lazarus1024

Yeah a mid grade video card is pretty much going to allow you to max out the graphics and have no graphically based limitations on your frame rates. Your limitation will be on the CPU side. There is probably no actual maximum where you'd see no return, but the question would be, how big? I can easily push 300+ part ships and stations with no evidence of slow down (IE never see under 30FPS) on my i5-3570 running at 4/4.2Ghz (4Ghz quad and 4.2ghz single core turbos) and 5570 (graphics aren't turned all the way up as my low end card can't handle them, but I do run at 1080p and medium settings, no AA). I can't recall a time I've ever had a ship over 400 parts on my desktop as I play almost exclusively on my laptop (i5-3317u). The laptop is high enough spec to handle most of what I want to do in the game and lounging on the couch or the easy chair in my living room with my wife is more important to me than sitting in my basement office. That'll probably change some in a year or two once we've finished renovating our basement (just finishing touches now) and we also have the spare change to get a nice new TV (until then, the 42" in the upstairs family/playroom and the 32" is in the basement). Then the basement will be the main entertaining room and I'll probably use my desktop a heck of a lot more. So all that said, my laptop can regularly run 200 part ships at over 30fps through most of game play. I do get a bit of launch lag and a bit of orbital horizon view and looking down on oceans lag on larger ships, but it probably still hits over 15fps most of the time and regular play is over 30fps (granted, 768p). The last few releases I haven't done much other than stuff right around Kerbin, Mun and Minmus with a couple of quick probe forays to Eve and Duna as I just haven't had much spare time the last few months (and a back injury over the summer). Earlier on, I did have a bit of a frustrating time with some really big ships in the 300+ part range for Moho and Jool missions, but I am a lot more efficient in design now, so I tend not to need them that big. I won't mind (hopefully) a couple of years from now getting a new laptop and fingers crossed, CPU performance will have gone up a moderate amount (at least 20-30%, preferable 25-35%) over what I have now.

If the planet had an atmosphere it actually might be thaattttt hard to do. It would be real tricky, but you could use the atmosphere for capture, or a straight descent and then for return to Kerbin you could do the same. It would be VERY tricky, but you could do significantly lower dV designs by utilizing aerocapture. Honestly I think that would be really cool to do.

Oh. When was this added, because I never noticed it before? I kind of wish it was closer to just the inverse-square law. That way for things Duna in, you can easily use solar panels. For stuff further out, you either have to go big with solar panels or switch to RTGs (supposing you want to do much). That said, I'd guess for Jool and especially Eeloo you would want to either go dense with solar panels or just use RTGs. This'll get more interesting as there are more things that require electrical power (and before that happens, please allow batteries to store more juice! They last an unrealistically short period of time for basically any load).

Not necessarily. Last I heard, some astronomers and astrophysists are starting to think we may have had a super jupiter/hot jupiter in our solar system's early history. Lithium is not a commonly occuring metal in stars, however it is (relatively speaking) common in most gas giants. Our sun has an overabundance of lithium in its photosphere compared to most stars. So early working theory is we had a hot gas giant that spiraled in and was absorbed by the sun probably billions of years ago in our solar system's history. I don't know if there is much backing other than a hunch in terms of orbital models of the other planets, but it is a very interesting idea.

Oh, so the later is going to be located at the ass end of the solar system Sorry, I couldn't resist a good Uranus joke. They could change their mind, but SQUAD did say when adding Eeloo that it would become a moon of the second gas giant at a later point, possibly with active geysers. They are also planning on adding a planet closer in than Moho with lava and heat damage. The second gas giant will have rings like Saturn. That is all I am aware of that SQUAD has mentioned in relation to future planets. After the above, I hope they make a small gas giant like Neptune with a couple of smaller moons as well as add in a few small dwarf planets equivelent to Ceres and Vesta right around where Dres is now (Dres is great, but too large to represent Ceres, and one really egg shaped dwarf planet would be cool). A couple of small moonlets for Duna would be neat too with even lighter gravities and SOIs than Gilly (hello Demos and Phobos). Lastly, but not leastly, one or more really far out Kuiper belt type dwarf planets (maybe one with a tiny moon) would be neat. I think that would be all she wrote though unless or until adding an actual asteroid belt (if ever feasible) and/or other star systems/comets

Suprisingly I've never been to Val or Tylo. Only two places I haven't visited (okay, I just lied, I haven't been to Eyloo either since it was added). Duna is lots of fun. The Mun can be too. For places like Minmus, Bop and Gilly I don't bother roving. I make little RCS jump vehicles that I can dock to the top of my lander for refueling. Pretty much just a panel, reaction wheel, battery pack, a few small solar panels, RCS jets, RCS tanks a light and a seat. Lots and lots of fun and you can cover a heck of a lot of the planet.

In addition it would be nice to select multiple tanks as sending or receiving tanks when shuffling fuel around on your ships. So for example I could pump fuel from one tank in to two at the same time/rate. That or allow you to type in the value of how much fuel/oxidizer you want to shuffle from one tank to another.

It would be cool if SQUAD made the anomolies their own distinct biomes for doing science at some point.

Its part of the reason I didn't mention a limitation on docking/per mission. Just a per launch limitation. If you have a limit per mission, its likely you'd never be able to build good sized space stations or bases at any point. Or if you could, ridiculous 1,000 part launchers would very much be possible. I am not opposed to ridiculous launchers, I just think that their place is more in sandbox mode than it is in career. Also juggling a mission based cost limit might be darned hard to do. You'd never be able to do anything permanent and so on. Though I suspose, I could see a "method" whereby you are limited by the total cost of your space program. Launching and losing missions costs nothing. However, for every in existance ship (outright debris doesn't count) it counts against your total space program budget. So if you have 50,000cr worth of ships, stations and probes out there and your budgetary limit is 50,000cr, if you want to launch another ship, you'll have to recover/end something to get the budgetary money back to launch something else. That or earn more budget through accomplishing a goal. That said, that'll neve really limit "ridiculous" designs, but it could also be a method that could work.

Instead of having a method to earn money and bank it, as well as possible penalties associated with ship that fail, how about having a set budget per launch? So as you achieve various goals, you get your budget per launch increased by your Kerbal overlords. Get to orbit, have your budget increased per launch. Recover a capsule for the first time and get your budget increased. Transmit science from a probe and get your budget increased. Orbit the mun, budget bump. Land on the mun and EVA with a kerbal, budget increase. Return said kerbal from the mun to Kerbin and get a budget increase. Etc, etc. That way there are no penalties for failed launches (which I think could cause a lot of problems, even with experienced players) and may, to a limited degree, prevent some really ridiculous launcher designs (in career). It also removes time from being any kind of factor in earning money.

Another option instead of simply earning an unlimited amount of credits and being able to spend it to your hearts desire, is having a credit limit PER LAUNCH. You can unlock higher funding by achieving certain goals in the game. For example, orbiting Kerbin once. Succesfully launching and recovering a flight. Landing a probe on the Mun. EVA'ing a Kerbal on the Mun. Returning a soil sample from the Mun. Orbiting the Mun, etc. I don't think it would be that difficult to add in achievements that would benefit your budget. Then, instead of simply accumulating money through a time based mechanic, you get limited funding per launch. That also prevents penalties from getting bad just testing out rocket designs (and possibly having to dabble in sandbox mode to get a good design and then create it in career). So, maybe you start with Tech tree 1, zero science and a budget of 2,000cr per launch. You successfully launch a rocket and earn some science AND since you managed to recover the pod, the Kerbals celebrate and you get awarded a higher budget of 2,500cr per launch. You get in to orbit on this one and now you've had your budget bumped to 3,000cr per launch. Or some such thing as that. It prevents you from having any mechanic where you can simply time warp your way to an unlimited budget and it also prevents failure from being a really bad penalty (though you could also have your budget limited if you have too many missions that get blown up or something). It would be my prefered method of economy rather than earning set amounts of money and would likely provide more flexibility in game play.

Well, at least Baytrail seems to bring roughly double the single thread performance and about 3-6 times the GPU performance of the old Atom based netbooks (with GMA graphics, not with Nvidia ION graphics...probably only roughly 25-75% better graphics performance than what ION was/is). Of course that doesn't necessarily turn completely unplayable at any settings in to bareable with everything turned way down.

Well that gives me a little hope that it could be playable, if sometimes rather laggy. At least from what I know of it, baytrail should be able to turn in at least twice the CPU performance of something like a 3Ghz P4 (and also still better single thread, not just multithreaded performance) and I can't imagine that the GPU isn't more powerful that X200.

Interesting, what kind of performance did you see?

I have also noticed much faster load times in .22. One of the nicest things behind research/science for me. Before (i5-3317u and a 500GB 5400rpm drive + 32GB SSD cache drive) I'd experience maybe 100-140s load time for the game itself and scene transitions from VAB to launch pad and worse anything to the space center view would typically take around 15-25s and SOI transitions would also take maybe 6-10s. Now I'd estimate it takes about 60-80s to boot up the game, scene transitions are only taking maybe 8-10s and SOI transitions are also only taking maybe 3-5s. It makes it much more pleasurable to play the game. That said, I am really hoping I can pry lose the spare funds this fall to get a 120GB SSD to drop in my ultrabook and pry out the 500GB HDD. I don't need all that storage (I'll cop to having about 150GB on the laptop drive though, but I could easily clean about 100GB of that off). I'd leave the 32GB cache drive in there and set things up so it would be two partitions of 120GB and 32GB (pre-formating of course). I may even set it up so the 32GB would be solely OS. Its fairly slow writes on the cache drive from what I have seen of the specs, only about 85MB/sec, but reads are in the 500MB/sec range. Then do all apps and everything else installed on the 120GB SSD.

The next couple of things I'd love to see are (in order of importance to me) Adding money/economy to career mode. Research and science were huge (it allowed career mode), but I think to round out the career mode experience, having to budget for your builds and a method to earn/get/accumulate money I think is also important. Adding that in I think will flesh out career mode significantly and likely would also "put a stop" to things that feel kind of cheaty on how to amass tons of science in short order. Moar parts. I do like most of what we have, but I do feel like some important parts are missing, especially once money/economy is added in and balancing/tweaking science. We need some 2.5m science parts and we need the option of some .725m probe science parts. I am thinking things like a soil scoop/analyzer for probes would be neat/cool/should be there and things like orbital research bays as 2.5m parts. Some more advanced parts, such as 2.5m NERVAs and nuclear reactors would be cool. Especially if other things like more powerful Ion drives or VASMIR is added in to go with nuke reactors (scaled roughly to in game ion drive standards...so something like a 1.5m ion drive/NERVA might use 10x the power of a probe based one, but also produces something like 10-20 thrust and a 2.5m one that uses 50x the power of the probe ion drive, but produces something like 40 thrust). To go with later suggestions, radiators too. Heat. Engines in space and some things like nuclear power plants and RTGs should produce heat that accumulates (even if very slowly, but dissipates super slowly unless you have working radiators). Radiators will radiate it away. Life support. Especially with career mode, I think it is time in the next couple of updates to consider implementing life support. 1.5m and 2.5m life support modules and small (monopropellant sized) radial life support modules would be nice. They use a small amount of power (very small) and have a finite amount of "life support gas" in them. Make it not uber challenging, so something like a small radial tank might be enough to last 4-7 days for a single kerbal and uses 1 unit of energy per minute or something. Probe sized provides double that, 1.5m gives 2 weeks for a single kerbal and a 2.5m gives a month for a single kerbal. Then also provide things like 2.5m hydroponics/aeroponics modules (hydroponics costs less, but weighs more and is unlocked before aeroponics in tech tree). Those hydroponics/aeroponics can provide unlimited life support gases for 3 kerbals, or recharge life support gases at the appropriate rate if there are fewer than 3. Great reason for space stations now too. Stick a hydroponics module or two on it, send it out to Duna and you can tank up missions being sent out there. You can also have "atmosphere reprocessor" units in radial size maybe, that use a lot of power, but also recharge life support gases on planets that have an atmosphere (can be used as an atmosphere scoop too, but the lower the air pressure, the slower the recharge). Deadly re-entry heat and heat shields/coatings. I honestly don't want this to be super difficult. However, if you are going fast enough to cause a heat trail, that should be fast enough to damage parts that aren't protected by a heat shield or tiles/coatings. Allow any part to be coated/tiled, but have the protection be a lot weaker than a heat shield (IE could withstand moderate aerobraking or gentle re-entry, but it isn't going to withstand a plunge straight in from orbital speeds. Heat shields can generally protect anything at nearly any speed...so long as not really ridiculous). Biomes for all planets and have them remapped (a la Kerbin and the Mun in .21). Possibly adjusting which planets there are and their setups like SQUAD things is closer to a final layout of the planetary system would be nice too. That's all I can think of for stuff I'd really like to see in the next few updates (say within the next 3-4 updates). There are many other things, but I feel like they'd add more after all of this stuff was done.

Roughly 1/5th the performance of HD4000. Or at least that is my expectation based on clock rate and number of EUs. However, looking at the few benchmarks comparing, it seems to do a bit better. Then again, the HD2500 which is 6 EUs and roughly similar clock rates to HD4000 with its 16 EUs seems to manage better than the ~38% performance it should. I suspect that is an element of CPU performance making its way in to frame rate numbers in the benchmarks though. Still and all, I'd expect if it is GRAPHICALLY limited, you'd see around 20-25% of the performance at the same graphical settings as an HD4000 based system. If it is CPU limited, then more like 30-40% of the performance of your average current ultrabook (single thread). I figure it would probably be unplayable, but I am still curious. I could live with frame rates in the 20s just to play around with it a little. Its not ideal, but it could be bearable as a simple experiment and happy fun time on rare occasions. You'd never be able to get an interplanetary mission together with a straight face (probably not even a Mun mission).

In seriousness though, what kind of super low end machine has anyone run this on before, and what kind of performance were they able to eek out? I remember reading on the performance thread a long time back (months ago?) at least one person who did try to run it on an old netbook. Got it to boot and run, but it ran at like 8-10FPS at best or something like that. That would make a newer Baytrail chip almost runable for small rockets and very low settings (being roughly 2-3x as fast in single threaded performance as the older atom chips and much better graphics).

I think you are going to want to back up on that Windows spec. The more and more Windows releases there are, the worse they seem to preform. Also the memory spec is generally paltry. The rest I'd be willing to accept at face value.

So, I am just kind of curious about the floor for the game. I don't NEED to run KSP on the following, but I am curious (and will be trying). I am looking at getting an Asus T100 Christmas-ish time supposing something nicer doesn't come out before then. I am vaguely curious if KSP would run on it at anything like playable frame rates, even with stuff cut way down. IE 1024x600, all graphical options turned down smaller rockets (say up through 50 odd parts). I know Baytrail has only been out a couple of days, but I am curious if anyone has had a chance to land a baytrail based tablet and run KSP on it, or when they do if they can give it a shot for the heck of it. Supposedly z3740 Baytail chip in the T100 performs roughly at the level of a Celeron 847 mobile chip (dual core, 1.1Ghz). About 125% of the multithreaded performance and around 80% of the single threader performance. Looks like the GPU should be roughly around 50-60% of the performance of the Celeron 847 as well based on clock speeds and EU count. So if anyone has run KSP on a Celeron 847 based laptop, I'd think that the performance would be roughly similar (with CPU probably being more of the bottleneck than GPU, especially with things turned way down). Alternately the z3740 seems to have roughly the same single thread performance as the current AMD Jaguar based chips out there. PS this is really a curiosity question and I don't really intend to play KSP on the T100, I am just curious if it could. I'll stick to mostly playing on my HP Envy 4t i5-3317u based ultrabook or my i5-3570 based desktop.

My wish is for something like protractor to be added in stock. Click on your intended target and it'll show you when/where to burn to achieve intercept. That said, I mostly work fine with regular manuever nodes.

Valid, but I probably would have lead with the actual question and then thrown in the reasoning behind the question. I paid for it back in .16, so I wasn't expecting resources back then. At least IMHO, the game is great as is. In each release SQUAD generally makes it a lot better in some way. Other than personal reasons (health, moving, etc) I've enjoyed every single update and each one adds something to the game that I have a lot fun with. I suspect that I'll be having a heck of a lot of fun with .22 and still not be done with the tech tree by the time .23 hits and probably makes me start out from scratch again (oh darn).

Allowing you to setup up space centers/launching points could be one of the features of bases/space stations. I agree that resources would be a major boon to giving them a purpose, but there are other things they can be useful for. If/when life support gets added to the game, even if resource extraction doesn't exist, having a space station/base can be a big thing to resupply fuel and air from. Or if you land the right modules on Duna, now you can start missions launching from there. Or maybe it just allows you to construct thing if you take the parts there. Not sure how they want to work it, but I'd swear I remember HarvestR saying they'd add in the ability to CONSTRUCT space centers. That could be a faulty memory though.

I think your stages are still messed up. HarvestR mentioned being close to a working economy with the .22 release, but he thought it wouldn't make the cut. So it sounds like it is likely to be in the next release as the main focus. Which puts it much earlier. Also a lot of your other things, like filling in the Biomes and stuff, all though I am sure a resonable amount of work, doesn't sound like anything that would even rise to a full release, let alone possibly several releases. For the later stage stuff, SQUAD has mentioned that reliability will NOT be occuring in the game. No worries that a part or engine might randomly fail. Wouldn't that suck if there could be random part failures and you just lost a mission after 3 hours of real world time? I'd be pretty annoyed. Bad enough if I screwed something up, I don't need random stuff breaking in game as a feature. Going back though, I do think the next 2-3 releases are going to be very career focused. After that it is just too up in the air. As HarvestR explained around .19, they are in asymptote mode. As features get expanded, adding meaninful game play value for that feature requires increasing amounts of work. I think until Career is to a point where it takes a lot more work to expand it just a small amount, I think they may refocus on sandbox...or focus on things that expand both of them roughly equally (like resources). I do think career mode is going to take a few more release to get it to where they are happy though. Research/science is big/huge and I love it. However, I think there is going to be a lot more than simply giving you money and a budget to finishing off career mode and the possible cross over a resources in to career mode (I hope there will be a mission system at some point among other things).

I basically expect KSP to still be in Alpha at this time next year as well. At their current release cadence, even if a few of the next releases are a little easier (tech tree and research was a damned complicated/big release). I'd imagine they'll still see a 6-8wk release cadence with some outliers one way or another. I'd bet my left nut (please don't make me give it up HarvestR!) that the next release is going to be adding in the actual economy/money to career mode, along with probably some part balancing and tech tree/research tweaks/balancing. After that, damnedifiknow. Probably still a focus on career. I'd expect either working re-entry heat and/or life support probably aren't too far away as both sandbox and career enhancements. I'd also expect to see a bit more focus on making space stations and dirt side bases (parts and reasons to exist). I think resource mining and utilization are several releases off. I don't think they are "right before Beta release", but I do think they are going to end up being several releases away. I think economy is the next big thing as it'll add a huge amount to career mode. Then I think they'll focus on more career mode polishing with some things that carry over to sandbox. Then I think it might be resources as that potentially adds a huge amount to sandbox and some to career if implemented there as well (possibly suplementing economy). I do think in the next release or the next 1-3 releases we'll see those biomes get add to all the planets and possibly having them cosmetically updated as well (ala the Mun and Kerbin). However, I don't think that'll be the focus of any of the releases, I think it'll just be add-on features. I could be wrong though, .23 could be all about balancing the tech tree, tweaking research/science and "filling in" the planets. Though if they are going to do that, adding in what they are think are the final planets and configurations would be nice in a way. Or we could have a version .22.1 that maybe fixes any outstanding bugs in .22 and maybe fills in some of the planets for all I know. Will we find out today? Who knows, but I am very eagerly looking forward to the Kerbal Report.