| Forum Index > General Games |
|
Cyro
United Kingdom20326 Posts
Yes. And i said, if you don't play with mods, that works. Clearly. I do. My install is quite highly modded, i just don't have much stuff that outright requires increasing the part count of rockets by a huge amount and i'm not really aware of mods that require that. | ||
|
m4ini
4215 Posts
But lets agree to disagree, if you can play fine - fine. I'll stick with my opinion (and squads, and the rest of the community) that ambitious stuff in the current engine is not possible while playing "normally". You could also ask Scott Manley on that, which openly said what i just said, too. | ||
|
Cyro
United Kingdom20326 Posts
Everyone agrees more parts = less performance = bad, but your experiences are not everyone's. Agreeing to disagree on something when you can easily check the facts and prove one side isn't the right thing to do - i don't mean that in a mean way, am happy to solve any problem that you present. I'm also on a fun little duna return mission at the moment to see if i can do it with a haphazardly thrown together rocket and transfer :D | ||
|
stenole
Norway869 Posts
On March 07 2016 02:41 Cyro wrote: There are some limitations that have important effects and performance is not where i'd like it to be (i sometimes spend more time benchmarking games with different configs and situations than i do playing them!) but it's not anywhere near the extent of a problem that you say it is - you can do single launch return to duna with a small rocket and run at 200fps, like i've shown. You can add a bunch of science parts, life support and drop a colony while you're there without killing performance. Everyone agrees more parts = less performance = bad, but your experiences are not everyone's. Agreeing to disagree on something when you can easily check the facts and prove one side isn't the right thing to do - i don't mean that in a mean way, am happy to solve any problem that you present. I'm also on a fun little duna return mission at the moment to see if i can do it with a haphazardly thrown together rocket and transfer :D I have to side with m4ini on the performance issues of KSP. It comes down to how the parts interact with each other. Each part you add will need more calculation than the previous one. The CPU cost of a 20 part rocket is in a whole other universe than a 400 part rocket. And this is something you can't get around by changing how pretty the graphics are or how beefy your computer is. Eventually you will reach a soft part limit where everything becomes slow and unplayable. The number of parts between a smooth, perfect experience and a game that is essentially paused is surprisingly few. | ||
|
Cyro
United Kingdom20326 Posts
The argument that i was using is just that you don't have to go to those part counts to do almost anything in the game. For that one example, a single launch duna return, you can fly that in a fun and efficient way with 10-20 parts. For some of the more crazy missions like a single launch mothership to the jool system with landers for tylo and laythe you can break 100 parts, but otherwise you don't have to or have much incentive to. It's very annoying and limits freedom of expression but it doesn't break the game. --- For the launch window thing - I think personally with life support, dealing with transfer windows is part of the fun. As an example for not needing transfer windows even for single launch low part count return ships, i took one of the most awkward windows possible for Duna and did a bi-elliptical transfer. Total cost from LKO (75x75km) to low duna orbit was about 1920m/s (1250 and 670 burns), so 880m/s more expensive than a perfectly optimal transfer. The return was at a very awkward time as well, but that's much cheaper. It was ~660m/s (hohmann) and wait a few years or ~1km/s and don't. | ||
|
Cyro
United Kingdom20326 Posts
![[image loading]](http://i.imgur.com/0GVoFwn.png) LKO with 5km/s delta-v left  I remember back when 1.0+ SSTO was considered to be hard - people just didn't understand the air breathing engines. Stats & thrust curves (with altitude and speed) helped a lot with that. | ||
|
stenole
Norway869 Posts
On March 13 2016 17:16 Cyro wrote: LKO with 5km/s delta-v left I remember back when 1.0+ SSTO was considered to be hard - people just didn't understand the air breathing engines. Stats & thrust curves (with altitude and speed) helped a lot with that. Nice. If we translate your remaining fuel into cargo it can haul about 60 tons to orbit (a little less to account for the return flight and margin of error). That assumes it was fully fueled at launch. It's probably a good work horse for a space station project. My only recent KSP achievement is almost getting Jeb up from Eve on a vessel launched from Kerbin. In other words, Eve remains undefeated... | ||
|
Cyro
United Kingdom20326 Posts
Space station project - i enjoy using far bigger launchers. I have a 100% recoverable rocket that can put 100t in LKO but that's a medium sized rocket for me  | ||
|
nimbim
Germany985 Posts
![[image loading]](http://i.imgur.com/JpdaFxZ.png) Off to Moho with 3m/s² and the most experienced team. Training crew in career mode takes forever. Mun, Dres and Duna landing aren't enough to fully train a Kerbal. | ||
|
Cyro
United Kingdom20326 Posts
Single craft still seem to be on one CPU thread as expected because squad are not wizards. There's also a huge loading pause when staging or loading a lot of parts which may be related to CPU and/or storage performance (guy with HDD was having 5 second pauses, but otherwise running at 20+ fps). This is understandable from an optimization point of view as they may have to consider each craft seperately and do a costly "rebuild" of everything that they're considering for it in order to drastically improve performance of those craft - having a 2-5 second pause but then having 50% more FPS when you're playing with hundreds of parts seems worth it. There are a lot of minor new features (exception to that is the wheels IMO, they seem a LOT better!) that are hard to go into without forgetting anything but working 64-bit and improved performance are on the top of the list ^_^ | ||
|
stenole
Norway869 Posts
| ||
|
Cyro
United Kingdom20326 Posts
The performance improvement seems to be bigger than the developers suggested early into the project. I guess they didn't want to promise too much They didn't know and didn't have much control. It's Unity's world, they're just living in it :D People have previously suggested that KSP could get 5-10x improvements like Besiege, which i was always very skeptical of. The devs were very quiet even late in the development process which was weird. It got to the point where they had said basically nothing, but we got footage of version 1.1 which showed an Xbox 1 flying a stock craft - the Kerbal X with 72 parts - at about 0.5 - 0.7x realtime speed during launch which did not excite many people. XB1/PS4 cpu is very bad - worse than a 6600k by like 6x if you're running 4 or fewer threads - but this news did not seem to imply strong performance improvements. It turns out they may help the most on craft that the XB1/PS4 can't run well anyway, and systems like mine may have performance functionally doubled on 250-part-craft or perhaps more with multiple craft. I'l be taking a much closer look once i can actually play the game and not just sift through hours of footage from the 24-hour-marathon VOD's on ksptv! Overall console experimence so far doesn't look good considering that performance and the lack of console UI & mods. They're placing their parts with a thumbstick-controlled-mouse-cursor.  | ||
|
Salazarz
Korea (South)2591 Posts
| ||
|
Cyro
United Kingdom20326 Posts
(but no working mods.. lol) | ||
|
stenole
Norway869 Posts
On March 30 2016 02:08 Cyro wrote: You'll have it very soon (but no working mods.. lol)It wouldn't surprise me if several of the most notable mods will be ready by release. Modders have had special access to experimentals and they have previously been very quick to update their mods with previous releases. The visual improvement mods might take a while to get working though. | ||
|
Cyro
United Kingdom20326 Posts
https://www.reddit.com/r/KerbalSpaceProgram/comments/4cgm5q/kerbal_engineer_redux_indevelopment_build_for_11/ Pre-release is live as of about 1:30am UTC (thanks squad.. ) & a small list of mods have working versions including kerbal engineer and spaceY AFAIK.Right click the game in steam, go to properties - beta - prerelease. You should probably copy/paste your KSP folder somewhere as a backup before doing that. First impressions: The game looks a little weird. Without testing, i'd guess that frametimes are substantially more consistent so that any given FPS on this new version of the game will look better than the same FPS on older version, which had quite inconsistent frametimes. It's a bit weird because it feels better at ~40fps, but i'm also getting the "woah this is smooth" feeling at framerates way into the triple digits. Also a little weird. I just dropped 4x110 part craft at the side of the runway and launched a 200 part craft. I get ~65-80fps when not looking at them, but 40fps while i'm looking at them. | ||
|
Cyro
United Kingdom20326 Posts
There are some interesting things happening in 1.1. From some testing, it looks like some parts of physics are ignored in some situations. I staged away 300 parts and my FPS immediately went to ~150-180 while they were within a few hundred meters of me and clearly in physics simulation range, they were ragdolling around due to drag etc (RIP kerbal joint reinforcement) but they seemed to lose the ability to collide with other parts that were attached to them. Parts of the stage were clipping or flying through other parts quite obviously https://www.reddit.com/r/KerbalSpaceProgram/comments/4ckysm/since_ive_seen_multiple_posts_about_mods_here_are/ | ||
|
shin ken
Germany612 Posts
In typical crasy KSP fashion, this monstrosity is the first stable man-portable tier 1 design I came up with :D ![[image loading]](http://i.imgur.com/hVX3QeX.jpg) using yaw is a bit tricky, but at least it keeps its course without diving/climbing! | ||
|
stenole
Norway869 Posts
On April 05 2016 21:52 shin ken wrote: First time experimenting with aircraft! I wish I had done this earlier, it's super fun. In typical crasy KSP fashion, this monstrosity is the first stable man-portable tier 1 design I came up with :D using yaw is a bit tricky, but at least it keeps its course without diving/climbing! Usually yaw control is something you do not need for a normal plane. You'd probably be better off taking away the front fins that stick straight up. You get better stability if your center of lift in the yaw direction is as far back as possible. | ||
|
shin ken
Germany612 Posts
On April 06 2016 06:44 stenole wrote: Usually yaw control is something you do not need for a normal plane. You'd probably be better off taking away the front fins that stick straight up. You get better stability if your center of lift in the yaw direction is as far back as possible. Yes! That was also one of my conclusions after further testing. That's what I love about KSP - the progress. At the end of the day, still using only basic parts, I was able to come up with a design that could fly further (400 units fuel instead of 300) without relying on gimmicks like above! And I learned quite a bit about aerodynamics and fuel efficiency along the way. Although gimmicky designs can be fun as well :D - I'm still proud of my rocket with boosters alligned in a swastika that could spin itself to orbit ^^ (at that time, I wasn't really understanding SAS, so my solution was building my own SAS by spinning the rocket while climbing through the atmosphere - swastika boosters are ideal for that) | ||
| ||
WardiTV Map Contest Tou…
Playoffs Day 3
Classic vs SHIN
MaxPax vs Percival
herO vs Clem
ByuN vs Rogue
[ Submit Event ] |
|
|
|
Ladder Legends
Bunny vs GgMaChine
ByuN vs Percival
MaxPax vs Krystianer
Solar vs Cham
uThermal 2v2 Circuit
BSL
Sparkling Tuna Cup
WardiTV Map Contest Tou…
Ladder Legends
BSL
CranKy Ducklings
Replay Cast
Wardi Open
[ Show More ] |
|
|
EPT
