Putting the 'role' back in role-playing games since 2002.
Donate to Codex
Good Old Games
  • Welcome to rpgcodex.net, a site dedicated to discussing computer based role-playing games in a free and open fashion. We're less strict than other forums, but please refer to the rules.

    "This message is awaiting moderator approval": All new users must pass through our moderation queue before they will be able to post normally. Until your account has "passed" your posts will only be visible to yourself (and moderators) until they are approved. Give us a week to get around to approving / deleting / ignoring your mundane opinion on crap before hassling us about it. Once you have passed the moderation period (think of it as a test), you will be able to post normally, just like all the other retards.

Kerbal Space Program TWO, DOS, DEUX, ZWEI....

dbx

Arcane
Patron
Joined
Dec 14, 2009
Messages
3,897
Meh:
1. Electric charge still measured in funny units. (Us to be exact on batteries, but EC/s on solar panels).
That kinda make sense, what KSP2 lack is a simple way to display total EC production and consumption (or at least I couldn't find it).

KSP2 also has another useful thing, powered time warp for electric rockets which you could have on KSP1 only with a very bugged mod.
 

Hellraiser

Arcane
Joined
Apr 22, 2007
Messages
11,353
Location
Danzig, Potato-Hitman Commonwealth
This needed at least an extra month or two of polish at least to squash some bugs if not also improve performance, but I guess (((suits))) gonna suit with one month to T2's fiscal year end.

I also have to say the technical state of the game at early access launch is very Kerbal and the EA release trailer song makes the whole thing quite meta :M

If it either ran well or had only minor bugs I would have said it is clearly visible it will be a good sequel with visible improvements in production quality or QoL, and that the content already in is decent. At the moment though it is a rather tiring experience at least for me and there's no telling how soon the needed hotfixes come, let alone the major fixes or the major content updates.

Probably best to wait for the SCIENCE! roadmap milestone to release first before buying if one is not completely opposed to early access. The low hanging fruits of performance improvements should be taken care of by then as well as the most obnoxious bugs, and for the other stuff modders will fix it(tm) in the meantime if the devs won't manage to, especially regarding UI shortcomings or missing parts. And it also should be clear by then if the devs know what they are doing and how much fault is theirs and how much the publishers.
 

RobotSquirrel

Arcane
Developer
Joined
Aug 9, 2020
Messages
1,961
Location
Adelaide
This needed at least an extra month or two of polish at least to squash some bugs if not also improve performance, but I guess (((suits))) gonna suit with one month to T2's fiscal year end.
I think it needed way more than two months. Looking at their roadmap this thing was always going to be a disaster. Now imagine if they charge extra as DLCs for those features. I fully expect it. This is what you get for giving a AAA publisher early access rights, it never ends well.
And it also should be clear by then if the devs know what they are doing and how much fault is theirs and how much the publishers.
Are we really going to play the "Activision-Blizzard" game when we know fully that this was a new team that usurped the IP. It is obviously both the developer and publishers fault here. If the original Squad developers were involved it likely would have just gone the same way KSP1 did. Instead you've got a bunch of scumbags running the show, so no one should be surprised that this is the outcome.
 

anvi

Prophet
Village Idiot
Joined
Oct 12, 2016
Messages
7,553
Location
Kelethin
To be what I consider a game it at least needs a research tree that unlocks better gear that you have to work hard to achieve. Without that it's just another fucking sandbox for retards and everyone who pays for it is decline enabling scum. It also needs to run at a decent frame rate with a variety of hardware. Lazy fucks.
 

Hellraiser

Arcane
Joined
Apr 22, 2007
Messages
11,353
Location
Danzig, Potato-Hitman Commonwealth
Are we really going to play the "Activision-Blizzard" game when we know fully that this was a new team that usurped the IP.

I'm giving them the benefit of doubt solely because of a change of developers mid-development and very likely worthlessness of any code they inherited from Star Theory*, if they inherited anything at all.
It looks to me like they had 2 years to code this from zero and I highly doubt they wouldn't push it back if it was up to just them and not the publisher.

I'll also add that no original Squad developers were around there anymore by that point (the bulk left in 2016), and Squad the company is even more scummy as they fucked over the original devs to the point all of them quit. Felipe Falanghe, the creator of KSP, got just a few years of royalties at some laughable percentage of revenue. His mexican bosses/owners of Squad sold the IP to T2 so they could make movies or some other pet projects from the money.

*unless properly documented it was properly more effort to salvage their code than make it from scratch. And I doubt a studio that got fucked over by T2 nicely gave them all the documentation on a silver platter afterwards.
 
Last edited:

RobotSquirrel

Arcane
Developer
Joined
Aug 9, 2020
Messages
1,961
Location
Adelaide
I'll also add that no original Squad developers were around there anymore by that point (the bulk left in 2016), and Squad the company is even more scummy as they fucked over the original devs to the point all of them quit. Felipe Falanghe, the creator of KSP, got just a few years of royalties at some laughable percentage of revenue. His mexican bosses/owners of Squad sold the IP to T2 so they could make movies or some other pet projects from the money.
I agree, I probably should've said ex-Squad members as that's who I was referring to.
What I can't agree with is the sink cost fallacy. In the case of No Mans Sky I still wouldn't forgive them for what they did, here its no different. You don't do that to your customers its that simple, they could have communicated much better but chose not to, the result is what we're seeing. And Early Access for AAAs is a scam no matter what anyone says. On top of this you have a limited support window for releases, everything they have to fix here takes away from the support that could be added later, I noticed this with Bannerlord hence the 1.0 being absolutely trash because it was half cooked from all the backtracking they did in EA, I actually liked bannerlord a bit when it first went into EA but the developers made some horrendously dumb decisions through development, it can happen, its very likely to happen. Post release support is usually trash for EA games and that's annoying because you go through an ordeal just to arrive at the end and the developer says "ah well see you next time" that sucks. You only get to release once, so make it a good one.
 
Last edited:

Hellraiser

Arcane
Joined
Apr 22, 2007
Messages
11,353
Location
Danzig, Potato-Hitman Commonwealth
I agree, I probably should've said ex-Squad members as that's who I was referring to.
What I can't agree with is the sink cost fallacy. In the case of No Mans Sky I still wouldn't forgive them for what they did, here its no different. You don't do that to your customers its that simple, they could have communicated much better but chose not to, the result is what we're seeing. And Early Access for AAAs is a scam no matter what anyone says.

I wouldn't compare these two cases.

Hello Kitty Games or how the No Man's Scam guys are called were independent though and blatantly dishonest on what will be available at launch or that they are in fact doing an early access release. Key point being that they didn't have suits above them that could fire them or pull the plug if they tried to be more transparent, 100% their fault and nobody else's.

KSP2 had a roadmap saying how much won't be in the release back in October (and the problem isn't the roadmap lying, at least at this moment, the problem is the technical state of the game) and at what price point. People were "concerned" with performance being crap on social media released videos since January. Even before that the gameplay footage shown last year to PC Gamer IIRC was choppy so the prophets of doom popped out already then but in smaller numbers. The ESA event footage and system specs validated the concerns, streamers invited over said outright the game has issues or have shown those issues.

They could have not done the event or just invited press over to keep issues under the rug, they didn't so I'm more willing to assume that if they could afford be more transparent without risking hell and fury from T2 managers they would have just said it out loud (if not ideally delaying the game, again). There was also a post made 5 days before release mentioning issues that are to be expected including that there will be no re-entry heat, that craft files get corrupted, that the performance needs work etc. This was all before they made any sales of the game and deterred at least some people. Hello Games meanwhile had hundreds of thousands of NMS pre-orders 1 month before release (pardon the plebbit link).
 
Last edited:

potatojohn

Arcane
Joined
Jan 2, 2012
Messages
2,646
Pretty sad to see that they kept the spyware bullshit in KSP 2, and there's no way to disable it.

Why do developers need all that telemetry and user tracking?

Do they not know what users want? They want a fun game that looks good and runs well. If you make cute merch and stealth market it on reddit, it will sell well.

There, I just perfectly summarized terabytes of telemetry. Morons.
 

Hellraiser

Arcane
Joined
Apr 22, 2007
Messages
11,353
Location
Danzig, Potato-Hitman Commonwealth
Well the first patch seems to have at least improved the maneuver experience, the long as hell patch notes did not seem to include it but you can finally sticky the AP/PE and closest approach point number tooltips to see them change live as you fiddle with the maneuver node, on top of the burn time indicator working and fixing the trajectories to render properly.

In general though the patch is far from fixing everything wrong, there are still some serialization/corruption issues with crafts and I again saw a random major orbit change on a non-focused vessel and the terrain performance will continue to suck at least until they replace PQS+ with the Concurrent Binary Tree-based terrain system (that's mentioned in a dev blog post).

Still if they keep the pace I think by the end of April with 2 more patches (that I assume go with every sprint since they seem to be doing Agile) at least the critical bugs like the craft corruption and random orbit shifts should be squashed, and the UI/UX gets more polish since they do seem to listen to feedback on that. I had fun with it after the patch without that much trouble, but the trouble did involve me having to manually edit the save file to fix fuel missing and a flameout state bug (something KSP1 also had at some point, possibly the craft was corrupted before the update or the update broke the old save) and a 2-3 extra reloads due to bugs.

They also fixed the damn rover seat EVA bug and made rover wheels actually have proper force. A bit too much proper for realism, they can easily reach over 30-something m/s, but I don't mind driving around at 100 km/h. So anyway I could finally build and launch a functioning one to Duna with my 4-kerbal lander:

bjCxz5G.jpg


The mothership was already waiting in orbit, used it go to the Mun earlier. Of course I am abusing the gas core nucular engine and two hydrogen spherical tanks like everybody else at the moment, while it is not gated behind science or resource progression which it probably will since it is both nuclear and speculative tech and has the best ISP after the ion engine but with a lot more thrust. Docking with the mothership was harder than expected, probably because the RCS was placed only on the lander, so I had to do the lazy method using only the main engine of the mothership and rotation towards each other, rather than flying via RCS with the lander+rover stack.

WPhhhxu.jpg


The lander lands seperately from the rover that is inside the aeroshell, both because I didn't want a tall lander and because the rover is light enough to land using only parachutes unlike the lander itself. I think even if re-entry heat would be added, this aeroshell would be enough for atmospheric entry.

Now this stack had to work around some current limitations to land, you see for the moment you cannot change vessels during atmospheric flight (they're working on it at least), the other is that the physics range is rather small (1,5-2km I think) and it seems that for the time being it deletes vessels in atmospheric flight that are not controlled and outside of physics range (probably connected to the first issue). In the end to land it I had to change the staging so that all parachutes deploy on both while still connected and seperate only during the final stage (1-1,5km to the ground) when I had to start firing the lander's engine.

VG8L3EA.jpg


That worked, despite a craft corruption/deserialization bug shifting that strut segment off. Also they fixed some wheel colision issue so now I could drive off the fairing floor with no issue.

However either the relays on the mothership don't work or I fucked up something with the antenna placement LOS and had to send a kerbal to fetch the rover, as the probe core had no radio connection.

Lander worked well, although the legs are a bit weak (don't know if max stiffness would help, probably should have used the one size larger legs for this mass) its parachutes didn't seem to do much but it had enough TWR and fuel to slow down from over 100 m/s during the final kilometer.

IUfZWuG.jpg


Drove around a bit, decided to do a hop with the lander because I had enough delta-v to spare (I think it is also Ike-worthy when fully fuelled, didn't test though)...

Q1eulvZ.jpg


...except I didn't and had to reload as it turned out the mothership was captured around Duna in a retrograde orbit, so I needed ~400 m/s extra to get back to the mothership from the surface :M

Good thing I didn't pack too much delta-v on it since I figured I don't need to mass optimize and I want it to be a six-legged spider lander.

BTW it seems they nerfed the Kerbal EVA pack, they used to have just above 1 TWR on Duna, now it is just below 1 TWR it seems as you can no longer lift off but kind of almost hover, so long jumping is an option.

DNDR8Cy.jpg


I liked docking this, without the rover attached it was very stable with SAS off so I did a proper autistic docking without abusing the mothership's reaction wheels to point towards the lander and flew around it to the docking port.

3IcjAMq.jpg


Flew by Ike, because why not. It now has volcanoes really sticking out of the surface as visible in the lower-right, although shame it lost its charcoal color. Although my shitty settings and GTX 1660 could also be to blame and it looks better on max*.

Also had to reload because I initially did a 8 km periapsis and it turns out Ike's terrain in KSP2 averages out at about 10 km above the datum :M On my second pass it seems 12km is quite common and those volcanoes probably go way higher.

yV5O7cG.jpg


Here's a smaller one from after the reload.

Should have kept the lander and edited in full fuel to land on it though, in KSP1 it's only 800-ish m/s for a landing+return trip to low Ike orbit.

*speaking of performance, I get 50+ fps in space when no terrain is visible with just the lander and mothership, Duna surface from orbit seems worse for fps than Kerbin as low orbit on the latter is 20+ fps but Duna was 15-ish. Patch seems only to have improved the KSC (I guess?) terrain for me, as I get 20+ fps consistenly now regardless during launch if I look at the empty sky or down to the surface at the KSC, but looking at the far horizon in flight does drop it to 15+.
 
Last edited:

Hellraiser

Arcane
Joined
Apr 22, 2007
Messages
11,353
Location
Danzig, Potato-Hitman Commonwealth
Well they released another patch today, after a long delay. I saw some optimization of compute shaders and what not in the changelog, so I fired it up to see if the KSC or clouds/surface from SPACEEEE lags shit to hell and back still with my cursed SSTO project or the nucular gas core mothership.

tl;dr (see spoiler below for more details)

rsF05tl.jpg


YR12jV3.jpg


Considering I'm running this on a 4 year old budget card (GTX 1660) below minimum requirements on max settings at 1080p this is better than expected, didn't try to see if lower settings would improve it.

It is almost at a state in which it should launch (ignoring missing content, like SCIENCE! etc.), but they didn't yet fix the random orbital shift/decay bug of non-focused craft. The good news is that it is getting mentioned in the dev updates as the top priority bug and apparently a fix is close but wasn't far ahead to get into this patch. The bad(?) news is that it shows how fucked up the code was if they had to squash a billion related issues and still not finalize a fix to it after almost 4 fucking months.

Overall I think that once they fix that, add in the heat simulation and have the space/flight sim part stable and drop the full SCIENCE! mechanics update the public opinion should turn around.

Science is supposed to mix the old experiment collection with the mission system from KSP1, so I think that there's going to be enough gameplay mechanics with objectives/rewards to pick in it to actually be enjoyable outside of sandbox fuckery.

I didn't really try the previous patch much as it just didn't want to load the launch scene with my shitty SSTO spaceplane attempt on it (did fuck around with the small stock plane though, and got at least 10 fps more while flying around KSC on the current patch then on the previous one). Not only did it load but the performance with that SSTO abomination is no longer a slide show in the single digits (usually 1 or 2 fps at launch) but rather a respectable 36 fps on average, despite setting everything to ultra/high at 1080p on a GTX 1660 at that:

TVI4f7p.jpg


On top of that either the performance helped or the fix of aero related bugs (they squashed some drag-cube bug this patch) helped as this fucking thing finally flew in a mostly controllable manner.

Yqfo7tn.jpg


Of course I managed to stall it/lose control by not being very gentle with the controls...

MYCD8fR.jpg


...and then one of the wings fell off which made Aldberry Kerman quite agitated:

G3xRgxT.jpg


Wish I had a gif of his full panic mode animation, screenshot doesn't do justice to the sheer terror he expressed as it went into an autorotating corkscrew into the dark ocean in the middle of the night.

It landed gently thanks to the autorotation but buoyancy was lacking (I need an inflatable floatation device mod) so it started sinking, unfortunately for the pilot.

I couldn't see shit but I assume the game crashed when it hit the ocean floor (or maybe fell through it into Kerbin's singularity in the planet's core).

Anyway quite surprised at the significant performance improvement to trying to take off with this thing from the runway. It was actually rather enjoyable experience to try to fly this as opposed to the agonizing experience due to the long loading time and slideshow after the first patch those ~3 months ago.
 
Last edited:

Hellraiser

Arcane
Joined
Apr 22, 2007
Messages
11,353
Location
Danzig, Potato-Hitman Commonwealth
The first content update/milestone with science/research, part heat and re-entry and the new campaign mode called "Exploration" was released about 2 weeks ago, and earlier in September I think there was a major performance improvement patch.

tl;dr summary: the performance after both seems a lot better and people below the infamously high minimum specs claim it has again and even more improved vs what I have seen in the June update*. Still buggy but most of the time a reload fixes the bugs (for example a craft not loading correctly), more rarely a game restart and reload (orbital paths disappearing). Map view/tracking UI/UX is a mess that fights the player half of the time and needs a rework, also it is a bitch to do interplanetary transfers without mods/external tools. I like what they did with science and contracts (see sperging in the too long spoilered section below). Re-entry effect has an at times janky animation, and the heat system needs some hotfixes (*badum tish*). Still it took them almost 10 months to get to this state since EA release so the pace was rather slow and clearly the game needed at least those extra 10 months in the oven.
I really underestimated how long it would take, I gave them 6 months at most until science would be out, it was almost double that :M.

*I can't judge myself as my GPU died during the summer and I had to push ahead with the GPU upgrade I scheduled for next year, but on high and at 1080p it does seem to chug only if I do complex craft and mostly from aerodynamics/the simulation rather than from graphical fidelity. Usually it's 60 fps for me, even when looking at planets from space.

Steam says I clocked 67,6 hours in the last 2 weeks (out of 107 hours total), but this is inflated due to quite a few instances of leaving the game paused while I do other stuff around the house. Still I definitely spent a lot more time playing it now after this update than the total time since launch. Anyway here's what I think about the new campaign mode after going pretty much everywhere in the Kerbolar system, doing most of the contracts, including all the primary/"main quest" contracts, and unlocking all but some expensive and not that needed at the moment tier 4 tech.

On a very broad level mechanically it is similar to KSP1, do experiment in biome on some moon/planet or do a contract to get science points, unlock a tech tree node with those science points. But as I played I realized how sometimes seemingly minor changes here and there resulted in a different approach to playing Exploration in KSP2 vs either Career or Science modes of KSP1. In general it grew on me and I can clearly say it is far better designed than KSP1's was. As for the changes...

The contracts are not randomly generated and serve a different function, they are meant to be the primary source of science (experiments are something you do along the way or if you need 200 more science for some tech to do a contract), throw challenges at the player and nudge them outwards to explore other worlds rather than grind the same few "easy" places for funds/science. They are no longer timed and they can't be declined/abandoned, rather they are like quests in a journal that you can track (KSP2 is now a RPG) and do whenever you wish (or ignore), unlocking based on some progression conditions.

KSP1's contract generator was annoyingly stubborn in not really giving you contracts to go to X and do Y unless you flew by X already. KSP2 avoids it and for instance actually made me land on Eeloo (for the first time ever, never did it in KSP1) as the cost vs benefit was very much in favor of going there and doing the contract (and get some science along the way from experiments) rather than just fiddling mostly around Duna or Laythe like I always did in KSP1.

Besides that there's a main storyline and "main quest" chain of contracts, which you can probably guess what it is about, it's basically a first contact/follow macguffins deeper into space plot related to "anomalies" like the mun arch, with obvious inspirations from 2001 a Space Odyssey and Contact in the plot, plus some secondary missions. As it is however the game doesn't really force you follow the main quest, but it is the easiest way to to gain the most research and I think it does unlock some new secondary contracts. Also after the first 2 or 3 IIRC tutorial ones they turn into a series of targeted landings on celestial bodies, so they're fairly challenging while providing much freedom in how to tackle them, especially if you lack rover wheels but decide you want to take them on anyway. I'm actually wondering if they won't be too hard on newbs, precision landings are almost as hard as docking to get right (and with less tools to aid you, like needing to guess how much the surface will rotate while you are in orbit).

For me at least doing them seemed the most fun and I decided to see how fast I can go through it without research grinding. At first the motivation was just to see where they go with the story, but then I actually had a ton of fun designing craft and launching missions along the way, and going to places I often ignored in KSP1's career (well it helps KSP2 has the nuclear gas-core engine, but not that much).

The main chain wasn't too difficult while trying to rush through it and do a path of least resistance/laziness with probes, but I have a decade of playing KSP behind me so go figure. Overall it takes you through the obvious places that is the Mun, Minmus, Duna and then at the (current, as it will obviously continue when colonies or interstellar travel gets added) end does something somewhat unexpected and makes you land on Tylo. Personally expected Laythe (obvious pick) or Dres (to subvert the "nobody goes there/doesn't exist" meme), but those are "optional", at least until resources are in and something will be an exclusive source of say Uranium for nuclear fission drives. With Laythe I'm even more suprised as I didn't see a secondary contract to even go there, but maybe that is because it needs another contract completed which I ignored, like the one where you need to land 3 Kerbals on Ike.

As the contracts are handmade they are sometimes rather unique challenges, such as:

nKTQahL.jpg


There's some speculation ongoing that the low science reward is deliberate and there will be some kind of colony-related reward on that one in the future, since it is a 200 ton at touchdown lander it asks for. There's a similar contract with Duna as the target available later and surprisingly independently from this one.

Or this rather hardcore one:

03l22pJ.jpg


Only one I wanted but didn't do so far, inflatable heatshield is bugged ATM and I can't make a lander that doesn't burn up without one. Maybe if they hotfix it.

The science reward balance is such that main story contracts>secondary contracts>experiments done with parts.

Experiments are last for two reasons, one is that obviously the payout from contracts is always better, the other is that the number of science parts got gutted (all the sensors are pretty much merged into the "crew survey" parts), each tech tier/page has usually just one new experiment type (there are 4 tiers currently, but the last one is only half-filled as it will unlock colonies or interstellar ships later), and as it is the game has five experiment types total with all but one having two parts that can do it (usually low tech heavy/large choice, and hi-tech miniaturized/more practical version). So especially early on the balance greatly favors contracts, and due to how the new parts look/work and how fast you can unlock them, you can't make a plane or hopper stuffed with every science part and just grind all biomes on Kerbin, Mun and Minmus. However I do miss having a real time pressure, gravity or temperature read out, also how do kerbals know the pressure on Duna if you are the director of the first and only space program on the planet?

Experiments are sometimes more involved than "click, get science". For example:

EqQwQoy.jpg


This large part with the antennas and beams/truss deployed is the larger, low-tech and crewed version (the probe version comes much later, in tier 4) of an orbital survey module which requires you to fly over a biome for 5 minutes (pauses after you "exit" the biome, can be resumed manually, but doesn't auto-resume yet, making it very tedious ATM for "crater" biomes for instance). I think ot will also scan for resources in the future. There's another one for sampling the atmosphere with a similar time requirement forcing you to either use a lot of parachutes or wings or hover with an engine to stay in flight in atmo for a few minutes.

In total you now have: crew surveys (can be done with the first science part), kerbal/robotic sample collection (robotic arm allowing probes to do it comes later after rover tech), orbital survey, radiation experiment (very bulky, I just leave it behind after done), atmospheric sampling, aquatic sampling (didn't unlock this yet, it comes in tier 4).

Another change to science is how now almost all experiment produce data and sample components. Data you can always transmit and get the full reward, no penalty anymore. The samples only give their reward if you return them to kerbin. So no more "fuck I clicked transmit all, reload" issues. Also you no longer need to manually transfer experiments/samples, experiments are now like a contagious plague, if a vessel or kerbal with one makes contact with another vessel (docks/board) everything gets cloned over to its inventory and once at least one "infected" craft is recovered on Kerbin you get the reward, and the experiment is cleaned up from all vessels.

Very convenient and one consequence along with the changes to science parts is that you can have a fully unmanned exploration program apart from the contracts that say otherwise, which is pretty much what I did in KSP2 now. Only landed with kerbals on Dres (optional contract made me do it), and on the Mun (very late into the game, ignored the contract for a long time, but wanted the later one from the same chain). I even did a Moho robotic sample return mission but it ran out of delta-v and is stuck in a Kerbolar orbit (I fucking hate Moho, goddamn eccentricity and inclination, delta-v maps/trip planner always lie).

Another minor addition, that personally I didn't bother with despite thinking I will, but I anyway like that it was added, is that there are "discoverables" that is special points of interest that count as a seperate biome granting unique experiments, for instance there are 2 or 3 special rock formations on Gilly.

I like what they did with the campaign so far from a design point of view and the direction they took, although certainly it needs a lot of QoL improvements in addition to fixing some still annoying bugs (I think at least the orbit shift bug is finally gone). Assuming T2 doesn't pull the plug on development within the next year, I look forward to seeing what they do with colonies, orbital construction and the related new content meant for the next milestone, tying it to the exploration mode gameplay just introduced.
 
Last edited:

As an Amazon Associate, rpgcodex.net earns from qualifying purchases.
Back
Top Bottom