Oolite Bulletins

Edit: that came out wrong. I meant I'm no smarter than Disembodied.

Is it Friday yet?

Diziet Sma wrote:

Anyways.. She remarked that most modern games with controls of this level of complexity, usually have a tutorial/training mode which teaches the player about the controls and how they function, and walks them through the first stage or two of the game so they can get a feel for things.

Well, some have a formal tutorial; others just start the player off in a situation where much of the functionality is unavailable/unnecessary, and introduce it as the game continues. I tend to find the "tutorial" approach leaves me very bored as they always seem very slow-paced (but there's going to be one crucial bit of information in there, so I don't want to risk skipping it either...)

An in-game manual for the interfaces screen might be a good idea, though, with a couple of quick-start pages, and maybe some "tips of the day".

Diziet Sma wrote:

Actually, the remark about the "complexity" of Oolite's controls is one I've heard from a number of members of the Playstation Generation

It certainly has a lot of controls, to the extent that for new functionality requiring a control the most difficult question is not "is this in the spirit of Elite" but "is this really a good use for one of our few remaining keys". As far as complexity goes, though, you only need about ten keys for normal play (up, down, left, right, faster, slower, laser, target, missile, torus) - most of the rest either relying on equipment that's not fitted or not being necessary for a first trip (and okay, you'll need to press 'hyperspace' once too). So I think there's a decent introduction curve for the controls.

cim wrote:

Well, different approaches for different people... I always (Elite and Oolite) went straight for Leesti-Diso, and then generally left the Old Worlds heading north through Reorte, then east through the Xexedi cluster to reach the rest of G1, or continue north to the worlds around Teorge for a bit - and by then I had a ship good enough not to need that safe a route. I only "discovered" Isinor-Ensoreus after reading a chance remark on this forum.

That's what I did in Elite as well.
It simply seemed logical to me - especially since they are rather close to each other.

cim wrote:

Mad Dan Eccles wrote:

It is hard for a noob, yes.

I think what we have to ask though, is that a bad thing?

Not as such - but the other side of it is that it gets very easy relatively quickly too. Even without particularly optimal decisions, you can probably have a full iron-ass (minus the mission-related equipment) by the time you're Dangerous ... and then you are almost never at risk, unless you add in difficulty-increasing OXPs. Really, "fore military laser + injectors" is enough to deal with almost every pirate pack in the core game.

I've been thinking about this. You're right, of course - and it's the prime engine driving a lot of new ships, and more powerful ships, and then yet more powerful ships, and so on. It would take a lot of work, but it might be possible to rebalance things. Your own [wiki]Skilled NPCs OXP[/wiki] helps here, making some of the bad guys smarter as well as better shots. To rebalance the whole game would mean using things like this to alter the types, behaviours and formations, of pirates encountered in various systems.

Like you say, at the moment, for a skilled player "fore military laser + injectors" (plus a modicum of fuel) is enough to cope with most pirates, or groups of pirates, you're likely to meet. But if it was assumed that, down the sliding scale of government types, you were likely to meet only small, desperate, lone-wolf bandits at the top end, people driven into piracy by circumstance, who are usually not very good at it; and put big, strong, Navy-defying packs, made up of skilled, well-equipped pirates at the bottom end, then the game could be made playable, and challenging, to players at a range of abilities. Corporate States, Democracies and Confederacies would be the shallow end, and Feudal and Multi-Governments would be deep waters. Anarchies would be deep, shark-infested waters. The Dictatorships and Communist worlds would hover in a grey zone in between. New players, and players still developing their skills, could splash around in the shallows, before venturing out into deeper waters. The game's geography would start to matter more, and route-planning would be more of a skill.

It would mean that players could make a lot of money, relatively quickly - but then they can do that as it stands, by leaving the spacelanes. And in any case, money doesn't matter, ultimately: in the game, it's more reputation, and kill rankings, that count, and to get those players would still have to get good at fighting, iron ass or no.

I just had an encounter with a Mamba escort, nothing too advanced, 9cr bounty, and yet he managed to take off half of my fore military shields in a matter of seconds!

another_commander wrote:

Regarding missiles: At the beginning they are indeed very difficult. However, we should not forget the tactic of launching one missile at them if they launch one at you, hoping that they will try to ECM it, destroying their own missile in the process. I've done it plenty of times and it works. However, I admit that although it generally helps, it is not a complete mitigation of the risk.

I'm of the opinion that few pirates would have missiles by the time the player sees them. If they've been lurking in the space-lanes, then they probably used them up on the first trader convoy they see.

We should see more pirates lurking just outside the normal shipping lane, either letting their shields and energy regenerate from their last fight and/or hoping to catch those trying to torus directly to the main station. Likewise, they should usually flee in that direction if losing.

Stable government systems could probably have a 'deadzone' within ~10 km of the line from the witchpoint buoy to the planet where pirates are not placed. They'd still appear on your scanner range if you're traveling down the exact middle of the shipping lane if they're within ~25 km of it, but just as likely they could be up to 50 km out. That's about the only way pirates could sensibly operate in a Corporate State system.

Most pirates (especially lone ones) should probably flee after just a short burst of hits from a pulse laser, if those hits come quickly enough to reduce them below half energy. Which is why they need shields also! Using a missile on a target means zero profit, so should definitely remain a last resort. They probably have few options of restocking missiles -- even if there's a Rock Hermit Asteroid willing to sell them missiles, the price is so much higher than the main station that it's almost piracy.

A question regarding NPC AI: do ships in groups (either pirate packs, or trader convoys) ever have any sort of group AI? Is such a thing even possible? Can packs of ships act intelligently as a pack, in other words - sending a couple of fast ships off to outflank and attempt to surround the player, for example? Can pirates "call in" other nearby pirates? (This seems to be something that'll happen in Elite: Dangerous; protracted combat will have the increased risk of attracting outside attention.)

Would this even need a "group AI", anyway, or could something similar be done with individual AIs for different types of ships, which took note of nearby "friendly" ships?

Great question. I'd like to know that, too.

They certainly don't appear to have any sense of pack-based self-preservation or strategic awareness. If I were a pirate in a pack flying, say, a Mamba, and I'd just seen my 'prey' wipe out five or six of my buddies without apparently breaking a sweat I wouldn't hang around to see if s/he had any fight left ...

There is some group awareness in AI, but it's pretty basic. One of my plans for the 1.79/1.80 releases is a new Javascript-based AI engine, which should make much more sophisticated AIs practical to write.

At the moment:

• ships in a group will sometimes come to each other's assistance in a fight, treating hits on each other a bit like hits on themselves for purposes of "should I switch target" questions. For escorts the targeting decisions are mostly directed by the mothership; for general groups it's much more individual choice.
• the initial decision of a pirate group as to whether to engage is based in part on the relative size of the pirate pack and the trader pack. But it's size only, nothing to do with armaments - and once combat is underway, it's never reassessed.
• at accuracy >= 5, there are some group tactics as a consequence of individual tactics. For example, if two pirates are fighting a lone trader, the pirate who is being targeted by the trader will try to close rapidly to dogfighting range, making it difficult for the trader to consistently target them. Meanwhile, the other one will hang back, firing on them from longer range where their evasive turns are less effective, trying to distract the trader from their colleague. These tactics are technically all individual tactics, so they don't require the attackers to be formally grouped or even aware of each other, and they work okay even if there's just one pirate, but much better with multiple ones.
• pirates can't officially call in other pirates from further away, but in a protracted fight in a dangerous system the steadily spreading debris field seems to have much the same effect. I've certainly had plenty of occasions in anarchy systems where a "simple" fight against three or four pirates has turned into a running battle against about twenty.

I'd rather not have group AIs as such - but that's only an implementation detail; an individual AI which takes group strategy into account and sends commands to other nearby individuals would have much the same effect.

Thanks, Cim - that sounds pretty sophisticated as it is! But the future plans sound really interesting ...

cim wrote:

I'd rather not have group AIs as such - but that's only an implementation detail; an individual AI which takes group strategy into account and sends commands to other nearby individuals would have much the same effect.

I think you're right. Not that I know squat about AI, but I can appreciate that an individual intelligence which is aware of its surroundings is both more elegant, and more flexible, than a rather kludgy "group mind".

Disembodied wrote:

Like you say, at the moment, for a skilled player "fore military laser + injectors" (plus a modicum of fuel) is enough to cope with most pirates, or groups of pirates, you're likely to meet. But if it was assumed that, down the sliding scale of government types, you were likely to meet only small, desperate, lone-wolf bandits at the top end, people driven into piracy by circumstance, who are usually not very good at it; and put big, strong, Navy-defying packs, made up of skilled, well-equipped pirates at the bottom end, then the game could be made playable, and challenging, to players at a range of abilities. Corporate States, Democracies and Confederacies would be the shallow end, and Feudal and Multi-Governments would be deep waters. Anarchies would be deep, shark-infested waters. The Dictatorships and Communist worlds would hover in a grey zone in between. New players, and players still developing their skills, could splash around in the shallows, before venturing out into deeper waters. The game's geography would start to matter more, and route-planning would be more of a skill.

I really like the basic idea that's behind that, but I'd prefer if the safety of a system would not depend on the type of government too much but only act as a rule of thumb. Saying that all anarchies are by definition dangerous places and corporate states are safe is, in my eyes, a prejudice. There might be anarchies whose inhabitants are willing to help and protect each other and visitors, too, making the system safe. There might be Dicatorships or Monarchies who have a wise and benevolent dictator / king who protects everyone from crime in his system. There might be failed democracies whose corrupt leaders look in the other direction when their favorite crime lords pick on some unsuspecting traders, as long as they don't overdo it. The leaders of a corporate state might for some reason have come to the conclusion that they gain more profit with less security, or they might be in a similar state like the corporations in the game Syndicate, which would mean that the ordinary person is not safe at all.
Therefore, I'd prefer if the safety of a system is not tightly coupled with it's government, but that the government only influences the "base security" of a system, with Disembodies' suggestions as a guideline. On the long range chart, the player should be able to see how safe the systems currently are. His actions (and those of other ships, of course) might even influence the safety rating of a system - if, on a trip, he destroys a pirate base, the systems' safety rating might rise. If, on the other hand, some Thargoid attack wiped out half of the systems police force, the safety rating might fall.

GGShinobi wrote:

I'd prefer if the safety of a system would not depend on the type of government too much but only act as a rule of thumb ...

I agree that there's perhaps an element of prejudice involved but this can be covered by assuming that these definitions are handed out by whoever writes the planetary guides, and they are intended purely as advice for pilots on what to expect in space. A planet classed as an Anarchy may be a delightful place to live, just as a Corporate State might be a hellish one - but for whatever reason, the Anarchy system doesn't supply sufficient funding to GalCop for extensive police patrols in the system, and therefore a lot of pirates use the local volume. The Corporate State, meanwhile, extracts lots of taxes from its groaning population and pays for plenty of Vipers to keep the trade routes running smoothly. The definitions are only meant as guidance for traders as to how much pirate activity them might expect to meet.

The problem with creating dangerous Corporate States and pirate-free Anarchies is that it would tend to blur all system types together, and the game might lose any sense of variety - especially if this was just done via dice-rolls, so that you might get lots of pirates in Corporate State X one trip, and none the next. If you want safe Feudal systems and dangerous, corrupt Democracies, they'd have to be hard-coded as such, and described as such in the F7 screen, to keep the sense of variety. And if they were hard-coded as "safe" or "dangerous" then the planetary guide might reasonably reclassify them according to the likelihood of encountering pirates. The safe Feudal system run by the good, wise king might be classified as a Corporate State thanks to his efforts in fighting crime and protecting trade, and the corrupt Democracy might be reclassified as a Dictatorship because they've cut a deal with pirates.

Disembodied wrote:

so that you might get lots of pirates in Corporate State X one trip, and none the next.

Well, I wouldn't want there to be no chance of this, either. A well-armed pirate band might well risk an occasional raid into a corporate system, where the traders think they're safe and the right strategy could distract the cops for the crucial five minutes needed to blow up a convoy, scoop the goods, and retreat into witchspace. It wouldn't happen very often (and should probably happen less often than it currently does...) but it should happen occasionally.

Narratively, perhaps more likely in some systems than others. So Ceesxe (G1), for instance, has few dangerous systems around it, and is probably consistently pretty safe. Xevera (G5) on the other hand ... well, there's a reason that region is called the Siege Worlds.

cim wrote:

ships in a group will sometimes come to each other's assistance in a fight, treating hits on each other a bit like hits on themselves for purposes of "should I switch target" questions. For escorts the targeting decisions are mostly directed by the mothership; for general groups it's much more individual choice.
...
I'd rather not have group AIs as such - but that's only an implementation detail; an individual AI which takes group strategy into account and sends commands to other nearby individuals would have much the same effect.

Group AIs at least in the sense that ships that are grouped together work ok together is currently an unfortunate necessity. The biggest problem currently with getting a "large convoy" of 10-15 ships heading from witchpoint buoy to system station is:
1.If they're grouped, they won't fly in formation without a lot of scripting that I don't care to dirty my hands trying to do. The AI logic inside the inaccessible-to-scripting/AI changes part of the core code is probably a better place to deal with that, especially with already-exposed reprogrammable formation definitions.
2.If they're not grouped, even setSpeedToCruiseSpeed will not get them to fly about the same speeds and stick together. You're going to need similar-speed ships converging on nearly the same distant point at the same time for that to even appear to happen. They're still going to trigger collision avoidance handling if/when they get too close, and one/some will veer off to avoid a possible collision and fall a little behind.

Ship groups can also be thought of as multiple layers of an array.
The Mothership is the 0th level of the array, its escorts are the 1st level. Simple enough!
But if you have a group of say 5 freighters, each with their own 2 dedicated escorts...that's 15 total ships. These 15 ships are not fully able to communicate with each other. 2nd Escort of freighter #3 gets a missile fired at it, informs its mothership (freighter #3) "help me, fire ECM!"...and freighter #3 just incidentally happens to be dead or low energy and can't fire ECM. The rest of the entire group that's not already busy might respond to a groupAttackTarget sent to target and attack the enemy ship that fired that missile or even the missile itself. But nobody else gets the message to use ECM.

Same instance also happens with Thargoid groups once they've launched a few Thargon drones. Thargoids are presumably going to defend their drones and they have ECM to reliably save them from regular missiles...but can't/won't because the Thargons are treated as missiles not escort ships of the Thargoid "mothership".

I've tried to tackle these problems with just clever AI scripting in Switeck's Shipping OXP, and I've got some very nasty surprises for anyone who dares look too closely. But some of this requires cranking out some quality .js code or changes deep in the core logic to really do it right.

While this could make the game a lot harder.
...Especially if rare "smarter" pirates (with higher accuracy skills) are grouping together in more intelligent manners, attacking in multi-prong timed attacks, using their ECM to save allied ships and missiles that have missiles fired at them, and playing dodgeball through asteroid fields or near the surface of a small moon or big station.

...Police and traders should regularly be doing the same. If you're clean and near a trader convoy or police, you get a little mutual protection and even a much-needed ECM if you have a pirate missile fired at you. You're giving the pirates and Thargoids another target to shoot at and hopefully fighting back also. It's a tough universe out there.

There won't need to be a core change of how large trader, pirate, bounty hunter (just really aggressive detached trader escorts?), police, or Thargoid groups are. Small groups are fine -- stations docking procedures are too slow to handle much at a time anyway. The few cases of bigger convoys can just be happenstance near-simultaneous arrivals of 2 trader groups, random placement of 2 big pirate groups, or hello Thargoid fleet!

Special OXPs can make epic battles with large convoys/fleets.
But the core game has to worry more about how single ships and small groups interact. The player currently is a group of 1, even a lone pirate will never flee that due to being "outnumbered" because of the stupid checkGroupOddsVersusTarget.

cim wrote:

Disembodied wrote:

so that you might get lots of pirates in Corporate State X one trip, and none the next.

Well, I wouldn't want there to be no chance of this, either. A well-armed pirate band might well risk an occasional raid into a corporate system, where the traders think they're safe and the right strategy could distract the cops for the crucial five minutes needed to blow up a convoy, scoop the goods, and retreat into witchspace. It wouldn't happen very often (and should probably happen less often than it currently does...) but it should happen occasionally.

Narratively, perhaps more likely in some systems than others. So Ceesxe (G1), for instance, has few dangerous systems around it, and is probably consistently pretty safe. Xevera (G5) on the other hand ... well, there's a reason that region is called the Siege Worlds.

That's fair enough, I think, as long as it's an occasional nasty surprise. Especially so if the types of pirates encountered on such rare occasions make sense (a well-armed pirate band making a large-scale raid in strength) and indeed if it was also connected to the larger galactic geography. Similarly, the same thing might happen in reverse: occasionally, a strong force of bounty hunters might team up to make a sweep through an Anarchy system. If there are several prosperous worlds nearby, perhaps local traders might club together and pay for a big anti-piracy raid. Maybe, even, such commissions might be available to players via the F4 screen? Go to system X, destroy Y number of pirate vessels?