What is Delta / Revamp, what is it doing, and where is it going?

[Zitchas]

First off, RisingLeaf created a great summary of some of features and additions to Delta at #15 so I'm not going to duplicate their work, but instead give a bit of a short introduction of sorts.

Revamp, properly said, refers to this PR from years back: endless-sky#5772 . It was an ambitious proof of concept intended to throw a whole bunch of stagnant PRs together, add in a whole slew of other changes, mix it all together, and let people experience it as a unified whole. The theory is that there are plenty of changes to the game that are individually either meaningless or potentially even a negative to the game; but which are collectively a positive, perhaps even a strong positive.

Technically speaking, the Delta project is Revamp 2; but I know several of us have just referred to it as "Revamp", and I doubt that's going to stop.

---

So, there's two things here, in terms of long term and short term.

Long term, Delta is intended to be a testbed in the same style as Revamp that will combine a lot of things together in order to test and discover things that might not always be apparent. It's also a means of testing a variety of things together that are too controversial or have too many pre-conceived notions about. Remember arrival distances? That was suggested for mainstream many times, and it was strongly condemned and rejected by a broad swath of the community. Until people tested it out in Revamp as part of many other changes, and now it's part of the game. Preconceived judgments of ideas are just as bad as stereotypes of people.

So, long term, I envision Delta being the place where a lot of testing and experimentation takes place on a whole range of topics.

Short term, there is need for focus. Right now, those of us who are most active here are focused on combat and related challenges. It's been a little slow as mass and engine changes are a really central element, and those are being worked on. But we're generally speaking focused on a heavily intertwined set of problems:

• Uselessness of fighters & drones
• The fact that big ships are better than small ships in all things in all places. (there's really no reason to ever fly a ship smaller than the heaviest ship of the type relevant to your current activity that you can get)
• Deathballs: Aren't fun to be in, aren't fun to watch, and don't make sense.
• Strategic roles: Fleets are more interesting to watch when they have varied roles and styles. Fleets are also both more challenging and easier to deal with when they have varied behaviors & stats: Less likely to clump up (so easier) but also more types of threats to keep track of and figure out counters to. (thus harder)
• Reward skill: Players should feel both incentivized and rewarded for their personal skill. Their ability to fly well should matter more than whether or not they have 10% more of a particular stat than their enemy.
• Solo vs fleet: The expectation is that the player is flying solo, and the gameplay should focus on that. However, combat is more interesting when one can come up against a range of threats, sometimes solo ships, sometimes fleets, sometimes packs. figuring out how to balance fleets vs solo play, and in particular how to prevent any possible ways of helping solo players enjoy the game equitably without enabling fleets of comparatively powerful ships. (Most nerfs to NPC ships also nerfs the player)

---

So, you say. That's a lot of stuff to look at about combat. Why are some of the first things merged into this scanner improvements and GUI elements for my flagship?

Information, the first principle of warfare, must form the foundation of all your efforts. Know, of course, thine enemy. But in knowing him do not forget above all to know thyself. The commander who embraces this totality of battle shall win even with inferior force.

Spartan Battle Manual
Accompanies the Secret Project "The Command Nexus"
Sid Meier's Alpha Centauri

This quote from one of the factions in Sid Meier's Alpha Centauri pretty much sums it up, but to extrapolate:

1. For players, in order to make informed decisions about how and when to engage, they need to be informed. Exactly how and when the scanner information is unlocked is subject for debate, but one way or another, players need a way to get information about their enemies. Compounding this, our variants system is intentionally setup to make it hard (or even impossible) to know how a ship is configured until after it's too late to matter. The more variants we get, the harder it will be. (for example, I consider all clippers to be the extremely overpowered Heavy Rocket variant. Are they all? no. But without a way to tell the difference, they might as well all be.) Without some way to access information about their enemies, there is no way to pick targets or to adjust one's own outfitting or behaviour to deal with them. On the flip side, it's one thing to look at your shipyard stats and see that your max speed is 500. it is quite another to be flying through space wondering "Am I at max speed? Or am I not quite there yet? What's that purple stuff on my ship, what's happening? Why don't my solar panels work here? Why are they suddenly working better over here? Providing numbers that people can look at enables people to understand their own performance, and understand the limited ways that system terrain affects them. Sure, we tell them in tooltips that solar works better closer to stars. Nowhere do we tell them that X type of star will provide a lot more power than Y type of star. Likewise, it never says how much better they will be. Are we talking maybe 5% better? 10% better? 200% better?
2. For content creators, and especially people working on combat balance, it is very helpful to talk about numbers. Figuring out why one ship beats another is considerably more rigorous when you can compare in-flight attributes and see what they are actually doing instead of just making guesses based on whatever attributes they've been given. It's also proven helpful in debugging and identifying inefficiencies in the code, too.

In short, information affects every aspect of both planning the encounters and flying them. If someone doesn't want to use information, that's their choice, nothing forces them to have it. But we should at least be giving people the option. Information lets people learn. For a lot of people, it's the only way they will learn. Very few people will ever learn the hierarchy of what stars will provide the most solar power if they don't have means of measuring performance.

---

So where do things go from here?

Well, there's some big things and little things

• What do we do about missiles? Lots of them seem to range from "inconsequential" to "overpowered" depending on if one is facing one of them or a dozen. Are they too fast, too easy to dodge, too pathetic?
• What's the balance point between guns and missiles?
• What are the roles for guns, turrets, and missiles, and how do we differentiate them?
• Arkhne's working on ship masses and engine thrust. Hopefully we'll have that all in soon.
• Engine balance: The bigger the ship, the more engine it can hold, and the more it can make use of anything engine in the galaxy. Hence we have more than a few heavy warships that can outrun any fighter or interceptor. Frankly, the current engine system is broken. We've discussed a bunch of different ways of balancing engines, but none are overwhelmingly better at a glance. Once the mass and thrust changes are in, I'll be implementing one of the various new ways of balancing engines. Hopefully it will allowed more freedom and variety in our engines, while cutting down on the abuse.
• Petervdmeer has restarted work on Formations. Those were a big hit in Revamp 1, and on the shortlist of things that we really want in Revamp 2. As soon as it is in working shape and he says it's ready for us, we'll be adding it.
• Samrocketman has a whole range of interesting PRs, and once we've worked out an efficient way of merging them, we'll be adding them in.
• Vitalchip has been providing feedback on some AI things and helping code in general. Hopefully we'll have some new circling fighter AI again soon.
• Departure distances
• ???

So, things are likely to move slowly until the engine and mass changes are implemented, then there's a whole slew of stuff (and this list isn't comprehensive. I know I've missed some things) that would be worth considering.

---

In a perfect world, we'll spend the next while working primarily on combat stuff. Eventually we will reach a point where the amount of changes is so great that it's hard to tell what's doing what. At that point, we'll be faced with a decision:

• We've managed to improve combat.
• We've failed to improve combat.
• We've managed a miracle and solved combat entirely.

In all cases we have a discussion with everyone who tested it about what is deemed a successful improvement, and we PR it over to mainstream, as well as into our main branch here. Then we pick through whatever's left, decide what's worth keeping, what needs to be tossed. Make detailed notes (and an archival copy of the branch for future reference), and then we reset to a fresh copy, based on mainstream+the PR'd best elements, and start again with new ideas and new combinations.

Eventually we will get combat to a good place, and call a halt to this phase of experimentation. At that point, we can identify a new focus, and work on that.

That being said, it's an open concept. If there's enough people interested and contributing, then we could certainly have more than one experimental branch. It would be quite easy to have separate branches for alternate approaches or focuses. There isn't enough people contributing right now for that to be the case, but it could.

[Zitchas]

As something of a quick summary of changes in Delta:

Endless Sky Delta
Changelist:

• Additional scan information & modular control
• Lateral Thrust
• In-flight drag reduction (spaceships move more like spaceships and less like boats)
• Additional player HUD information on the flagship.
• Limited turret arcs (now also in vanilla)
• Adjusting mass, thrust, and turning across all ships. (a weaker version of this was implemented subsequently in vanilla)
• Implementing engine slots
• Fixing problematic upstream changes to Far Osk 27.
• Implementing station keeping capabilities (all ships have small amount of reverse thrust in addition to the lateral they already have)
• Sol has been scaled up to provide a larger game map area.
• Hai Info bridge implemented (this re-enables access to Hai Reveal and reverts a number of problematic changes in upstream)
• Expanding map zoom ranges
• Expanding the main screen zoom levels
• Smoothing the zoom progression at far out levels to prevent tiling and performance problems.
• Cloak differentiation for the Archon and Remnant
• Splitting Ship Info into Ship Info and Hardpoint Info.

By request, the same list broken up a bit:

Endless Sky Delta
Changelist:

Pure code changes:

• Additional scan information & modular control
• Lateral Thrust
• In-flight drag reduction (spaceships move more like spaceships and less like boats)
• Additional player HUD information on the flagship.
• Limited turret arcs (now also in vanilla)
• Expanding map zoom ranges
• Expanding the main screen zoom levels
• Smoothing the zoom progression at far out levels to prevent tiling and performance problems.
• Splitting Ship Info into "Ship Info" and "Hardpoint Info"

Mixed changes:

• Note: Most code PRs do touch the data files. Interfaces.txt, for instance, and addition of images.

Pure data changes:

• Adjusting mass, thrust, and turning across all ships. (a weaker version of this was implemented subsequently in vanilla)
• Fixing erroneous changes to Far Osk 27.
• Implementing station keeping capabilities (all ships have small amount of reverse thrust in addition to the lateral they already have)
• Sol has been scaled up to provide a larger game map area.
• Hai Info bridge implemented (this re-enables access to Hai Reveal and reverts a number of problematic changes in upstream)
• Cloak differentiation for the Archon and Remnant
• Implementing engine slots