Gilmoy

Join Date 12/2009

+41

Gilmoy's 0.41.2 Notes & Tweaks: Benevolent, Approval

I micromanage, so my turns take longer to play. I'm up to only turn 88. (Terran, Large, 3 Godlike)

I hope these comments illuminate for those of us who do pay attention to the numbers.

1. Ideology costs lowered, but traversal now breadth-first

The Ideology tree is revamped. All costs got lowered. However, a new rule seems to be in effect:

At each unlocking, (new cost of that tier) is deducted from your Ideology points.

I had reported this as bugs in 0.40 because it happened inconsistently. Looks like it's now a feature.

Corollary: Ideology tree traversal is now breadth-first (instead of depth-first). When you re-reach a non-empty tier's (new) points, you must unlock again in that tier. So you can't skip choices and beeline for tier 5.

Assuming 10 points per colonization event, and that you don't dabble across ideologies, this effectively correlates "tier choices" to "colonies", until you reach tier 3 and take your ideology's point-generating center.

15/22/33/50/75 = initial tier costs. (Don't worry about 75 being cheap: it will cost much more by the time you get there.)
2nd colony (turn 28) = 20: unlock tier 1 (cost 15). Costs increase to 18/25/x/y/z. Deduct 18 = you keep 2 pts.
5th colony (turn 70) = 32: unlock tier 2a (cost 25). Costs increase to 20/28/42/62/92. Deduct 28 = 4 pts.
8th colony (turn 91) = 34: unlock tier 2b (cost 28). Costs increase to 23/32/46/68/101. Deduct 32 = 2 pts.

See how insidiously they calibrated this so your 12th colony = 42 pts just misses the next threshold!! they are evil

13th colony = 52: unlock tier 3a (cost 46). Always, always take your point-generating tile improvement here, or it'll take you another 5 colonies to get back to tier 3.

Corollary: For Benevolent players, it is worthwhile to beeline for a 2nd colony ship just to unlock your tier 1 Benevolence for the Approval +10%. Do this before you even finish building out your homeworld.

2. Approval tweaks: Entertainment Center early; new starbase modules

The Governance/Commerce tech path now has more Approval bonus items. This makes it easier to run your empire at 100% approval for the monster Production bonuses.

Entertainment Center (tier 1 tile, +10%) moved from tier 4 Xeno Entertainment to tier 2 Xeno Commerce.
Tier 4 Xeno Entertainment now gives starbase module Recreation Center (Approval +10%!).

Typo: It says "Approval: +0.1". So its formatting is floating-point, instead of percentage.
Bug: Tier 4 Xeno Economics' module Orbital Shopping Center (wealth +15%) replaces Recreation Center (and vice versa?). They don't stack. So "upgrading" a starbase can suddenly lose an approval bonus for multiple planets. Gahh.

Tier 5 Interstellar Tourism now gives starbase module Orbital Resort (Approval +20%!!).

3. Gameplay notes: the Technological Capital gambit

Ironically, to build up to a Techno Capital on a research world, it must also be a pretty robust manuf world.

I had a candidate planet: class 18, triangle+1 = manuf (Solar + 3), nigh-hex of 6 = something big. I pondered going for a Tech Capital.
During preliminary build-up, just put a tier 1 Research Laboratory in the middle. The +10 levels of mutual adjacency bonuses (5 each way) pays for itself many times over (until you reach the point where you're ready to replace it).
To upgrade, you destroy the RL in the middle and build the TC. This loses 10 adjacency levels during construction. The simplest way to shrug off that loss is ... to set 100% manuf and get it built ASAP. Then you don't care about those -10 levels
My build time was surprisingly short: only 8 turns. Let BT denote this build time. Considering that my Age of Expansion goes ~110 turns, that's potentially a sound investment, if I can earn back those 8 turns.
Omitting my empire-wide rp% bonuses, I calculated:

+205% with a Research Laboratory (rp +10%) in the middle. Adding the base 100% from your unpumped pop, and letting E denote your off-planet bonuses, then let RLP = (305% + E) denote the base case, i.e. your rp% if you don't upgrade.
Upgrading to TC bumps that up by an additional +190% (lose +10%, gain +200%), +50% for an extra +10 levels of adjacencies (5 neighbors each went from +1 level to +3 levels). Let TCP = +240% be the gain you get after the big build is done.
Hence, my net rp% (ignoring pop growth) increases from RLP to (RCP + 240%).

So when do I break even? Let that be BE turns. Then I could have had:

(BT + BE) * RLP without the gambit, or
BE * (RLP + TCP) with the gambit.

Solving for equality, we get BT * RLP = BE * TCP, or BE = BT * RLP / TCP = 8 turns * (305% + E) / 240% ~= 11 turns. Empire-wide bonuses make this slightly worse, because your gain after the build is a smaller fraction of your total capability. Plus we should add in the 4-6 turns it took me to fully upgrade my manuf capability first (e.g. going to tier 3 Mega Factories, which is really a costly way to squeeze out only +5% above Xeno Factories), pro-rated by the fraction of RLP I was getting in those turns.
Even so, this instance of the Technological Capital gambit "breaks even" in about 20 turns. Thereafter, it's all profit. So my research lags for turns 66-76, but by turn 100 I'll be way ahead of a non-TC economy. This gets me through the Age of Expansion faster than not doing it.

Upkeep and Approval throws a monkey wrench into the above -- but it's a transient penalty, which soon goes away. The Technological Capital's whopping upkeep of 10 bc will hammer your Planetary Wealth approval penalty. My TC world was running at Approval +40% bonus tier while building it, but had to drop to +10% bonus tier afterward (since that was the bonus tier that maximized rp).

What boots it to gain +60% to your techworld's total rp output (+335% to +550%), if it forces you to drop -22% in Approval production bonus? (actually that would still be a profit ... but it means your turns to break-even BE goes way up, maybe from 20 turns to 50)

Solution: Entertainment Network +20%, starbase Recreation Center +10% ... and Mindfulness +25% I actually got my techworld back up to Approval 100%, even on the rim of the wheel.

This game is a devious ploy to make us solve algebra problems. Will one of the add-ons encourage us to use calculus? (in the complex plane? You collect 2i * pi * k for every black hole within this circuit, for each lap you complete) A closed-form expression for Ideology costs at tier k involves a summation ...

Locked Post

Advanced Search

Subscribe to this post

Do not email me updates for this post Email me updates for this post

Successfully updated karma reason!

You are my hero!

Successfully updated karma reason!

Larsenex

Join Date 04/2009

+191

Gilmoy, did you initially just build the 3 factories and Solar plant by themselves THEN the Research center and finally the TC all in linear order?

Successfully updated karma reason!

Quoting ,

quoting post

That's OK as long as we don't have to calculate the residues of the black holes by hand.

Successfully updated karma reason!

Quoting Larsenex,

reply 2

Gilmoy, did you initially just build the 3 factories and Solar plant by themselves THEN the Research center and finally the TC all in linear order?

Basically, yes. I build roughly:

growth: Colonial Hospital (and promptly upgrade to XMC if I have it)
manuf: For a world that needs "large" build capacity (e.g it has 1+ nigh-complete hex rings; or it's adjacent to an AI), I go for a triangle or line of manuf, with Solar in the middle.
as needed: Farm(s) next to the Hospital. Approval to let me stay on the rim of the wheel (at +50% bonus tier).

Then I start building the world's specialization. N.B. all of my large planets look oddly similar until this step

pass 1: tier 1 Research Laboratories to fill 1 ring, in triangle order (first finish 1 triangle, then go cw or ccw around the center hex). My manuf is usually enough to complete 1 tile per turn + excess to its specialty, e.g. to research (so filling in the center hex with a "sacrificial" tier 1 is worthwhile for the extra adjacency bonuses x duration of build-up). Maybe upgrade the outer ring to tier 2 (in 5 more turns).
pass 2: Destroy the center hex and build the major improvement thereat. This usually takes several turns x 100% manuf axis. I prefer to finish the build ASAP, rather than splitting some pop to use the remaining specialty tiles in their "crippled" empty-center-hex state.

Proof sketch: envision rectangles on a 2D time axis, where x = t and y = +% bonus. Then area ~= total output. Extend rightward to the 300+ turns it takes to win. How wide do you want that short rectangle to stretch before you get to the fun tall rectangle? For constant game duration, you maximize area by minimizing the short rectangle's width.

It takes some good fortune to have 2 land masses that are just right for this, e.g. a small island of 3-5 and a huge island with the ring + more.

For small planets with sufficient non-tile +manuf% (techs, starbase, planet), I skip the Solar triangle and go straight for the pop/specialty. These usually don't have a nigh-ring anyways, so they're not candidates for your precious Wonder-class improvement.

Successfully updated karma reason!

Larsenex

Join Date 04/2009

+191

Thanks, can you restate or clarify your build(s) for plants Class 8 or lower, assume no more than 3 hexes are 'touching' the rest are scattered or islands.

Successfully updated karma reason!

More 0.41.2 game-play experiences.

4. Gameplay: Micromanaging really, really, really, really sucks

~25 hours(!!), Turn 124. Each turn takes 20-35 minutes now. (The save file actually says "34.5 hours", but some of that was me AFK without hibernating the laptop.)

OTOH, my economy has already surpassed 2 Godlike AI (Drengin, Iridium), and I'm within 2% of the 3rd (Altarian). And that's without any major combats. Maybe 12 ships have died in the whole galaxy so far. They don't have the range to get to me, and all I build is slow weenies (but with range). So whatever they're doing ... be no match for my micromanagement skillz.

5. Gameplay: Lower build costs, higher upkeep costs

From Alpha 0.31, the economy was tweaked to make things easier to build, but costlier to maintain. That's a step in the right direction. It makes wealth slightly more relevant (from afterthought to annoyance). But I expect all numbers to get further rebalanced (i.e. changed), so I won't nitpick, except in broad terms.

It's still too easy. Set aside 1 world in 10 as a moneyball, and it'll pay for everything. Also, research Xeno Tourism and your tourism income alone will overcome almost any amount of maintainance.
It's still not enough. Wealth is unique in that it accumulates across turns. But it (currently) doesn't buy you any game-breaking (e.g. game-winning) advantages, in the way research does. Obviously, diplomacy et al. will change this.

If there is some "storefront" of benefits that you can only purchase with bc, and cannot obtain in any other way, then wealth generation for store buys suddenly becomes analogous to research in a tech tree. (Galactic auctions for small Precursor relics, maybe?)

6. Gameplay: Approval moneyworld asymmetry

Approval makes us tweak. But it seems weird that the way you tweak your Approval higher is by stockpiling a benefit that you want to do anyways. After some experience, I confirmed my initial impression that this is ... overloaded: it conflates two benefits, which causes synergy.

Specialist worlds become asymmetric. Forgeworlds and techworlds may need to tweak wealth-ward, but bankworlds don't, because they already maximize in that direction. Hence, a bankworld's default state is to have maximal Approval, for zero opportunity cost: they do exactly what they would have done regardless. That introduces an asymmetry in the way we handle worlds. (For a micromanager, it means banks are oases: You never need to tweak a bankworld. They're maximally happy, and there is no queue.)
The real-world parallel might be: an entity (government or corporation) gives a tax cut / kickback / bonus / other bribe to make its citizens happier. But the 0.41.2 implementation is that you save it all, i.e. the underlying premise is that saving makes us happy. That's true for me, but not for most of my peers! The more common societal effect is that spending a bonus makes us happy.

To make it truly symmetric, add a rule such as: "bc spent for Approval are not saved"; they're simply lost. Then all 3 axes would be symmetric ... if that's even a design goal. But this would probably be ugly in proliferation of game elements, e.g. distinguishing between "bc of savings" and "bc of approval".

7. Technological Capital Gambit addendum: 17 techs in 30 turns

From my scrap paper, here are the turn numbers in which I completed these techs.

T076 Terran has completed the Techological Capital.
T078 12.G3 Wealthy Population
T080 13.G4 Xeno Entertainment
T084 14.G4 Xeno Economics
T087 15.C5 CO/Optimized Manufacturing (but I wuz robbed -- can't upgrade to Interstellar Collectors until C6)
T089 16.C4 Environmental Engineering

a pause here while most of my planets build a Soil Enhancement

T094 17.C5 Research Coordination (robbed again, same bait-and-switch until C8)

I fear having wondrous manuf, and no combat ships worth building ... so the Small Ship Stack:

T095 18.W2 Weapon Systems
T096 19.W2 Defense Systems
T098 20.E2 Orbital Manufacturing
T100 21.E2 Interstellar Travel

This takes me into Age of War, by a few rps. N.B. Ages are strictly based on total rps spent: 3000 per Age.

T101 22.E3 IS/Hyperdrive Specialization
T102 23.E3 IS/Sensor Amplification
T103 24.E3 IS/Advanced Recirculation
T106 25.G5 Xeno Tourism
T109 26.G5 Interstellar Banking
T111 27.C5 Practical Fusion
T115 28.C6 Precision Manufacturing (now I can also build the C5 Interstellar Collectors)
T118 29.C5 Environmental Research (bug: has no effect? None of my research improvements show the +1 level.)
T121 30.C6 Terraforming

Iridium dang near invades me with one naked speed-7 transport!! I design, build, and shoot it down. Whew!

T122 31.M2 Militarization
T123 32.M3 MS/Enhanced Training
T124 33.M4 Planetary Invasion

And I actually have too many manuf worlds and not enough techworlds. In fact, of my first 12 colonies, and not counting the Technological Capital world itself, I had only 1 other techworld + 1/2 of Earth. 2.5 techworlds is not enough; I should have made 4-6. So I think this is actually slow progress from an imperfect start, and could be sped up.

8. Gameplay: Small ship stack and miner swarm

My small ships are just that: Small (50 capacity), with the +1 speed, +1 sensor, and +2 range. Since I'm way more aggressive than all 3 AIs combined, I have to go to them. I replace all redundant weapons and defenses with Life Support for moderately long range. I prefer Cannon for now (as you can see from above, I've never researched a weapon tech, like ever), and a Chaff vs. the AI's fixation on missiles. So my ships are minimal 0-0-3/0-7-0 with sensors, or 0-0-3/0-14-0 with a 2nd Chaff, and range 30 (= 34 for Terran). That's still not enough to penetrate into enemy backfields.

Workaround 1: Research the Medium Ship Stack, with capacity 75 and tier 3 components. Nag nag, I'm getting to it.
Workaround 2: Miner swarm. Find a resource (e.g. Durantium) between me and AI. Send fast constructor (+ fleet of weenies) to it, and preferably 5 hexes beyond it. Create a starbase. Follow up with double-constructors. Specialize to Mining Ring, and take the defensive modules.
The 1st constructor must have sufficient range. Thereafter, follow with a stream of (at least 2) double-constructors with no range (default range 11).

All of your ships can exist and move beyond your range, with no penalty. You just can't issue a movement order to them in any out-of-range hex. You can freely order them to move from one friendly ZOC to another ZOC, no matter how far the gap between them. The only risk is that you lose some control over them while they're "in between" ZOCs: you can't creep them 1 hex at a time (so they can just walk right up to an enemy snub fighter), and your evasion is limited to full-move straight-line paths back to any ZOC you can click in.

I need to mine resources anyways. So this tactic lets me use a single starbase for the mining role and the range-extension role. One good mining starbase lets my Small Ship fleets swarm over enemy space.

I really hope the AI will know how to do this to us! It requires a bit of a paradigm shift to think halfway to the goal.
Godlike AI does not (yet) counter this human tactic ...

9. Gameplay: Large galaxy just barely not large enough

I need 13 colonies to get to Ideology tier 3 and take my point-generating tile improvement. Well ... I hit 11 colonies, and then Iridium outraced me to 2 of them in my backfield. Now I'm hemmed in by all 3 AIs. If this galaxy is a bridge hand, then I'm West.

There may not exist another 2 empty colonizable planets!! Grumble. But there's hope, I found unsurveyed artifacts along Altarian's border. You know what that means!

Technology, shift to C8 Integrated Research
move Survey Ship
engineer's head explodes
Technology, shift back to E4 Life Support

No way I'm wasting an Artifact on something I can finish in one turn myself ...

Successfully updated karma reason!

Larsenex

Join Date 04/2009

+191

Amazing play/turn by turn list. I actually missed a customers phone call due to being engrossed in your post Gilmoy.

Successfully updated karma reason!

DARCA1213

Join Date 01/2014

+106

Wow.

Successfully updated karma reason!

First, I'll give a perspective on Beta. I'll use this to wave off entire categories of bugs as "straightforward, not worth harping upon". Then I'm going to steer toward the gaps, like a survey ship on Meander.

A. Metaphor: Beta as onion, with unfinished freeway ramps

Of course, I'm just a (newbie) player, with no knowledge of Stardock IP. But it seems to me that we've been entrusted with this Beta to give it a spin. So I'll first explain how I perceive it to be, and why I don't sweat the obvious placeholders at the boundaries.

Dev builds are like an onion: they expand from inside out. There's a core engine that runs, which we can test. It's surrounded in every direction by placeholders, which are null or stub modules for further development.
A placeholder is like an unfinished freeway ramp. It's a tiny bit of quick-and-dirty code, e.g. trivial algorithms with text-only output. It exists so that your Alpha/Beta game doesn't just halt when you fall off a ramp. It's deliberately cheesy, like they spent only 1 hour on it, because they already plan to replace it with a 12-story condo, and it's not worth another 5 minutes.
Some obvious placeholders include:

ship combat stub (all defenses are ignored)
ground combat stub (attacker has insane +% advantage)
Alpha/Beta AIs (that mismanage tiles and stack ships onto rally points)
disabled tech branches: no diplomacy, trade, ...

Corollary: Don't get hung up on the placeholders. They're supposed to be unfinished. They just mean you've reached the limits of this Beta iteration. Replacements for them are already in the dev queue.

So instead of heading straight for a placeholder and complaining that it's unfinished, ponder what we do have (in the center of the onion). GC3 Beta 1 consists of, at least:

Graphics engine: a DirectX(?) engine that renders solid(?) models, with zoom/pan/rotate, animations, hither/yon clipping

From the dev streams, I see Paul et al. showing the Ship Designer, where he has the 3-orthogonal-arrows widget showing axes of rotation, exactly as in some commercial solid modelers, e.g. SolidWorks. I haven't snooped GC3 files to see if they actually use a solid modeling format I'd recognize. (Solid modeling in-joke: anyways, it's owned by Dassault Systemes!)
Corollary: Graphics woes are useful feedback, esp. the textless issue and drivers fix(?). I've had zero problems here.
Fog-of-war feedback falls here, esp. black hole flicker.
In the 08/28 stream, Paul found an Artifact inside an Elerium!! that was way cool, let's keep that plz -- you must install the Survey Mining Module to mine it

XML scripting. Vast amounts of game data, including techs, components, improvements, races, etc., are simply XML files, which you can browse.

Corollary: Lots of tech-related bugs are, to me, just glorified typos: they're wrong in an XML file. Hence, the fix is a trivial typing job of ~1 minute each: somebody will just type in the correct values. (Far more interesting is the design decision to determine what the correct settings shall be, which drags in design intent and game balance issues. We're along for that ride.) Here are some bugs that I just don't worry about. It suffices to point them out, but probably they're already getting re-worked anyways.

text typos (but srsly, guys -- run a spellchecker )
improvement descriptions not matching their effects, e.g. Environmental Specializations and high-end terraformers(?)
starbase module tangled prerequisites, e.g. approval/wealth
numerical or signed typos

We deduce that the developers have an XML editor that lets them bang out these definitions.
I'll detour to note that most of the player-buildable game data consists of intermingled trees: tech, tile improvements, ship components, and starbase modules all have one-way(?) dependencies. Techs enable buildables, and buildables are upgradeable to themselves. So the minimal representation of GC3 is a bunch of trees, with tech at the top and many non-tech trees branching off of them (which means it's still just a tree, although you might say it's a "3D tree", like a pyramid).

Actually, is it a pure tree? i.e. are all arcs strictly outbound from techs? More generally, it would be a graph (although I can't think of anything in GC3 that qualifies). Whimsy: suppose you must research tech T to get improvement J, use it to build a ship with Survey Mining Q , and then mine that for enough turns to enable new tech U. That would induce graph-ness.
Ideology could be a disjoint tree unto itself (e.g. Benevolent's Missionary Center requires, and confers, no other tech), so actually we have a forest (disjoint trees). We could hypothesize a one-way graph, which has no self-loops as in a true graph, but is a forest where trees can have multiple parent nodes (which is no longer a simple tree). For example, if you need Ideology C and tech T to build some new component, then those trees would merge. Either way, your XML editor would need a UI more powerful than one-tree-at-a-time.

Hereafter, I'll use intermingled trees to encompass { tech; improvements; components; modules; anything else }. Surely the dev XML editor's UI depicts the entire tree, and lets them sweep their eyes across it, to ensure consistency and non-tangled hierarchies.

I wonder if their XML editor actually lacks a good tree-wide UI. That would explain the tangled hierarchies we've all noticed. As 1-at-a-time XML files, tree-wide consistency is basically impossible for any human to maintain. With a good visual UI, tangled loops would stick out as red arcs.

I conjecture that the engine also has a C++(?) meta-object protocol, which is populated by slurping in the XML definitions. Hence, the fields we see as attributes in XML get mapped to class methods or game engine mechanism in the code. This is illuminating because we can deduce whether those game mechanisms work properly: if all techs with some <attribute>Foo</attribute> fail the same way, that suggests a broken or missing mechanism. Ergo, each of the XML attributes is a valid topic for bug reports.

Unique with upgrades: We know that this allows "staircasing", where you upgrade a unique to tier 2, and then tier 1 reappears (and you can build it again). Likewise, when you get tier 3, you can upgrade both, and then rebuild a tier 1. We think this is the design intent, not a bug.
Wonder: Known to be broken. Each player can finish 1 wonder. I have 2 Manuf Capitals, mine and Iridium's.

Wonder-racing generally needs better UI throughout, e.g. to goad other players, and auto-cancel all failed wonders in all queues when one gets completed. Surely this is a placeholder, and they're on it.

<Scope>Global: Seems to work correctly. We ignore here whether the numbers are game-balanced.
Ideology +1 pt per turn: Known broken. Mechanism not implemented?
others ...?

UI API. Hexes! Sidebars! Production wheel! Progress bars, buttons, tooltips, tabs. Report any vagaries for these features.

Fleet tooltips often lose metadata, e.g. the bottom 2 rows of text (including fleet movement pts) get truncated away. I save/restore often (ahem), and I know that after a few iterations of save/restore, this bug is inevitable. Work-around: Exit/restart.
Production wheel resolution is clearly finer-grained than 0.1 of each axis, or 1% of your production. It may go down to 0.1% of production ... or maybe it goes down to whatever your screen resolution in pixels is, and then rounds nearest. Grumble.
Generally, after the "golly" phase wears off and you get into gameplay, consider whether the UI works for your desired way of playing. Is information you need readily available? Do you have to drill to get information occasionally? Incessantly? Can you issue your gameplay intent quickly and conveniently? etc.

Game state calculator. Equations galore, and math techniques to compute them.

Round-off discrepancies. Many places in the UI show annoying discrepancies in numbers ... as if they're passing through two separate developers' code paths. (Or the same developer, in two different months ... and he cut-and-pasted code, oh no)

Approval %: Planet bar vs. Planets sidebar; empire average %
Queue items claiming "1 turn" to complete, but not completing

Game engine. I guess this has to be somewhere, eh? This includes all in-game considerations, i.e. what we as players encounter when we're actually playing to "win".

Shipyards: orphaned; shut down (yaay!); renaming them to "shipyard" ; crash when too many candidate sponsors; need better sponsor management.
Ship Designer. I have no artistic inclination, so I'll defer to others here.
Artifacts in Eleriums, hehe
Conquests: can't invade planet; can't fight post-conquered ships. Almost surely just a byproduct of the combat placeholders.
Mystery crash-to-desktop (CTD).

I think the consensus of 0.41.2 is that ... the core engine works. In fact, the Alpha/Beta game, even with all of the placeholders, is already playable and (gasp) addictive. We're having fun and so we sometimes (often?) lose sight of the larger goal that we should be celebrating the bugs (as a success of our testing methodology) instead of griping about them. I think many of the demands for quick fixes are justified in this light -- even Paul has mentioned some of these in the streams.

Ship/planet combat "placeholder 2" pans to each combat

Maybe it is worth an extra 30 minutes of dev time to make a "patch to the placeholder" (for the ~2 months before ship combat is fully exposed), if it gains an enormous qualitative difference in the playability of Beta 1.x/Beta 2. I can see it from Stardock's pov: maybe they expected Alpha to consist of a crash every other game, and could not anticipate it working so robustly that ship combats would become the bottleneck.

Successfully updated karma reason!

More 0.41.2 gameplay comments. I will try to steer my comments to what I call a gameflow level. That is, I'll take a tour through my actual in-turn play-by-play methodology, and note whatever sticks out as being ... not smooth. Generally, anything that forces me to pause and drill will interrupt my trail-of-thought, and become a gameflow obstacle. These lean toward repetitive player tasks, and whether the UI handles them well.

I will skip most "typo-class bugs" and placeholders, as outlined above.

This is the same 0.41.2 "Bene Darda" game as in my original post (with the Technological Capital gambit), which I started on the day of 0.41.2 release. I'm up to ~50 hours(!!), turn 143. I micromanage every queue, every turn. I'm winning, limited only by the slow speed of sending transports across a Large map.

B1. Move all ships

I'm West. Drengin is South, Iridium is North, Altarian must be East. I think there are no more empty colonizable planets anywhere.

Ships sidebar is too tedious. I almost never use the Ships list to find a ship.

Flat list in chronological order. Surely it's a placeholder, so I won't gripe further.

[UI-request] Group tabs by "theater" (player-defined, oh please). A theater is whatever logical grouping I want. Mostly, it would be a subset of ships + fleets in one map area, whose gameplay decisions must be planned holistically. It could also include ships that are far away on the map, but with same player intent, e.g. en route.

I prefer to zoom way out, so that my eye can sweep.

Even with ~50 ships, I have had no real trouble keeping track of them all. Dunno if that scales up to 200+ ... maybe my head explodes
Let abstract mode denote the zoomed-out icons, and rendered mode denote zoomed-in detail.

For each ship/fleet, click it on map, and move it. I go roughly ccw around the rim of the Large map, from Drengin (sw) to Iridium (nw).

Abstract mode UI is insufficient to inform me as I sweep my gaze. This forces me to drill, which is like the gap from Excel to Concentration/Jeopardy.

[UI] In abstract mode, currently selected ship/fleet needs a highlight.
[UI-request] In both modes, show ship/fleet status icons.

if idle, show hopping ball icon (or other attention-demanding). I want to sweep my eye and see what needs my attention.
if upgrading, show hourglass?
[UI-addendum] These could be player-selected overlays. Example: I choose a palette of "HP bar && troops bar && idle", bind it to hotkey Ctrl-7, and whenever I type that, it toggles my palette of ship/fleet icons on/off. Then I can tailor a bunch of palettes to the info I want to selectively view.

Region-based views. Sometimes I want essentially a database query, which is way beyond a click-to-drill UI.

In conquering-horde mode, I often wonder: what ships do I have coming here? That could be decomposed into a couple of idioms.

[UI-request] Define a region as a 1st-class query, e.g. drag to mark a circle. This is potentially overloaded, which may require a shift key or hotkey to disambiguate.

Multiple select. It selects all friendly ships/fleets in that region.
Region define. In this mode, it simply defines a floating circle, which persists as a UI element (until the next command?).

[UI-request] Comefrom. Right-click(?) on a region, and choose a query such as: show all inbound ship paths that end within this region. Then I would use this frequently around Iridium planets I've bypassed, to verify that I have enough transports en route. Other obvious commands could include: show all outbound, ...

For a game in which distances in hexes is so important, it's tedious to have to manually count hexes in space. Elegant UI would make this elegant.

[UI-request] Tape measure. Click once to anchor. Thereafter, a ruler overlay follows your mouse around, with a floating distance display. Wave it over the map, and it shows you the straight-line distance in hexes. Click a shift key (maybe literally hold down Shift) and it plots an actual ship path (i.e. avoiding impassable hexes, taking shortcuts), and shows you the distance.
[UI-request] Circular region. Click once to center the circle. Thereafter, dragging illuminates a circle of that radius. This merges with the region-define UI mentioned above. It lets you envision starbase effects before you actually get there, without having to count hexes.

Successfully updated karma reason!

B2. Upgrade ships

I wonder if the ship upgrade mechanism is a placeholder. We didn't have it in Alpha 0.31, and it got added for Beta 0.40.

[Bug] Upgrade costs are not deducted! (yaay! until Beta 2) Possible cause: it looks like the same window as a Quick Build, so maybe it's the same bug.
[Bug] After upgrading all ships in a starbase(?) (and after it's done), each ship's hover still shows its old stats, and its Details still shows the old stats, but the new components. Work-around: Eject the ship(s) to update their hover stats.
[Exploit] A ship can move its full movement in a turn, then upgrade, and still get credit for "1 turn" of upgrading. Hence, within your own ZOCs, your ships upgrade "instantaneously", because it takes "1 turn", but you can use them normally in the same turn, as long as you issue the upgrade command afterward.
[Exploit] An upgraded ship refreshes its old stats. If you downgrade its movement (e.g. from 8 to 4), it will have "8/4 movement" for that turn only. I guess it's symmetric: if you upgrade to a faster design, you get the old speed for 1 turn.

We can infer that the game engine's turn-updating algorithm goes something like:

this->stats = this->design.stats
if (!--this->turns_to_upgrade) { this->design = this->upgrade.design }

Separately, the auto-generated ship designs in Designer annoy me in several ways.

Sometimes they're just ... not interesting. e.g. if I've chosen Rock as my weapon focus, I shun Paper and Scissors anything. I delete most auto-generated designs. But they keep coming back. Every time they re-clutter my ship designs list, I have to delete them again.

[UI-request] Unsubscribe/Do-Not-Call. When I really don't want to see an entire line of ship designs, let me unsub from that product line. All future upgrades to that design are routed to /dev/blackhole/.
[Bug] Upgraded Ship Design is identical to previous version. In general, completing a tech seems to trigger Upgraded Ship Designs too often, with no sanity-check for identicality. (You did overload operator= = () for class ShipDesign, right? )

[Bug] I completed C7 Technological Focus / Manufacturing Focus, which gives only "+15% Research". This triggered a tranche of Upgraded Ship Designs ... where every design was identical to the previous version. Not a thing changed, because that tech level is disjoint from ship componentry. Why did it even trigger? I don't think this is an XML attribute, which suggests that it's more deeply wired into the game engine.

I am a rudimentary Ship Designer: I'll spend 30 seconds to tweak 1 component and make it look nice. The only functionality I use now is scaling, rotation, and occasionally an offset. I got my double-constructor to look like a pair of binoculars, which makes me laugh every time I build one.

[UI-request] Clone component. Select existing component, click Clone, and you get a copy with the same parameters (scaling, rotation, etc).
[UI-request] Set slider granularity. Most of the time, I want the Scaling slider quantum to be 100%, not 3%.

Successfully updated karma reason!

B3. Govern all planets

For each planet, I check its build progress.

New planets (colonized or conquered) impose light cognitive load, because most turns they don't complete anything. Glance and skip.
I always disable auto-upgrade for every planet, and do it manually.

[UI] Govern global option.

I rarely assign many tiles in advance.

Double-click to place a tile is good. The "Build" button is also good.
[UI-whimsy] Temporal "script" that can act on tiles not existing yet. To truly automate a complex tile build-up, we must have a way to specify "action-in-the-future", i.e. targeting tiles or improvements that do not yet exist (because they haven't been built yet). A sample script will illustrate:

@A, build, Colonial Hospital
@B, build, Basic Factory
@C, build, Basic Factory
@A, upgrade, Colonial Hospital to Xeno Medical Facility
@D, build, Soil Enhancement
@D, build, Solar Power Plant
@E, destroy, Slave Pit
@E, build, Xeno Farm
@D, upgrade, Fusion Power Plant

This may be outside the scope of GC3's design. I've never seen a 4X include a temporal action language. So the first one to ship would be a 5X ...
[UI-request] Maybe replace the Build button with an Insert button, which lets you insert a build request into an existing list. Or Reorder up/down to edit a list. Currently, we must delete everything in the list, then start over, which is maximally clumsy. (Guess which competitor's 4X I've played ...)

For each planet that can-complete a queue this turn, I optimize its queues. (Remember that I'm an off-the-deep-end micromanager.)

A planet's 3 queues include:

social: its own manuf queue. This is always 1-to-1.
military: its shipyard queue. Complication: This is N-to-1 in the general case, where N is the number of co-sponsors.
technology. This is M-to-1, where M is all planets in your empire.

Lossy queues will become carry-1. All queues in 0.41.2 are maximally lossy: all excess production is discarded. This is already slated to change; rejoice!!

In the 08/22(?) stream, Paul mentioned that social and military queues shall become "carry-1", i.e. all excess production will carry over, up to the build cost of the next queue item, but without ever finishing two in 1 turn. So any production beyond the 2nd item's build cost will still be discarded(?). But this means that most of us, most of the time(?), will no longer have to micromanage fractions of 1 item's cost.
[UI-request] 2-item Technology queue, plz
[Exploit?] This suggests that we can Rush the current queue item, and still keep building toward the next item.

This one feature probably moots most of my per-planet tedium. (N.B. moot, v. to render moot tenses mooted, mooting, will moot #wordcrime)

B4. Govern all shipyard-planet cliques

N-to-1 queues (shipyards and tech) still exceed a click-to-drill UI. Clicking/hovering is inherently 1-at-a-time. The canonical 0.41.2 gameflow problem (for me) is: given lossy queues, and that my M-to-1 tech contributors may intersect some or all of my N-to-1 shipyards ... how do I conveniently optimize all of those queues concurrently? Work-around: I scamper like a mouse to do a database join, run Excel on my 2nd monitor with some macros to compute the roundoff breakpoints, and do pixel click-hell on every queue. Yah, it sucks.

[UI-request] Planet includes shipyard link. Fit it somewhere. It should show which shipyard, and what the spending is.

In the planet title bar, next to the "<" and ">" buttons, add a "^" button for the shipyard, which takes you directly to the Shipyard list. If this planet has no shipyard, gray it out.

[UI-request] Shipyard/Tech shows all contributors. Just as Shipyard has a hover with all sponsors and their contribution, let Technology have a hover with all active contributors.

Or: Customized Planets list views. In Alpha 0.31, the Planets list showed each planet's mp, rp, and (net) bc, which let me sweep my gaze without drilling. In Beta 0.40 (and 0.41.2), it now shows bc, pop, approval. So we've lost a gaze-sweeping way to see all rp contributions to Tech.

[UI-whimsy] M-to-N production control. Dunno what this interface should even look like. Maybe it has to embed a complete graph renderer, and show you a clique at a time (which could engulf everything if all of your cliques also produce tech). Each node of the graph would be a mini-wheel(!! -- heh). The key might be to auto-calculate the minimal contributions to exactly-finish a local queue, and split each producer's mini-wheel into 2 abstract chunks: (a) earmarked to complete X, and ( b ) excess. The user's control would be something like rearrange excess chunks. For example:

A clique of 3 are each building locally (social), jointly feeding 1 shipyard (military), and all produce tech.
Their clique pies earmark exactly enough to complete each one's social build. These portions are grayed out, and can't be selected.
The shipyard's remaining build is a partially-filled pitcher. I click the pitcher, then click excess portions from its sponsors, which fills up the pitcher, and earmarks each planet's excess with a pitcher icon.
If the shipyard's pitcher overflows while I still have planets left, that planet's excess is fragmented into an earmarked portion and a new excess.
Finally, Technology sits as a huge pitcher (bucket? swimming pool?). Double-click that to earmark all remaining excess to it.
If the Technology bucket fills up, all remaining excess is truly excess. Click a moneybags icon to set all excess to wealth.
N.B. earmarking to a queue automatically sets pop in the planet's wheel to exactly achieve that much along whichever axis is relevant: manuf for social & military queues, research for tech, wealth for moneybags.

Finally, around mid-way through the Age of Expansion, I start to juggle multiple shipyards. By this, I mean the trick of having a cluster of 3+ planets share 3+ shipyards, and switch madly to 1-2 per turn to exactly meet remaining production or exploit proximity. Shipyard sponsor management could really use some drag-and-drop UI.

[UI-request] Drag-and-drop shipyard sponsorships (literal). Make a sponsorship arc be a 1st-class UI primitive, with 2 anchors. Click an anchor to attach or detach it to/from a shipyard/planet. Then we can change our sponsorships entirely from the map, without entering Shipyard or Add Sponsor.
[UI-request] Ibid (abstract). The above UI idiom isn't pretty when you want a far planet to "hold a leash", e.g. at distances of 20+ hexes. Also allow an abstract drag :

[UI-request] between the Planets list and a shipyard on the map;
[UI-request] between a planet on the map and a shipyard in the Starbases list;
[UI-whimsy] directly between the Planets list and Starbases list. (and show both lists at the same time, hehe)

B5. Turn, Reload, add 0.1, Turn

I hate it even worse when the UI lies to my face.

[Bug] Social/military/tech queue says "1 turn to complete", but next turn it is not completed.

Work-around: Save, Turn, check all queues to see which ones lied to me. Reload, add 0.1.
Repeat again and again and again, as needed.

Successfully updated karma reason!

DARCA1213

Join Date 01/2014

+106

You are awesome!

Successfully updated karma reason!