First time experiences with 0.24 - UI/UX feedback

[ringerc]

Hi folks

I just bought KSP (then spent the whole weekend building rockets and exploring). I'm delighted by it, WIP-status though it is - it's already a 4.5 stars game and it's not even done yet.

I know that when you're looking at something all day you don't see rough edges and counter-intuitive UI elements because they're obvious to you. So I thought I'd share my first-48-hours experiences.

Overall, given the complexity of the game the UI is a wonder of usability. I'm really impressed. I had issues, but not nearly as many as I expected.

(In the following I've left out things that I saw in the "already suggested" and "what not to suggest" pinned topics, but this isn't a feature request list so much as a list of things I found counter-intuitive, confusing, or hard to figure out from a UI and gameplay perspective)

Things I got really stuck on and had to dig around on the 'net for help were:

• Why my probe just stopped responding to controls. (Out of batteries, hadn't invented solar yet). The UI should respond to attempted control input in a way that indicates that the craft can't respond. Showing a partial pulldown of the resources menu and flashing the power bar would be very helpful.
  
• How to collect the kerbal astronaut in orbit once I'd mostly matched orbit with him. Having to press the [ key to switch control to nearby craft isn't discoverable or intuitive. Should be able to click on craft within range to switch control via menu. If it's out of range or not remote controllable the disabled menu text should say so.
  
• Recovering boosters. Or rather, that you can't. The tutorial needs to cover this IMO, I got really frustrated wondering why my boosters with their adequate parachute loads still seemed to just vanish. Especially when the description of the booster talks about it being designed to be recovered...
  
• EVA controls. It's easy to get into EVA, but what then? It doesn't help that the kerbal sometimes likes to leap away from the capsule when you exit, and did when I first tried EVA. So you have to figure out how to enable RCS on their suit (and that you have RCS on all EVA suits) and how to fly them back to the capsule. This is the kind of situation where situational help would be good - "Hey, this is your first EVA, want a hand?".
  
• Docking. I had to hit the 'net to find out about docking mode and how to use my RCS for precision alignment. Still have no idea how to yaw/pitch when in docking mode, without switching out and using normal flight mode controls for pitch/yaw.
  

Things I worked out myself, but that were barriers, were:

• How to recover a vessel once it's landed, and that when a mission is done you can just go to the space center then visit the tracking station, there's no explicit "end of mission" unless you crash and die horribly. The orbiting tutorial should be extended to cover landing and recovery.
  
• My capsules plummeting to the ground and exploding after their chutes were open and I was on time warp. Problem was that in time warp, especially with radial chutes, the physics seems to break quite easily at the moment of chute opening. Should disable/pause time warp for parachute opening.
  
• When you accidentally enter cockpit view, getting out isn't obvious.
  
• So, those lander legs, they work better if you extend them! (They're pretty handy even when they're closed, as my first Kerbal VTOL lander demonstrated). It's non-obvious how to open them.
  
• Navball "target" mode. I think I've worked it out, but some docs on it would be really nice. Especially since it snaps into target mode when you approach something you set as a target, and suddenly prograde/retrograde aren't *orbit* prograde/retrograde anymore but they look the same...
  
• Orbit matching the rescue kerbal. Got there in the end, but a tutorial for matching orbits would be nice, and could be combined with navball target mode and with EVA flight controls
  
• Radial symmetry in construction mode. Awesome once you figure it out.
  
• Tweaks in construction mode, for changing booster fuel, etc.
  
• Subassemblies in construction mode.
  

Pretty sure I found some bugs too:

• Electrical charge level resets to 100% on a vessel without solar panels, so short lived probes can go longer than they should. Sometimes. Seems to be to do with returning control to flight center and launching other vessels or switching control to others. Possibly also reverting to design after crash? Haven't pinned it down yet.
  
• On my system (Intel graphics, Fedora 20 Linux x64) the labels above the crew pictures are unreadable, there's no visible text in the buttons for EVA etc.
  
• When constructing vessels using symmetry, if you build a ring of boosters then want to put something in the middle, the central object won't snap to the surrounding ones.
  
• Same as above, when trying to put an object inside a ring of boosters (big fuel tank, big liquid booster, whatever) sometimes it seems to work, but actually you've created one copy of the object for each radially symmetrical booster. The inner objects overlap and ignore clipping, taking up the same space. I had six LV-12 boosters overlapping so they looked like one earlier. (In case you're wondering why I'd do such a thing: I was trying to get an LV-12 booster to escape velocity for a contract, so it was payload. Which meant a ring of huge multi-stage boosters to get it up there.)
  
• Sometimes when I'm trying to place a reinforcing strut it seems to disappear inside the structure of the rocket. It still exists, but you can't find it (there's no parts list to manipulate separately).
  
• When trying to use multiple vertical decouplers for structural integrity on large booster stacks only one of them would snap the two objects together correctly. The other one wouldn't "click" to the other side.
  
• The game puts gantries at the rocket stage they're connected to, even though it never makes sense to put them anywhere but on the first stage. It's easy to rearrange the stages, but annoying if you don't notice and the rocket goes nowhere. Or BOOM.
  

Phew!

Finally, some more general comments.

• Career random mission rewards are so lavish that there's little incentive to go exploring except for fun, and you tend to get technology trivially easily by doing ridiculous things.
  
• Random missions need a filter for absurdity. Can I launch an LV-12 integrated liquid booster stage to escape velocity and fire it? Yes, I can, I did it, but why would anybody do it? (Restrict first-stage heavy booster rockets from requests to test them in orbit / at escape velocity).
  
• Drag is given as a number, but it's not clear what it means. If I put an aerodynamic nose cone on my big flat booster top will that add 0.2 drag? Or reduce net drag? (I suspect this just isn't implemented/finished).
  
• Why are there rocket adapters when I can just connect a big rocket stage to a small one and vice versa?
  
• If my vessel will explode violently the moment the physics engine starts, is there any chance of warning me "hey, this might want a gantry"? Not that I mind the light show, but it gets tiresome having to go back and add a gantry. And I'm forgetful.
  
• Motors should show burn time and total impulse, both in the specs for the default and for a given fuel and thrust limit when tweaked. It's a pain working it out if you want to sync up some solid boosters with a liquid core or similar.
  
• It's really hard to work out stage weights and total impulses to sanity check whether a design might work without flying it.
  
• Why do control system torque rings seem to help with pitch/yaw too?
  

... and after all that, again, thanks for an awesome game. I'm still so hooked it's hard to stop playing before the birds are waking in the morning.

[Joonatan1998]

"So, those lander legs, they work better if you extend them! (They're pretty handy even when they're closed, as my first Kerbal VTOL lander demonstrated). It's non-obvious how to open them."

I also didn't realize that about lander legs at first, I thought that the small ones were for really tiny probes and that the medium ones were meant to be used with the manned pod.

[ringerc]

I just flew a manned pod with the small legs on top of a small liquid fueled engine. Took it hopping around KSC doing short VTOL flights. I did very, very careful low horizontal velocity landings and worked on the principle that any landing you walk away from is a good landing.

I didn't have ladders yet, so once the pilot hopped out he couldn't get back in anyway.

[kujuman]

To several of the first comments, especially those about controls: had you read through the "Input" tab in the Settings menu? If so and the labels didn't make sense (at least not without the context of having played), that's one thing, and different from not looking up the key bindings (of which we are all guilty ). Were you aware that probes required electric charge (1.7 EC/minute or whatever) on the expanded VAB part menu?

• Career random mission rewards are so lavish that there's little incentive to go exploring except for fun, and you tend to get technology trivially easily by doing ridiculous things.
  
• Random missions need a filter for absurdity. Can I launch an LV-12 integrated liquid booster stage to escape velocity and fire it? Yes, I can, I did it, but why would anybody do it? (Restrict first-stage heavy booster rockets from requests to test them in orbit / at escape velocity).
  
• Drag is given as a number, but it's not clear what it means. If I put an aerodynamic nose cone on my big flat booster top will that add 0.2 drag? Or reduce net drag? (I suspect this just isn't implemented/finished).
  
• Why are there rocket adapters when I can just connect a big rocket stage to a small one and vice versa?
  
• If my vessel will explode violently the moment the physics engine starts, is there any chance of warning me "hey, this might want a gantry"? Not that I mind the light show, but it gets tiresome having to go back and add a gantry. And I'm forgetful.
  
• Motors should show burn time and total impulse, both in the specs for the default and for a given fuel and thrust limit when tweaked. It's a pain working it out if you want to sync up some solid boosters with a liquid core or similar.
  
• It's really hard to work out stage weights and total impulses to sanity check whether a design might work without flying it.
  
• Why do control system torque rings seem to help with pitch/yaw too?
  

1) Career mode is very new, so there's only basic balance on it so far. It'll get better probably, but I think the intent is that career mode is more of a way to get your program up and running in a semi structured way, and late game it's basically sandbox with some cash and science. So contracts are a means to an ends, not an ends in and of themselves.

2) Absurd missions are fun XD Although occasionally impossible contracts have come up (not sure if it's still an issue in .25, but when contracts first came out, people might get contracts to test the launch clamps somewhere other than landed on Kerbin. Yeah, impossible ).

3) Drag's been a placeholder basically forever. The actual force of drag from a part is also partly mass based, which is why a ship of all .2 drag parts won't automatically orient during free fall.

3) Style

4) Since one gets 100% refund of any launch clamps (since they are landed at the launchpad), just always add a few of them until you get an idea for what won't need them.

5) There are mods to do this, otherwise you can do it by hand/spreadsheet. This is due to the game's design being more guess and check experimentation than exacting measures (boosters found by the side of the road and all). If Squad does think basic calculators should be in the final game (I don't know where that discussion is ATM), they'd probably wait until it's almost ready. No point spending time now adding features a few very well maintained mods already add. Further, with a bit of experience, you should be able to get a feel for what will work. If in doubt, add more boosters Addendum: they do currently have thrust ratings and fuel consumption. Not sure by what you wrote if you're familiar with the extra detail menu in the VAB.

6) See 5

7) They seem to do so because they do. The torque SAS modules are reaction wheels. So rotation about all three axis is appropriate since we can assume there is more than one wheel.

The Mk2 Drone Core's (new in .25) model features the 3 axis' wheels.

Edited November 3, 2014 by kujuman

[ringerc]

Thanks for the detailed responses, I appreciate it.

To several of the first comments, especially those about controls: had you read through the "Input" tab in the Settings menu? If so and the labels didn't make sense (at least not without the context of having played), that's one thing, and different from not looking up the key bindings (of which we are all guilty ).

It's one thing to know which keys to press - after all, they're the same keys you use for controlling your vessel normally - and another to understand how they *work* in the context of the completely different model of control in EVA flight. Everything behaves differently.

You also have to know that EVA suits have RCS and that you have to turn it on.

So it's not a matter of "I can't find the controls" so much as "er, I'm drifting away from the ship and none of the controls do anything, what do I do?" then after figuring out that the suits have RCS, "auuugh, everything flies crazy!".

Were you aware that probes required electric charge (1.7 EC/minute or whatever) on the expanded VAB part menu?

Yep, so I figured out why it'd stopped responding after a while.

It'd be nicer if the UI told you why it was ignoring control input, that's all. If you're out of fuel your stage diagram shows empty fuel bars. If you're out of power, there's nothing visible until/unless you think to pull down the resources menu.

I'm talking about things that reduce the "friction" for new players, the "um, ..." moments.

1) Career mode is very new, so there's only basic balance on it so far. It'll get better probably, but I think the intent is that career mode is more of a way to get your program up and running in a semi structured way, and late game it's basically sandbox with some cash and science. So contracts are a means to an ends, not an ends in and of themselves.

OK, that makes sense. I've since read a bunch of the forum posts and gathered that impression as well. It'd be nice if it was labeled "Career (beta)" mode.

2) Absurd missions are fun XD. Although occasionally impossible contracts have come up (not sure if it's still an issue in .25, but when contracts first came out, people might get contracts to test the launch clamps somewhere other than landed on Kerbin. Yeah, impossible ).

Yeah, it's true that some of the ridiculous ones can be fun challenges too - I did enjoy the challenge of getting a monstrous booster stage to escape velocity.

I guess if the financial constraints were more real, so you couldn't solve every problem with "build an even bigger rocket" it'd be cool to keep the ridiculous ones. You get to judge if you can actually achieve them or not. Right now, it's a matter of "build a bigger bunch of asparagus". (I love that term!).

3) Drag's been a placeholder basically forever. The actual force of drag from a part is also partly mass based, which is why a ship of all .2 drag parts won't automatically orient during free fall.

Right, so again it's a "that's not finished". Fair enough; be nice if that was somehow visible in the UI so it wasn't quite so confusing trying to figure out what's going on when you start playing though.

3) Style

Fair point. It's certainly prettier. I guess when drag is implemented properly it'll matter more too, big flat surfaces and sharp edges create horrible turbulence.

4) Since one gets 100% refund of any launch clamps (since they are landed at the launchpad), just always add a few of them until you get an idea for what won't need them.

That's what I started doing. It's pretty minor really. Also, it's actually hilarious when rockets go boom, the main issue is that my poor pathetic laptop can't handle it so I get a slide-show explosion.

5) There are mods to do this, otherwise you can do it by hand/spreadsheet. This is due to the game's design being more guess and check experimentation than exacting measures (boosters found by the side of the road and all). If Squad does think basic calculators should be in the final game (I don't know where that discussion is ATM), they'd probably wait until it's almost ready. No point spending time now adding features a few very well maintained mods already add. Further, with a bit of experience, you should be able to get a feel for what will work. If in doubt, add more boosters Addendum: they do currently have thrust ratings and fuel consumption. Not sure by what you wrote if you're familiar with the extra detail menu in the VAB.

Yeah, I use thrust and consumption data from the extra detail menu, it'd just be nice to have the option not to have to do quite so much manual calculation for the simple and obvious repetitive stuff, so the brain work can be left for the "interesting" bits of rocket design.

I have a notepad with total burn times and total impulse from the engines I use, for example, but it'd be handy to have that in the UI. Adding this sort of thing through the research tree and/or difficulty options would be nice. Blind experimentation is awesome fun for getting to orbit and the Mun, but gets pretty old by the time you're trying to get five-stage probes to the planets.

The lack of a parts manifest for your rockets (per-stage or otherwise) makes it harder too, because you have to visually hunt around to figure out what the components are. Especially when using subassemblies where you might not immediately have the manifest for it to hand. And if there was a manifest, it'd be a pretty un-fun part of the game to go down it with a calculator looking at the parts details and adding stuff up. Why not hide orbital tracks and make us do all the calculus for those too? (OK, that'd be kind of cool in hard mode, in an infuriating kind of way).

7) They seem to do so because they do. The torque SAS modules are reaction wheels. So rotation about all three axis is appropriate since we can assume there is more than one wheel.

Aaah. OK, that makes sense. The flat little model had me envisioning a single big wheel for rotation around the main axis.

I wonder if part descriptions should mention that they aid control in all three axes, or if that's just a "well, duh" issue.

Edited November 4, 2014 by ringerc
Fix quote formatting