So let’s have a look through my illustrations folder, shall we? Do note that most of these are very much in-progress, though of course I’m not going to show you any of the truly failed compositions. Still, I do hope you find it interesting to get a glimpse into my process here.

“Put it in Bay 12 and be careful – those uranium rounds are hot!”

Now these illustrations are intended to be shown in-game as an accompaniment or decoration rather than used as promo art so they’re much smaller than earlier work I’ve shown on the blog here. This has the added advantage of making the process of creating them much faster than full-sized drawings. It’s more interesting for me, too, because I can experiment more with theme, composition, and technique.

Location Illustrations

Starsector is created by a very small development team so it can’t very well have separate game modes for walking around every space station and backwater colony the player comes across. Many different locations need to be distinguished as cheaply as possible, basically. So: we offload description to text and a sort of “Starsector Holiday Snapshot” illustration — and that’s where I come in.

Arcology interior: Yes, I played way too much Sim City 2000.

This one certainly owes to Paulo Soleri by way of, yes, Sim City 2000. Arcologies are intrinsically cool and I swear I’ve been drawing them for upwards of 20 years now. Enough said.

I must admit that the shape of those towers comes from some building in the intro video to Quake 2. I never got over it for some reason. Also once I got this far into the sketch I realized that I really shouldn’t put too much information about what is in the sky of the planet because that means Alex couldn’t use this for an outpost in orbit around, say, a blue gas giant. Oops.

This is probably a good time to dissect the exact dimensions of the Hound landed in this image.

So there: change up the perspective so you can’t see the sky. Again with the huge inter-tangled cities, yes.

Not every planet has proper facilities — these ones seem to have been bombarded into a pile of scrap some time ago. But people make do.

Piles of rubble and ruined cities are also good. I really must make a note to start some images of barren worlds and space-stations.

A Little Something Extra

Part of my goal with the illustrations is to expand the emotional scope of the game. Depending on the interpretation of a concept, a player can be given an entirely different feeling about the subject. Take for examples these two approaches to an image titled “Marine Preparations”:

Marine Preparations 1 : Faceless robotic killing machines

Marine Preparations 2: The band of rogues

One set is a pack of inhuman killing machines, the other is a sort of roguish band of mercenaries with faces and individuality. Are you so quick to burn through marines to take enemy ships if you’re reminded that they’re meant to be people? A touch like this can nudge the meaning of the game away from what the raw mechanics tell you.

(By the way, my interpretation of space marines is probably somewhere between Aliens and Planetes.)

Hull Breach: It’s not a good day for the Deck 3 maintenance crew.

(Alt title: “Systematic Abuse of Radial Blur”)

(Alt alt title: “The captain assured me that overloading the cargo hold would be perfectly saaaaaaHHh…”)

I’m very much looking forward to exploring how the setting and narrative impact of Starsector can be expanded through illustrations as Alex creates new mechanics and starts rolling in the higher-level campaign.

I’ll leave you with that; It’s going to be fun.

Comment thread here.

Travel Drive
Conceptually, ships use a different, much faster means of transportation for travel than they do for battle. Presumably, they either have to pop out of it every so often to navigate, or are forced to slow down by targeted nav system interference/debris fired across the projected trajectory/need to divert engine power to defensive systems due to the threat of enemy fire/<insert your favorite technobabble here>. Personally, I’m in favor of it being some kind of nav interference, whether it’s due to hypervelocity debris or jamming. The point is, a fleet can be forced to slow down and engage in battle.

On the flip side, a fleet disengaging from battle means that they’ve been able to re-engage their travel drives and put on speed. Maybe the nav computers finally calculated a safe trajectory. Maybe enough distance has been gained to eliminate the threat of enemy fire. The concept was always there, even if it wasn’t directly explored in the gameplay.

But now, there’s an opportunity to use it to improve the experience. With that in mind, the new mechanics:

• A ship or fighter wing being deployed into combat comes in with travel drive on, and it remains on for about 5-6 seconds – enough time to move around 3 grid squares in the command screen
• A retreating ship or fighter wing engages travel drive when it gets to within a couple grid squares from the border it’s heading for to make its getaway

A Medusa and a Wolf entering battle with Travel Drive (possibly pending cooler name) on.

This does a couple of things, all of them to do with reducing the impact of borders on gameplay. One, maps can be bigger without battles taking too long to heat up, because the extra space around the border is traversed very quickly using travel drive. This extra padding means objectives – which fighting often centers around – can be much farther away from borders. Two, retreating ships separate from pursuit far away from the border, almost entirely eliminating the “chase ship all the way to a border” situation.

Escape Scenario
When one fleet wants to run away but the other fleet doesn’t want to let them, what happens is a normal battle, but with modified rules.

First of all, regardless of whether it’s the player or the AI, the escaping side starts out fully deployed, a good ways up the map – 7 grid squares or so, less if the “harry” option was chosen by the attacker after a previous engagement (see previous blog post). The escaping side has to retreat at top border of the map, and the map itself is taller than the standard one.

Second, the pursuing fleet always deploys from the bottom of the map, but their frigates and fighter wings also have the ability to deploy from the left or right. This can be used to secure an important objective (such as a Nav Buoy; as you might imagine, speed matters here) or just to more quickly achieve contact with the retreating ships. The left/right deployment is far enough up the map that any ships being so deployed are slightly closer to the enemy than they would be if deployed from the bottom.

There are a few special cases here that need to be addressed.

What if the disengaging fleet is huge? Fleets beyond a certain fraction of the battle size can’t attempt to disengage – not maneuverable enough, let’s say – and have to fight. They can then attempt to disengage if they take enough losses.

What about ships that aren’t combat ready, what happens to those? They are also deployed, but can’t do anything other than try to retreat. They can’t fire their weapons, use their shields or ship system, etc.

What if the fleet is beyond the disengage size limit, but none of the ships are combat-ready? That’s very much an edge case, but the system still needs to handle it – let’s say it’s allowed to try to escape. It’ll likely get shot to pieces in the process, what with none of the ships being combat-capable, but at that point it’s not exactly drowning in options.

Initial deployment in the Wolf Pack mission

All in all, we have something that works like you’d imagine a chase to work – the attacker trying to pick off the slower ships, while the escorts for the retreating force try to delay pursuit long enough for some of their ships to escape.

Boarding
Boarding is tricky because if it’s too easy, it can trivialize the rest of the game. If the player can get a ship at the cost of a few marines, then why would they use other, more expensive means, such as buying or manufacturing one? Right now, the odds of being able to board a ship are low, but that’s not enough because there’s no cost to rolling the dice over and over until they come up “board”. Presumably, the player can create a fleet that wins against some type of enemy without any casualties (hence “no cost”), and then they could use boarding as a means to grow their fleet. Lowering the odds just makes this a more time-consuming process.

Now, time itself is a cost, especially when it starts to matter in the campaign – i.e. not only do you need more ships in an abstract way, but you need them now, to do something specific before it’s too late. But, they’d still get a random trickle of “free” ships from whatever combat do engage in while going about their business. It’s also difficult to put a value on time, since it changes so much – from not being worth very much early on to being worth much more later.

(I’m going to completely ignore reloading and trying until one gets the boarding results they want. The game is balanced for iron mode, where that’s not an option – there’s nothing wrong with save scumming, but it’s not something you can balance for, either.)

Ultimately, though, it comes down to this: I don’t think getting “free” ships has the right feel. Things are looking grim for the denizens of the Sector; it’s a struggle just to get by, and one has to fight for everything they get. The mechanics spitting up a shiny new ship every so often is at odds with that. (Ok, fine, not so shiny, but easy enough to repair.)

Still, I’d like to keep boarding around. It’s a neat concept, and it’s also a safety net because it ensures more ships are attainable even if they can’t be bought; either due to lack of stock, hostilities with the faction that would sell them, or some such.

With that in mind, boarding should be a way to get new ships rather than a way to get more ships. Thus, there’ll be a very high cost to successful boarding. Over time, this cost should be at least on par with, or higher, than the cost to acquire this ship by other means.

New Boarding Mechanics
One disabled enemy ship may be randomly picked for boarding after combat. (Currently, fighters can’t be boarded. Need to reconsider some mechanics around fighters anyway; stay tuned.) Presumably, the surviving crew on that ship have got some systems working and might be able to get it underway unless the winner interferes. Having more disabled enemy ships to pick from increases the odds of one of them being boardable, but only one ship can be boarded per encounter.

Once that ship is picked, the winner has these options:

Organize a boarding task force
Select ships that will participate in the boarding action. The ships need to be combat-ready and will lose additional CR. The ship selection determines the maximum number of crew and marines that can be sent in a boarding party, but also exposes these ships to danger if the enemy ship self-destructs.

Order nearby ships to engage
Do not board, shoot the ship down instead. Requires some ready ships, but there’s no risk, though the enemy ship might get away. The AI always picks this option if it’s available. Note that this leaves the player a chance to escape with a single ship, even after a total defeat!

Let it go
Not much reason to choose this at the moment, as with the “Let them go” option from part one, but other options might not be available.

About to try boarding a Venture. Wouldn’t want to be on that boarding party; the odds aren’t so good.

If the player decides to board, they get to choose how many marines and crew to send, and what type of strategy to use. Before making this choice, or, indeed, before even making the decision to board, the player can see roughly how many lifesigns there are on the ship about to be boarded. They can’t see whether these are crew or marines, though, and marines are drastically better at repelling attacks.

Hard dock and attack ship-to-ship
The best combat effectiveness for the boarding party and the highest odds of a successful capture. However, the ships involved could suffer serious damage or be destroyed if the ship being boarded self-destructs. The damage is based on the flux capacity of the ship being boarded, so larger ships have better odds of surviving the self-destruct of a smaller ship.

Launch assault teams from a distance
Poor combat effectiveness results in much higher personnel losses, but there is no risk to the ships in case of a self-destruct. The enemy ship also has a chance to make a clean getaway before the assault teams can land.

There are four possible outcomes to a boarding action: the ship is captured, self-destructs, gets away, or is captured but sustains too much damage in the process (and goes back to the “disabled” state). I mentioned earlier that on average, boarding a ship should have a cost higher than its value. Adjusting the probabilities of the outcomes and the cost of marines is an easy way to control that.

The player also has ways to reduce the cost. Using ships with higher crew capacity (such as the Valkyrie troop transport) means less ships exposed to danger. Fitting them with hull and armor enhancing mods further reduces the risk.

Overall, boarding is a series of gambles that stretches out over multiple battles. The player can affect their odds and how much they wager. Sustained boarding attempts aren’t going to be something a small fleet does on a regular basis, though. It’s more the domain of larger fleets where losing a ship or two (or suffering heavy marine casualties) to get a new ship back in the long run is acceptable. Smaller fleets focused on boarding could be successful, too, but would be more susceptible to bad luck if they try for something out of their league – such as trying to board a cruiser with smaller ships, which is just asking to lose all of them to no gain.

Comment thread here.

First, a quick recap of how things work now. When the player encounters another fleet, they can choose to attack them or leave. If either side wants to attack, the fleets engage, and each side has three options: “attack”, “defend”, and “escape”. Without going into too much detail, there’s a rock-paper-scissors mechanic there where attack beats escape beats defend beats attack. (“Escape” beating “defend” simply means a clean getaway, with “beating” in general meaning having an advantage in the battle, not automatic victory.)

Playing rock-paper-scissors vs the computer isn’t fun. The computer is either predictable or random, but in either case you don’t get the mind games that make it interesting vs a human opponent. The “escape” mechanic also doesn’t work well. The escapee has to run their ships across the map and retreat them off the enemy side, which is much more difficult than “attacking” and then retreating ships off their side of the map – which they could do without seeing a single enemy. If they do that, their ships take some automatic, random post battle damage to simulate a chase after this retreat, but that’s a problem in itself.  Either it does enough damage so that retreat from a real lost battle is disastrous, or it doesn’t do enough damage to stop “attack to retreat” being viable.

Adding in CR, the current setup is also open to being gamed. For example, a single Hound-class frigate fighting against an Onslaught-class battleship – the Hound could engage, then immediately retreat, causing a CR loss for both ships for being deployed in battle. However, the Hound both costs less CR to deploy and recovers it faster. So, it could reliably beat an Onslaught without firing a single shot but wearing down its CR.

There are also some concerns that would be nice to address given the opportunity. One such is that civilian ships – freighters, tankers, transports – don’t get to see much combat, because the “escape” scenario doesn’t work well. That’s a shame, because escorting/attacking those would add some new dynamics to combat, and the civilian ships have nice graphics that don’t get to see the light of day often. It’d also be good to see some attention being paid to outfitting them, rather than them serving as simple cargo/fuel/personnel capacity boosters.

Also, even if there was a reason to pick the “escape” option, part of what makes it not work well is that chasing down a ship often results in firing some last, desperate shots at it as it crosses the map border to escape. Borders in space? A necessity of the engine, perhaps, but not exactly a fact you want to rub the player’s face in repeatedly. Which this does.

New Mechanics
The most important consideration here is making CR impossible to game; that is to say, making it so that you can’t win a battle by getting the AI to waste CR and avoiding actual fighting. With that in mind, let’s dive into the new system.

When the player encounters another fleet, each side has the choice to either engage or attempt to disengage. If both sides want to fight, there’s a normal battle. If one side wants to run away and the other wants to fight, there’s something we’ll call an “escape scenario”. Multiple engagements are possible within the same encounter. We’ll get to the details of all that in a minute, but first let’s talk about the CR mechanics surrounding these choices.

Deploying a ship into battle reduces its CR, and ships with CR below a certain threshold can’t be deployed into battle. That’s the basic rule, and everything else is built around that.

So, let’s say both sides decided to engage, one side won, and one side lost. Any retreating ships from the losing side take an additional CR hit as they put on extra speed to disengage and push their systems to the limit. The intent here is to provide a real cost to retreating, so that it isn’t something the player does to, say, reload their missile ammo after a massive alpha strike, with the intention of coming right back for another round.

The winning side also has the following options (note that these options are available to both the player and the AI, whoever is the winner):

Harass the retreating enemy ships
This is an offensive option. The CR of the enemy ships left in reserve is reduced as if they’d been deployed into battle. (Presumably, they were called into action to cover the retreat of friendly ships off the field and/or had to scramble to avoid danger themselves.)

This takes care of strategies like deploying a single frigate to bait out as many deployments from the enemy as possible, and then retreating – if the player does that, the AI can pick this option and their reserves will still suffer the CR hit. The ships on the winning side will also be able to pursue and force an “escape scenario” battle if the losing side decides to disengage, and will have an improved starting position there.

Send out salvage teams
The middle-of-the-road, economic option. The winner gets improved salvage from any ships disabled or destroyed – on either side, during the last engagement only. They can still pursue if the loser decides to disengage, but won’t get an improved starting position.

Stand down
The defensive option. All ships that were deployed into battle regain a percentage of their deployment CR cost, with the percentage being lower for a harder-fought battle. The downsides to picking this option are the inability to pursue a disengaging enemy and the opportunity cost of not sending out salvage teams.

The number of enemy ships disabled or destroyed (rather, their combined deployment point cost) is used as  metric for how hard-fought the battle was. Other stats, such as hull damage taken or ship losses suffered, might make more sense conceptually – but would be more open to being gamed. For example, a ship that’s set up to avoid damage could then stand down to regain all of its CR. In practical terms, using “ships destroyed” works well to prevent harassment tactics from being used to drain CR, since it can’t be done unless ships are lost. This also takes care of any gamey CR-exhaustion-based “deploy and retreat” strategies; the winner can always pick this option to have some fighting capability left after the engagement.

The Hegemony Patrol is soundly defeated. (Graphics/text not final.)

So, the engagement is over, the winner picks one of these, its effects are applied, and we’re back to square one: each side can pick to engage or disengage. Let’s say the losing side decides to cut their losses and run. The winner now has the following options:

Pursue them
Only available if the winner has combat-ready ships remaining and did not pick “stand down” after their last victory. Also unavailable if the slowest loser’s ship is faster than the fastest winner’s ship. Starts an “escape scenario” battle.

Harry their retreat
Available even if pursuit isn’t, this deducts another chunk of CR from every ship in the retreating fleet. That way, even faster fleets have something at stake when a  more powerful enemy catches up to them, even if they can’t be forced into battle. The only time this option is unavailable is if the fleet that’s disengaging is the fleet that actually won the last engagement. That allows a clean getaway.

Another reason for this option is to avoid the situation where the player is forced to pick “pursue” so the enemy fleet has to deploy to escape (using up CR), and then exit out of the battle without deploying anything themselves, just to achieve the same effect this option has.

Let them go
The losing side gets away without any repercussions. Not much reason to let them do so at the moment, but it might come into play when reputation is an issue, or if you simply want to leave the enemy fleet combat-capable for some devious reason of your own.

After one of these is picked, the winner gets to pick between sending out salvage teams and standing down, and the encounter is over.

All in all, an encounter consists of zero or more engagements, and at most one “escape scenario”. I should probably talk about what that is… but this post is getting a little long, so I think I’ll split it into two parts. Stay tuned for part two – coming sometime next week, and talking about travel drive, the “escape scenario”, and boarding.

Comment thread here.

Comment thread here.

Option one is unquestionably faster and easier, but also seems likely to result in a tacked-on campaign because of the compromises that we’d have to accept to make it fit. Option two is more work, but is the one most likely to result in a campaign that’s a game in its own right, on par with combat. Do I even need to say which option I find more appealing? (Hint: it’s the second one.)

Enter combat readiness (“CR” henceforth), a mechanic specifically intended to improve the connection between the campaign and combat layers. There are already mechanics that link the two – for example, the persistent fleet, ship damage & repairs between battles, and character skills. So, we’re not talking about anything radically new; just taking stock of what’s there and cleaning it up, smoothing over the rough edges. While this involves some changes to combat mechanics, the goal is to enable the campaign to exert a greater influence on combat (and vice-versa), rather than to rework combat for its own sake. We’re giving the campaign tools to work with, levels to push, knobs to turn, analogies to abuse.

Conceptually, CR represents weapon and system maintenance and repair,  securing the magazines, making sure there are no cargo crates knocking about in the hold, that sort of thing. In-game, it’s presented as a percentage. A good way to think of it is as a “stamina” mechanic on the campaign level.

So, how does it work?

First of all, the maximum CR for a ship is determined by several factors, such as the crew level and the logistics situation in the fleet. A ship with 100% CR is a rare sight (and has significantly boosted stats), while values in the 50% range are more standard. A ship with very low CR can’t be deployed for battle at all, and deploying a ship results in a CR reduction afterwards. Recovering CR consumes supplies, much like repairs do. CR can also be reduced by campaign events, such as being hit by a radiation storm or suffering an accident.

What does this actually do for the gameplay?

Deploying ships now effectively costs supplies, and is a risk because those ships will need some time to prepare for the next fight. So, even if the player’s fleet outnumbers the enemy, there’s still incentive to get the job done with as small a force as possible. If the cost and risk are properly tuned, this can lead to challenging fights throughout the game, even if the player’s fleet has an advantage as a whole.

A couple of mechanics get streamlined, too. Crew level no longer has a direct impact on ship performance, but acts through CR. Refitting ships no longer has an entirely separate timer with obscure rules like “no hull mods take effect until a refit is finished”. Rather, a refit just knocks the CR down, based on how extensive it was.

In general, CR gives the campaign a way to affect ships without being too blunt about it. Doing hull and armor damage to a ship just isn’t right for many situations (i.e., a radiation storm or running out of supplies), and CR covers all those situations nicely. Systems knocked offline by a power surge? CR reduction! Not enough ball bearings to keep up with wear and tear on turret 23C? You get the idea.

What about the link going the other way, from combat to campaign? The obvious effect is that deploying ships costs CR, but there’s room for more nuance. For example, using missile weapons in combat could result in a greater cost, since they’ll now have to be reloaded (and, perhaps, missile weapons might not even be loaded when a ship is deployed at below a certain CR level). EMP weapons might decrease the CR of ships they hit.

CR also provided a good opportunity to address a long-standing balance problem. Ships like the Tempest or the Hyperion (both high-tech frigates) can be kitted out for speed and then take on arbitrarily-sized enemy fleets, provided the player is patient enough to spend a long, long time kiting and chipping away at the enemy. This takes skill to pull off, but is still a bad combination of “boring” and “effective” (video evidence here).

You want effective tactics to also be fun, and this particular hole in the design always bothered me. Now, this could be addressed by adjusting the stats of the offending frigates, but they are a lot of fun to tear about in! Taking what makes frigates distinct (i.e. speed) in the name of balance would hardly be ideal.

Enter some additional CR-based mechanics:

• Particularly low CR causes weapons and engine to have a chance of suffering a “malfunction”, which is the same as being knocked out by damage, without the damage
• Frigates only have a certain amount of “peak effectiveness” time in combat, after which their CR starts going down gradually

We had a slight engine malfunction, but uh… everything’s perfectly all right now. We’re fine. We’re all fine here now, thank you. How are you?

What this means is that frigates have a limited window of opportunity to do their best work, and past a certain point (when malfunctions start to occur) leaving them on the field is a big risk. Balance-wise, they can now be made faster without the fear of creating additional ships that can kite entire fleets to death. So, frigates become more distinct, as their strength becomes more pronounced, but also comes with a weakness that encourages more aggressive play.

CR also gives another way to differentiate ships. For example, some ships might be more effective in combat, but come with a higher maintenance cost and a high deployment cost. The Hyperion (again) comes to mind as a good candidate. It’s an extremely powerful ship because of its teleporter, but if it could only be deployed once before needing some time for repairs, it could stay dominant in combat without becoming the universally best choice in the campaign.

All in all, CR is a flexible mechanic, and lots of things fall under its domain. I should probably say that the details of the above are very likely to change as a result of playtesting and balancing. In fact, I can’t think of a single blog post where the mechanics outlined therein didn’t change at least a bit.

Comments
I’m going to try something new here and shift the comments over to a dedicated thread on the forum. Keeping the discussion in one place is good, but the main reason I’m giving this a try is spam. Right now, I’ve got it set so that a comment from a new poster requires approval before showing up, but I get an email *every time* there’s a spam comment in the queue. One time, a bot was set to post literally every two minutes. That was especially fun because I was away from the computer for a while and didn’t see it right away, and lots of emails accumulated.

I could turn off email notifications, but then it’d take me longer to get to legitimate comments from first timers. A conundrum! So, off to the forum we go – which has its own spam issues, but there the moderators help out, for which help I’m very thankful.

Right. Please leave your comments here.

* He actually got it sorted out pretty quick. No need to panic. I just have to express everything in terms of Starsector, eg. my first tech advice was “Transfer command!”. We have fun!

So yes: I’ve been digitally painting some large scale illustrations for use in Starsector, both as promo art and cropped down to little scenes to show for game event choices and so on.

Above: the final painting of the “Hound Hangar” faded into the first sketch.

Old games like Darklands (to the right, thanks Mobygames) and King of Dragon Pass used illustrations to show game events that were basically described in writing but not detailed gameplay simulations. I mean, say your game has a detailed tactical combat model: should you also put in a really boring wall-climbing minigame? (Answer: no. Though it’s funny how huge-budget modern games manage to cram everything possible in, to what one might describe as varying degrees of verisimilitude and/or success.)

Same idea here. Starsector has one artist — yours truly — so, for example, making a whole game mode where you’re a character walking around a space station really doesn’t make sense. This set of interactions can be expressed through some buy/sell type UI and illustrated event dialogs for if something special comes up. Makes sense, right? Let’s get back to the art itself.

When drawing these things I enjoy creating a new layer each time I sit down for a drawing session or make a major change. This lets me show the progression of the illustration over time. And now I shall show you, dear reader, this very process!

Above: very first sketch. So what are we doing here … how about some kind of hangar, that’d be cool.  Dramatic perspective, naturally. Choosing colour palette, rusty oranges and browns for that ‘completely not antiseptic and perfectly maintained’ universe of Starsector. Think Millenium Falcon in Mos Eisley. Serenity in … well, wherever Serenity went. But more rust, even if in space there’s no oxygen to actually oxidize what probably isn’t even iron armour. That’s not the point!

Moving on.

Here I’m filling in major areas of colour and lighting. Off to the left, cargo storage: still a warm colour, but something off from the orange going on in the main hangar. Above, florescent-green lights in the hab-interiors of the ship giving both an unnatural yet somehow life-filled glow. In the ship at dock, some bright cold blue lights to offset it from the rest of the image.

Rendering and details are filling in here. You can see I’m slowly botching the perspective on the ship. At some point around here I posted an in-progress image on twitter. Naturally, this image was dissected in far too much detail in the forums.

I admit, it’s slightly troubling to have my work picked apart in excruciating detail as if it were some canonical guide to the game universe and the exact scale of ships vs. people. My answer: it absolutely isn’t and doesn’t really matter! I’m just drawing stuff.  (But do pardon my defensiveness; I appreciate that the motives at work here are an enthusiasm for the game and a desire to explore the universe of Starsector more closely. This is actually awesome.)

But that said, it still bugged me enough that the scale was out of whack that I changed the size of ship by extending the drawing over to the right. There! It’s more like a proper Hound now. Kinda. Perspective is utterly bollocksed now. And even for that, there’s more detail going in all over.

Apparently I love that shade of orange in the background of the main hangar; I used a similar effect in my art for the Dungeons of  Dredmor title painting. I even had a tube of paint in that colour in art school that I abused thoroughly. I just realized that I keep coming back to it. Neat eh?

Alright, let’s fix this perspective. There: aligning things to be one-point rather than two-point.

I actually had an overlay layer of perspective lines drawn to use as a rough guide but as with all painting I tend to play fast and loose with the rules. I’m a rebel like that, you see. Or, more honestly, I can’t be terribly bothered to make the drawing a perfect CAD render because that’d take forever. I think it’s more important to draw the feeling of an image, whatever I’m trying to evoke, rather than holding with unnecessary fidelity to (anyway) made-up schematics. Admittedly, if I were doing this properly as many digital painters do I ought to block out the rough forms as 3d models then paint overtop a rendering. Maybe someday.

Yet more detail throughout, especially darkened the roof, added that girder lattice to emphasize the perspective, and added some spotlights to set it off.

More rendering and cleanup, especially perspective on the left. Made the big yellow supply crate into two smaller crates that might have a chance of actually fitting into the ship’s cargo bay. Scale, see – I’m paying some attention to it!

And so on.

At this point, Stian contacted me about some kind of Thing being cooked up which would be well served by having the image be at a higher resolution. So I said, sure, let’s rock and roll. Image -> resize.

Here I’ve up-sized the image significantly and filled in the extra space at the top and bottom with Photoshop’s “content-aware fill” which guesses what the selected area should be filled with based on what similar areas elsewhere in the image are adjacent too. It does a serviceable job at the very, very broad strokes and I’ve been enjoying using the content-aware fill as a starting place for when this sort of thing comes up.

And here we are. Lots of clean-up done for the up-sizing as well as tweaking of various details. It is still, you may note, not at all perfectly rendered down to each pixel. Lots of heavy brushwork. This is intentional because 1. it’d take forever to draw this pixel-perfect or photorealistically (and there’s so much more to do) and 2. this illustration will generally be shown at much lower resolution when used in game. And probably even with portions cropped for different situations in-game.

I’m looking forward to doing more of these, maybe not all quite so large. I’ve always loved science fiction illustration since I can remember and now I get to draw a whole bunch of them! What more could I ask for?

Just for fun, I’ll post a crop from another one I’m working on:

Onward to what’s new in 0.54.1a. It’s a bugfix and polish release, much like the other .1 releases that follow up a major one. Unlike those, though, it’s taken over a month to put out. That’s the case for two reasons. One, the name change took some extra doing. Two, the 0.54a release was stable enough that there was no urgency in putting this version out – so, we’ve been able to work on the campaign design and do some prototyping along the way, laying down some groundwork for the next major release.

The major items in 0.54.1a are:

• Redesigned character screen (screenshot)
• New “emp arc” mechanic for the Tachyon Lance and the Ion Cannon, improvements to the High Energy Focus ship system
• Lots and lots of bug fixes and modability improvements

You can find the full patch notes here. Please download the new version using the links below.

Alternate download links: Windows Mac Linux

The 0.54a version of Starfarer is now out! The main features in this release are:

• Character skills and aptitudes (19 skills and 4 aptitudes to choose from)
• Redesigned fleet command interface
• A slew of new features available to modders

You can see the full patch notes here.

Download the new version using the buttons below – you’ll need to reinstall the game, but shouldn’t have to enter the activation code again.

Alternate download links: Windows Mac Linux

For this written interlude I shall discuss some of my thoughts around drawing Starfarer faction icons. I shall give you my lore disclaimer right now: the story & background of Starfarer is not something I make up, though I have certainly contributed thoughts to Alex and Ivaylo on occasion. Unless explicitly stated otherwise, anything I say here about the setting of Starfarer is going to be what goes on in my head while I’m drawing rather than what is any sort of canon which will appear into the game and which you should read closely for overarching literary themes.

So: the factions in Starfarer need identifying symbols, colour themes, and a sort of general ambiance – and these all must support one another. One faction may be warlike and value strength; this must be shown in how they represent themselves. Another may be peaceful and value stability; likewise, it probably wouldn’t do if their icon was a burning red skull menacing with spikes. Pirates could be into that though, because they’d like nothing more than to appear scary and ruthless. I’ll say again, each faction has a few core themes that need to be shown with both symbols and colours which support said themes as well as the setting of Starfarer in general. It’s futuristic heraldry, if you like.

Now let’s have a look at some sketches and the process of taking a faction icon to a presentable state.

(Click on any image to enlarge it.)

My faction icon sketch sheet

Here’s where I threw down some ideas and pushed them around a bit. Some more than others. We’ve got a small set of factions to start with which I think Alex gave me a list for at some point as well as some background material floating around in a shared document somewhere. And of course you’ve met a few of these factions in the game already, so those were my starting points.

Hegemony

Some ideas feel right the first time you draw them. I’ll have you know that this is the exception rather than the rule.

I see the Hegemony as something like a lost Roman legion (which their background story resembles) – they are practical, aggressive, and value strength due to their military origin but also aspire to the glory of a lost empire. They’re not unlike any number of European nation-states that saw themselves as successors to Rome who borrowed the symbolism of said empire to identify themselves with its strength. Hence the eagle.

The circular element? Oh, I don’t know. Maybe it looks futuristic, maybe celestial. Things divided into threes seem futuristic to me and I just realized that I used it all over the place in these faction icons. I also pulled the wings back a touch from the geometric look in the original sketch because I didn’t feel they were working very well there; too much geometric greebling (like any number of the ships I draw).

The Hegemony

So there you go. Some kind of Eagle or Falcon — oh, hey, what are the names of the cruiser hulls that the Hegemony uses all the time? Right; you see what happened there.

… and I’ll probably touch up the talons a bit …

(Why orange? No one uses orange. Except the Dutch, I guess. Someone should use orange.)

Tri-Tachyon

These guys are The Company from Alien, one of the zaibatsu from Neuromancer or insert-cyberpunk-title-here. The ruthless corporation, profits above people, etc.

But it should be a bit more complex than that. They’re not stupid. They are wealthy beyond imagination, for what that means in a world of chaos. They value elegance. They have taste. To harken back to that 80′s-90′s theme in fiction of eastern corporations taking over the west (see again: zaibatsu, Weyland-Yutani), they have an almost zen corporate ruthlessness (as opposed to something much more crass and American like Enron).

I looked to Japanese Mon, the clan symbols which were the equivalent of European feudal heraldry but, to my eyes, much more restrained and sophisticated in design. Some are even corporate logos to this day — see Mitsubishi. (And the symbol of the Hojo clan may be a very familiar symbol indeed to gamers.)

Tri-Tachyon Corporation

So: Simple, elegant, restrained, cool, division into three, a round element, the implication of high technology and sophistication. This took a lot of iteration and maybe I’m still not happy with it, but there you go.

Anyone else want one of these on their desk?

The Coalition

Ah, this is a new face, isn’t it.

Information is light, but these guys are basically a coalition of worlds that have joined together for mutual defense. They’re interested in trade, prosperity, trying to build something from the rubble. Maybe they can do it, maybe they’ll be torn apart from a strong enemy from without — or perhaps from fighting within, where one member puts their own interest ahead of the others to the downfall of all.

No doubt you, the player, will help determine this outcome.

The symbolism here is pretty obvious: shield, ring of celestial spheres, unifying star representing force/power/command. This design came pretty quickly. Plus rounded future-corners, of course.

The Coalition

And gold — why not? That’s what they’re all in it for.

Pirates

Pirate flags are fun. And they’ve got a rich tradition of naively executed brutal imagery not so unlike metal album covers. (Not knocking them; I’ve got a few myself). Red and black are obvious colours for a threatening faction, maybe too obvious, but pirates aren’t known for subtlety.

The question boils down to how traditional do you go with a skull flag? This is space, not the high seas. So I tried for a skull-like space helmet, and oops, did the division into three again. And spikes. But not horns, that would have been too easy.

Space Pirates

I’m actually not entirely happy with this one, just something about the lines and shapes of the bits at the top, the interaction of the eyes with the cuts, and the mouth-piece seems off. I’ll give it a few more goes. (And yes, I realize that I’m sounding like a typical perfectionist artist type by saying I’m unhappy with every single design as soon as I show it.  Comes with the territory.)

And it should be said: Pirates are not known for being unified as one group. It may well be appropriate to crank out a big pile of variations on skull flags to represent all kinds of different pirates. (This will be fun!)

Cult of Lud

I’ll talk about this one even though I fully intend to revisit it.

The theme of this one gives me some trouble. I know what it’s going for, something like a faction of deep-green militaristic primitivists who want to return everyone to a state of nature – by force. Not quite the Sid Meier’s Alpha Centauri Gaians, not quite the Sid Meier’s Alpha Centauri expansion’s Cult of Planet, but they all probably hang out at the same coffee shop, with Lud being the creepy one that sits in the corner polishing their guns.

Lud

So with Lud I was thinking of something both natural and spacey, both hippy and Starfarer. The double-ended tree didn’t quite work (and is an existing symbol for something or other, I discovered), so that is out. I was finally taken with imagery that alludes to cosmic swirls, like galactic arms and nebulae. Though I’m sure no one particularly shares the viewpoint of Lud, I’d imagine the world as they see it being a whole interconnected being with its own greater needs whose importance outweighs the destructive hubris of human technophiles.

It needs more work to come together, probably needs to be kicked up a notch or three in aggression — hippies simply aren’t militaristic enough for these guys — and I think some of the contradictions in the faction need to be worked out in my head for a design to work.

… And then, after writing a draft of this post, I emailed Alex about all this and he said that rampant environmentalism was not an essential feature of Lud, so I’ve kinda been running with the whole idea in my head the whole time. Whoops. So for this design I’m going to take a step back, re-interpret the themes, and play up something that’s more about an apocalyptic cult lead by a charismatic messiah figure which holds the excesses of technology (and some other stuff, likely) to be the cause of the downfall of Humanity … and these are the End Times and so on and so forth.  More cult, less nature.

So that’s an artists’ work and my work is cut out for me!

Hope you enjoy seeing your favourite factions in-game (and blowing up your most hated factions, at that). I’ll make sure they’re pretty, either way.

The very first version of the UI used the standard RTS model – control groups, right-click to order ships around, etc. It didn’t work very well – there’s a strong incentive  to keep checking on how your ships are doing, so that you can adjust if they’re doing something you don’t like. Optimal gameplay was, then, constantly interrupting the flow of combat to open up the map, check on your fleet, and tweak their orders.

The next version – the one that’s in the current release – solved that problem by limiting how many orders can be given to ships (via “command points”) and adding the concept of “assignments”. Instead of ordering ships about, you’d create tasks – capture this, defend that, rally a carrier here. The ships would then work out the details on their own. You could also give a few direct orders if you saw the ships doing something undesirable.

This worked much better – you could create an initial set of assignments at the start of the battle, and then just focus on the combat, only occasionally adjusting them. Because you couldn’t give unlimited orders, you were freed from the burden of having to constantly give orders to feel like you’re playing optimally.

The new approach had some issues, though. When it worked (the AI doing the right thing in “working out the details”), it worked well. When it didn’t, it could be frustrating trying to fix it using the limited direct orders.

The bigger problem was (and, I suppose, still is) accessibility. RTS-like controls are the go-to assumption when one sees a map with units on it – but they didn’t work. If you clicked on a ship, hoping to tell it to do something – you couldn’t!  You’d be presented with a context menu that let you create assignments that target that ship – i.e., you could tell your fleet to escort the ship you just clicked. You couldn’t tell that ship to escort something else, not without creating that assignment first (and then using a direct order from the context menu to assign the selected ship to it – a somewhat clunky process).

Telling the game what you want done – i.e. creating an assignment – is reasonable, in the context of commanding a fleet. An admiral wouldn’t tell every frigate in the fleet exactly what to do – that’s the job of his subordinates. But being reasonable, as it turns out, doesn’t get you far when going against UI convention. Games don’t typically ask “what do you want done”, they ask “how do you want to do… eh, whatever it is, I don’t actually know/care.”

The new version – that’ll be in the next release – aims to combine the best aspects of the two approaches.

The assignments system remains in place, but RTS-like controls can be used to interact with it. For example, if you select some ships and right-click on an assignment, the ships will be assigned to it. If you right-click on an empty space (or an objective without an assignment), the game will automatically pick what it thinks an appropriate assignment. A single fighter or frigate right-clicked on an objective will result in a “capture”, a larger group of ships on an enemy-held objective will create an “assault”, right-clicking on a friendly ship will order an “escort”, etc.

It will undoubtedly get it wrong sometimes – this is fine, since you can always change the assignment by clicking on it and selecting a different type. Creating assignments by clicking on the target (not the actor(s) that will carry it out) is still the preferred way.

Command points also remain, but work a little differently – creating an assignment or giving a direct order opens up a “command channel” for a few seconds (of game-time, the channel will stay open indefinitely while paused), or until the command UI  is closed. While the channel is open, you can give direct orders for free. Creating assignments still costs a command point, though.

In other words – spending a command point gives you something akin to a “turn”, during which you can give orders. The number of these “turns” is limited, but what you can do during one isn’t.

As usual, the devil is in the details. “Left click to select, right click somewhere to give an order” sounds great, until you consider that in the current build, right-click and drag moves the view around. Oops! So, how do we get around that?

One option is to introduce a new way of panning – the way RTS games do, by moving the mouse to the edge of the screen. I did actually add that, but I’ve always found it a bit awkward to use in RTS games  (lots of moving the mouse around to pan a hard-to-control-exactly amount), and wanted to keep right-click-and-drag as well. It’s something that all of the UI supports – weapon lists, the campaign map, the trade screen, etc – you can right-click and drag to pan/scroll, everywhere. Removing that as an option from the command UI felt wrong.

The answer was to take a right-click action when the mouse button is released, instead of pressed. If the mouse moved more than some threshold between the press and the release, it’s a pan. Otherwise, it’s a right-click. Sounds good, right? But, when you’re giving orders quickly – click to select a ship, move the mouse across the screen to right-click on an objective – it would oftentimes interpret that as a pan, just because of the inertia of your hand moving when you right-clicked. The same amount of movement done over a longer period of time would be intended by the player to pan the view, though. Considering whether the interval between the button press and release was short – and increasing the movement threshold if it was – helped things along.

Left-click had its share of challenges, too. You can select ships  - but can you clear the selection? You might want to, because when you have stuff selected, an accidental right-click could give unintentional orders. So, how? Clicking on an empty space seems like the most intuitive way to do it (“click on nothing to select nothing”), but that already has an action attached to it – it creates a waypoint. The answer was simpler this time – a click when the selection is empty creates a waypoint. Otherwise, it clears the selection.

Getting the little details right is what’s going to make the UI feel polished.  There’s still some playtesting to do – and I’d like to improve how the ship AI actually follows orders - but for the first time, I’m happy with the way the command UI feels. It doesn’t just get the job done, but also isn’t fighting me along the way. A lot of that has to do with improvements to the assignment system, not just streamlining the controls – but the details of these could probably make for their own post, so I’ll refrain from getting into the details.

Since you’ve stuck through to the end: here is a screenshot of the new UI. Please pardon the placeholders for a couple of the icons. Yellow arrows indicate direct orders, green lines are automatic assignments. The letters on the buttons are keyboard shortcuts.

Burn drive on the Tarsus… whose bright idea was that?

And now, if you’ll excuse me, I’ve got a fleet command tutorial to rewrite. Grumble.