Devnote Tuesday: Back to Work!

[panzer1b]

I VERY much hope that we will finally be able to deal with slightly higher part counts without it devolving into an unplayeable lagfest. Im not asking for 3000 parts at once rendered or anything, but if the lag started at say 1000 instead of teh current 500ish, thatd be a great step to allowing us combat oriented players to have massed 5 vs 5 capital ship on ship battles all in the same area!

Im also looking forward to improved physics, and hopefully some of the odd bugs like the claw can be fixed as well (i just lost a HK-103 interceptor in a battle thanks to the claw kraken explosion bug and it almost cost me a victory!). Either way, im hoping squad doesnt focus as much on extra features and adding more parts as much as bug fixes, performance improvements, and similar issues.

Ohh, and i also hope we can get a small hotfix to the aero. 1.0.2 really overdid the drag increases, and made reentry heat a utter joke even without any heat shields, unless ofc you do something unbelievably stupid or come in at 30km/s. Right now you can reenter pretty easiuly from a minmus trajectory with a command pod+parachute combo, this really should burn unless you make multiple braking passes without a shield on the bottom! i know 1.0.3 is planned, but we dont exactly know what will be in it and if whats in it will actually fix both the super ANNOYING mem leak as well as the aero that is arguable a little too much closer to 0.90's souposphere then i personally like. 1.0 was not perfect, heat was a tad over the top, and perhgaps the atmo was a hair too thin, but 1.0.2 felt like an overreaction that was too much of a decrease to heat and too much increase to drag. Something halfway betwene the 2 is what id love to see stock.

[TORCHWOOD]

Mine is when putting info in the description. I end up ripping off about half of the rocket, and having to re-build, or have the presence of mind to just control + Z it.

[Tourist]

Glad to see Squad got some R&R and a chance to catch-up in person. Its important for teams, especially ones based in different countries to meet face to face occasionally.

Unity 5? gotta be honest I don't understand all this programming stuff, but the smart kids seem to think this will make the game better ... somehow.... so... woohooo, ya Unity!

[K3RBAL KOL4]

Back to work, good. Maybe you can finish the Beta now.. You know, Aero that works and isn't broken... massive memory leaks from a "feature" that can't be turned off permanently you have to disable every [snip] launch... Oh and Science Bay destroying entire craft... just because. Oh yeah, fail Aero model...And oh yeah, contracts.. "Explore Duna and test Rhino at 35000m after getting temperatures at 3 locations on Duna" after your first orbit of Kerbin.. Cuz yeah that makes sense, just like testing first stage only rockets at specific altitudes on anything but Kerbin before even unlocking 4th science tier... Wow, just wow.

Edited May 20, 2015 by Vanamonde

[mattinoz]

.....snip ....

One thing I didn’t expect would be a problem though, was a small but far-reaching change: Rigidbodies using Mesh colliders would previously collide just fine with other rigidbody mesh colliders, as long as one of them was a convex mesh. This is no longer the case. Now, if you have two rigidbody objects, both mesh colliders have to be convex. At first glance, this seemed fine, as all our parts are already convex, and the terrain isn’t a rigidbody object, so it can still have a con-convex mesh. So all is good then… except asteroids.

Asteroids are the one exception to this new convex collider rule. They have to be able to collide with other parts (for obvious reasons), and they also have to have rigidbodies, otherwise they wouldn’t be maneuverable. That put us in a bit of a quandary, because using a single convex hull around an asteroid is just not accurate enough. Asteroids can be very large, compared to spacecraft and Kerbals, and we wouldn’t want objects hitting (or grappling on to) invisible walls over and around the rock surface.

We half expected we might run across problems like this, which is the reason why we couldn’t take this upgrade as a small job, or even estimate how long it would take. This one issue in particular did have a solution, which took two days to implement. It’s a piece of code I’m actually quite proud of, 700+ lines to procedurally generate solid collision geometry, and all it does is make asteroids behave no differently from before. It can be a thankless job sometimes… But it was a lot of fun to code, in a geeky, procedural-mesh-enthusiast kind of way.

....

So would this lend itself to other uses within KSP?

Like say Surface features on the planets that could be interacted with like appear on the wish list all the time.

More so I'm thinking buildings around the space centre having more natural looking ground interactions that change with each upgrade. Not just a flat prepared base that exists from day one.

Even maybe and yes I'm clutching at straws here, player place-able buildings so we can have our own layout even leave the odd building out if we want. Open up the possibility of mod-buildings being placed at KSC or elsewhere.

[tntristan12]

So would this lend itself to other uses within KSP?

Like say Surface features on the planets that could be interacted with like appear on the wish list all the time.

More so I'm thinking buildings around the space centre having more natural looking ground interactions that change with each upgrade. Not just a flat prepared base that exists from day one.

Even maybe and yes I'm clutching at straws here, player place-able buildings so we can have our own layout even leave the odd building out if we want. Open up the possibility of mod-buildings being placed at KSC or elsewhere.

+1 (though I also believe this will never happen). A man can dream! Anyway, maybe we can have crater generation now? Huh? Huh!?

[Engineer of Stuff]

I’ve been organizing the update plan for the next few updates (yes, plural) while figuring out how to most effectively organize the team to speed up Multiplayer’s development.

Oh, gosh. It is here!

This deserves a HypeTrain right now. (jk)

EDIT: hey, post number 555!

[uglyduckling81]

have massed 5 vs 5 capital ship on ship battles all in the same area!

That already exists. It's called Dreadnought

https://www.greybox.com/dreadnought/en/

[Francois424]

Ahhh Unity 5 ...

I may moan about stuff I am not happy with every now and then, but never forget this: You guys are the best !

[Tourist]

Ugghh... multiplayer, why does everything have to have multiplayer. Do you know what would ruin KSP for me.... landing on Eve and finding thousands of "ShaDoWKILLers" and "DEthhatER8s" have already placed their flags all over the place.

I know, I know, it will be optional, its not going to be an MMO etc etc.

Still, KSP seems one of those games uniquely suited to solo play. But of course... everything has to have multiplayer tacked on otherwise Gen Y loses interest. grumble grumble, you kids get off my lawn!!!!

[Hcube]

"i even got to jump/fall out of a plane."

It´s licky that 1.0 is out, otherwise you would not have been able to climb back up in the plane !!

[Hobbes Novakoff]

Ugghh... multiplayer, why does everything have to have multiplayer. Do you know what would ruin KSP for me.... landing on Eve and finding thousands of "ShaDoWKILLers" and "DEthhatER8s" have already placed their flags all over the place.

I know, I know, it will be optional, its not going to be an MMO etc etc.

Still, KSP seems one of those games uniquely suited to solo play. But of course... everything has to have multiplayer tacked on otherwise Gen Y loses interest. grumble grumble, you kids get off my lawn!!!!

The point of multiplayer for me, I'd say, is doing co-op or competitive play with a couple of close friends. Theoretically you could have a gigantic gazillion-core hub (and since multithreading will be a thing, the gazillion cores will make a difference) and have a bunch of people on at once, but I think the point of MP is just to play with a few others.

- - - Updated - - -

Also... Multithreading. Yahoo!

[T-Bouw]

It sounds like the Unity 5 engine replacement will result in a better everything update!

[worir4]

What if when they finish the upgrade to unity 5, Unity 6 comes out???!!!

Just joking, Awesome stuff guys!

[ShadowZone]

I VERY much hope that we will finally be able to deal with slightly higher part counts without it devolving into an unplayeable lagfest. Im not asking for 3000 parts at once rendered or anything, ...

But I am!

I am currently working on another >1000 part vehicle. Rendezvousing with one of my resume 500 part stations is impossible at the moment.

Dev Notes are back. Yay! Looking forward to the U5 stuff

And yeah, please fix the memory leaks, thanks.

[KerbMav]

Drag was increased to raise the dV needed to orbit to closere where it was back in soupy-days, right? What about raising the atmosphere instead of cheating some factors in?

(I would like to try a somewhat bigger Kerbin, but even if the mod does not include adding and moving planets and moons around, I do not like the need to build rockets with x-times the number of parts than in stock ...)

[WeirdCulture]

However, it’s not without its own set of changes that are (you guessed it) also backward-incompatible

will the update break savegames?

[sal_vager]

Far too early to know that WeirdCulture, but as the save format is custom made and not dependant on Unity, it's very unlikely.

[Alewx]

Just as a sidenote, the kerbalspaceprogram.com site is broken when it comes to the links of the news with the IE11.

[Laie]

Drag was increased to raise the dV needed to orbit to closere where it was back in soupy-days, right?

If I understand it correctly, it was supposed to increase stopping power. IIRC, in 1.0.0 a simple free-falling Mk1 pod would make ~350m/s at 5km. But if parachutes can fail for going too fast, you need a sufficiently thick atmosphere to bring vessels down to deployment speeds.

[LordFjord]

Great to see the devnotes again.

And Unity5, beside having lots of great new things, it will also mean a heck of a lot of porting work for nearly all mods.

[Linear]

Explanation?

We've had regular updates take 8 months, and this update is rewriting or tweaking a LOT.

Plus, they're likely gonna add stuff besides just the U5 port. I still believe it will be quite a while.

Updating to U5 doesn't and will not take too long (on the scale of a year...)

Pay my wages for a year and i'll rewrite KSP for you.

[Ival70]

Ah, fantastic! I hope that maybe the new UI will look a bit more sleek and space-y. Also, I have a feeling that a LOT of mods will be broken when this releases..

[J.Random]

<b>Felipe (HarvesteR): </b> ...

One thing I didn’t expect would be a problem though, was a small but far-reaching change: Rigidbodies using Mesh colliders would previously collide just fine with other rigidbody mesh colliders, as long as one of them was a convex mesh. This is no longer the case. Now, if you have two rigidbody objects, both mesh colliders have to be convex.

...

This one issue in particular did have a solution, which took two days to implement. It’s a piece of code I’m actually quite proud of, 700+ lines to procedurally generate solid collision geometry, and all it does is make asteroids behave no differently from before. It can be a thankless job sometimes… But it was a lot of fun to code, in a geeky, procedural-mesh-enthusiast kind of way.

Isn't it more logical to make asteroids procedural a bit, building them from multiple parts, each having a convex collider? It could affect performance a bit, but it could also add a bonus feature of partial/gradual destruction on impacts or reentry.

[heng]

10 pages and noone reacts to the audio revamp?!? *weeeeeee*

does that mean Chatterer will get to be stock finally? my favourite and absolutely essential mod!