stream is not preferable. In a combat situation, group drops are preferable for several reasons. Perhaps the primary reason is that it gives the enemy more targets, so while some bots will die, most will get to the drop, whereas if they stream in one at a time, it's likely that all of them will be destroyed before they reach the drop. This is a real world tactical fact, but also applies to Factorio bots. Another reason is that it's better to receive a stack of magazines at once, rather than one at a time. Technically, one could wait until they have a stack, but that may happen more slowly with a steady stream because of the first problem, and it also just never seems to work out that way.

Group drops -- combat drops -- are widely recognised to be superior in a battlefield situation. And the implementation of this is not overly complex.

[Martc]

it's better to receive a stack of magazines at once, rather than one at a time.

If you have no ammo, you will be grateful for every magazine you get. Bitter will not wait until your factory produce whole stack of magazines.

They are called Logistic drones, not Combat logistic drones, so they are designed for logistic not for combat situation.

Group drops -- combat drops -- are widely recognised to be superior in a battlefield situation. And the implementation of this is not overly complex.

I think that main problem is not about implementation, but about finding consensus how should they behave and find time to do it

[Adamo]

If you have no ammo, you will be grateful for every magazine you get. Bitter will not wait until your factory produce whole stack of magazines.

If the bots do a steady stream you will receive less total ammo than if they do groups, because of the death problem I laid out in a previous post. With group drops, less bots will be destroyed, and you will be thankful that you end up with more ammo.

[Khyron]

Hey. Let me make sure I have this straight. You are a software engineer, but you say you do not understand this statement:

We should make a definition for what is a combat drop, such as "is a request for weapons" and "has aliens in adjacent chunks" or "is near the player" -- something like that. Given the conditions, the robots know to coordinate such that they make drops simultaneously, rather than piecemeal.

That is concerning to me. Any trained software engineer should understand the basics of establishing conditions and making decisions based on those conditions. I have my doubts about your claim. It seems you are trying to object to my suggestion, or show support for others' objections, for erroneous reasons.

Honestly, your suggestion is confusing. You posted a solution to an undefined problem and gave no context or example for how it solves the problem. The question is why do you want this? There is a saying "that is a solution in search of a problem".

After reading the rest of this thread, my best guess is you're talking about having constructor bots place a power pole and laser turrets as part of a turret creep strategy. Your solution would seem to have the most effect in that scenario. Personally, I hope no changes are made that encourage the use of that strategy because I think turret creep is lame. I'm also pretty sure the devs are planning to implement changes which discourage or outright prevent that strategy. Search for a thread on turret creep to find out more.

If it's not that, then I don't know what you're on about. The usefulness of your suggestion in pretty much any other scenario (resupplying the player with ammo, or turrets with ammo) seems pretty minimal to me. There are complications to this like what ssilk said, and each complication stacks the scale against implementation. By that I mean - the perceived improvement to Factorio on the whole vs. the time spent making it. There are plenty of other suggestions vying for the developer's time. Why is this one important?

Also with that quote above, your response is pretty poor (almost rude) given how badly you explained the problem in the first place. You obviously had a idea in your mind but you need much more practice learning how to communicate that to other people. If you want your ideas and suggestions to be implemented, the burden is on you to explain it properly - not on the software engineer to waste time guessing what you want.

[Adamo]

If it's not that, then I don't know what you're on about.

It's not that. What I'm on about is suggesting that the bots be given a simple "combat awareness." They should understand combat drops. My suggestion is nothing more than what is written in the first post.

The usefulness of your suggestion in pretty much any other scenario (resupplying the player with ammo, or turrets with ammo) seems pretty minimal to me.

The usefulness is that bot survival rate and drop:death ratio are significantly improved when delivering objects to areas that have aliens. Why do you need a specific instance? This is a general suggestion.

There are complications to this like what ssilk said, and each complication stacks the scale against implementation.

You guys are imagining complications that don't exist in my original, extremely simple, suggestion.

By that I mean - the perceived improvement to Factorio on the whole vs. the time spent making it. There are plenty of other suggestions vying for the developer's time. Why is this one important?

Yes, good question. It's not very important compared to even my other active suggestion, although without this the bots are pretty terrible around aliens (essentially just slow-streaming themselves, and their cargo, into oblivion). If you want to prioritise, I recommend looking at my other suggestion regarding bots, which is more pressing, and also simpler to implement.

If you want your ideas and suggestions to be implemented, the burden is on you to explain it properly - not on the software engineer to waste time guessing what you want.

I would prefer if you guys stopped guessing what you think I mean and look at what I said. I am suggesting this:

We should make a definition for what is a combat drop, such as "is a request for weapons" and "has aliens in adjacent chunks" or "is near the player" -- something like that. Given the conditions, the robots know to coordinate such that they make drops simultaneously, rather than piecemeal.

This is the suggestion. It is very straight forward. If you guys don't understand the usefulness, OK, but I am hardly concerned with that. There are infinite strategies by which this game may be played, and just because you don't feel the strategies you know will benefit, doesn't mean that others, or even yours, won't. I assert that any strategy will benefit from having robots that handle themselves just a bit better in combat zones.

[Khyron]

They should understand combat drops.

Why?

[Adamo]

Why?

So they die less and more drops get through when they're under attack.

[Khyron]

So they die less and more drops get through when they're under attack.

My drones never get attacked. Why are your drones getting attacked?

[Adamo]

My drones never get attacked. Why are your drones getting attacked?

This suggestion only applies to bots that are in a combat situation. It has been stated that, among alien behaviors, one is that an alien will switch its attention to bots, turrets, and players when they encounter any of these things. If your bots are not being attacked, it's likely because your bots are not in range of any aliens. In that case, this does not apply to your bots.

[Khyron]

This suggestion only applies to bots that are in a combat situation. It has been stated that, among alien behaviors, one is that an alien will switch its attention to bots, turrets, and players when they encounter any of these things. If your bots are not being attacked, it's likely because your bots are not in range of any aliens. In that case, this does not apply to your bots.

Bots should not be in combat situations. That is part of the challenge of the game.