Devnote Tuesday: An Overdue Break

[SQUAD]

Hi everyone!

Today’s devnote will be very short indeed, the reason for that is a planned break for the ‘core’ developers, who after months of coding get a week off. While the experimental testers are carefully documenting the first batch of bugs development does continue slowly and we’ve already seen several builds since the weekend.

Chris (Porkjet) spent time fixing a few graphical glitches, after helping Bob (RoverDude) animate the inflatable heat shield. It proved to be an unexpected challenge to skin and animate it in a way that would look somewhat properly ‘folded together’ and still correctly manipulate the mesh normals. The latter is important as these are not only for looks, they also directly influence how drag cubes are generated.

Obscure issues with the part search functionality are this week’s nightmares for Brian (Arsonide): this weekend he ran into some tag “collisions” while refining the search functionality. In essence, the game would return unexpected search results in some cases: the word “station” would match with the search term “ion”, which may be a small problem in stock games, but part mods might increase this exponentially as the search algorithm takes part descriptions into account. Brian decided to create a system whereby we can deviate from the standard matching algorithm on a set of defined problem tags. You can now prepend a question mark to get exact results: “?ion” only matches the word ion exactly. For players, this means your searches will be far more accurate. For the modding community, it means that you will have plenty of options when dodging tag collisions on your own parts. There will be a guide after release on how to use this system.

We’re still fixing bugs that exist in 1.0.5 and prior as well: Nathanael (NathanKell) solved a longstanding issue with parachutes and occlusion. As it turns out when we changed the occlusion system to be based on contact area rather than node area during 1.0 development the occlusion multiplier handling got broken. When you (semi-)deploy a parachute, what’s supposed to happen is that occlusion will no longer apply because the parachute is waving around way far from your stack, so it doesn’t make sense for your stack to occlude it. However, since the multiplier was not used, that disabling of occlusion was not occurring and thus your stack will occlude many (stack-mounted) parachutes when they are only semi-deployed. That is now fixed for 1.1.

Finally, Nathan (Claw) mended the “deploy” option on control surfaces. Previously the deploy direction was dependent on the control surface position relative to the control part, now it will deflect based on its own orientation. This also means that control surfaces can now be deflected based on their symmetry attachment method. Mirror attached parts will deflect in a mirrored way and radial attached parts will deflect radially.

Meanwhile Andie (Badie) and Kasper (KasperVld) are working to set up the Media Group and KSPTV group for showing you previews of KSP 1.1 close to the release of the public pre-release testing branch. Aside from that a new group of forum moderators may appear amidst you soon, and planning for the console releases has already started.

Finally, a poem by Joe (Dr Turkey), who is also stuck at work.

Vacations is all I ever wanted,
To roam free, boxer clad and deep in thought in my war room.
Wondering whether to make friends with the Yor or the Altarians,
while feasting on spirits and pizza.

Vacations is all I ever needed,
Crush the Torians and their puny fleets, educate the Drengin on the finer points of xenocide,
Learning the hard way to not mess with insect people.
Chugging chocolate milk, cramming cookies till the end of time.

Or at least the end of the day.

Alas, vacation was not in my fates this day!
Curse the heavens, stomp the earth,
The Big Three have come knocking,
it’s certification’s time to pay!

GDC looms it’s hydra head,
problems and logistics cloud the hours,
Type, type, type, has my dog been fed?
Send, receive, send, should send my girlfriend some flowers…

[Guest]

gg squad

 

best devnote

[Dres]

5 minutes ago, SQUAD said:

Meanwhile Andie (Badie) and Kasper (KasperVld) are working to set up the Media Group and KSPTV group for showing you previews of KSP 1.1 close to the release of the public pre-release testing branch.

Can't wait! (Now I wish I had gotten KSP on steam )

[CAKE99]

Just now, Dres said:

Can't wait! (Now I wish I had gotten KSP on steam )

Yeah, I think I got KSP before I got Steam though.

[Superfluous J]

9 minutes ago, SQUAD said:

Vacations is all I ever wanted,

Vacation, had to get away!
Vacation, meant to be spent alone...

Sorry. It's in my (social) contract that I have to answer that call.

[regex]

 

Sounds pretty cool.  Looking forward to the release.

[Laie]

Good call on the control surfaces. Ran into that the other day and getting it to work as I thought it should wasn't easy.

6 minutes ago, SQUAD said:

You can now prepend a question mark to get exact results: “?ion” only matches the word ion exactly.
 

Hmm, could you maybe consider using some kind of established regex syntax? Like, perl or grep or whatever? That would be ^ to match a beginning and/or $ at the end.

Would help those who are used to it, and those who are not may as well learn the right stuff.

[Sanic]

Fourth? Hooray?

Seventh. Triple ninja'd.

Love the poem, Turkey! 

[regex]

4 minutes ago, Laie said:

Hmm, could you maybe consider using some kind of established regex syntax? Like, perl or grep or whatever? That would be ^ to match a beginning and/or $ at the end.

Would help those who are used to it, and those who are not may as well learn the right stuff.

This.  So much this.

E: Although I suspect this is to be used in tag definitions within parts, not in the search box.  Still, the syntax just doesn't look ... right.

Edited March 9, 2016 by regex

[StrandedonEarth]

Jeb, Bill, Bob, Wehrner, Gene, and all the koys in the VAB are enjoying a vacation while they wait to see what 1.1 will bring.

That poem seems like it should be to the tune of "Vacation" by the Go-Gos, but I can't quite seem to make it fit. I need to refresh my memory... I was going to post the video, but on second thought I won't inflict that on the forum community

[DuoDex]

4 minutes ago, regex said:

This.  So much this.

E: Although I suspect this is to be used in tag definitions within parts, not in the search box.  Still, the syntax just doesn't look ... right.

Someone wants to get a special mention in the game.

[godefroi]

So, a question that I've never seen answered: does the giant inflatable heat shield protect things that aren't stack-mounted? I.e. if I have radially-mounted stuff that is visually "inside" the shield radius, will it still burn up?

Also, is there only one size, or is there one for each size?

[terryammon]

Yaaa let the party's begin  Great work :-)

[theonegalen]

Speaking of control surfaces, is it at all possible to do something about control reversal when the center of mass moves from behind to in front (or vice versa) of the control surface? It's entirely too possible with midwinged planes, especially if they are made to resemble real-world aircraft.

Or maybe I should just get good.

Thanks for all your hard work, SQUAD! Excited for 1.1, and to play the game unmodded again!

[Superfluous J]

23 minutes ago, Laie said:

Good call on the control surfaces. Ran into that the other day and getting it to work as I thought it should wasn't easy.

Hmm, could you maybe consider using some kind of established regex syntax? Like, perl or grep or whatever? That would be ^ to match a beginning and/or $ at the end.

Would help those who are used to it, and those who are not may as well learn the right stuff.

While I agree it should be like this, it won't fix the stated problem, that searching fro "ion" finds "station"

I suggest something in quotes finds the exact word while not in quotes would find the extra stuff. "ion" would not find station, but just ion would.

Or allow wildcards and require finding only full words. So ion would not find station, but *ion would.

[Claw]

17 minutes ago, godefroi said:

I.e. if I have radially-mounted stuff that is visually "inside" the shield radius, will it still burn up?

Yes, radially mounted things hidden behind the heat shields (or any other parts) should be more protected from direct heating.

18 minutes ago, godefroi said:

Also, is there only one size, or is there one for each size?

One size for the inflatable.

5 minutes ago, theonegalen said:

Speaking of control surfaces, is it at all possible to do something about control reversal when the center of mass moves from behind to in front (or vice versa) of the control surface?

Deployment now follows this, but control for the ControlSurfaces will still auto-flip based on their relationship with the CoM. That's so as the CoM moves forward, Pitch and Yaw inputs still give proper pitch and yaw response.

28 minutes ago, Laie said:

Good call on the control surfaces. Ran into that the other day and getting it to work as I thought it should wasn't easy.

They will hopefully be a bit more intuitive now.

[Guest]

42 minutes ago, SQUAD said:

We’re still fixing bugs that exist in 1.0.5 and prior as well: Nathanael (NathanKell)

<SNIP>

Finally, Nathan (Claw)

Sooooo confusing.

42 minutes ago, SQUAD said:

Aside from that a new group of forum moderators may appear amidst you soon

*Looks around*  Fun!

If there are moderators that are going to be new to the forums, welcome.  Otherwise, congrats to anyone who gets it.  

[Dman979]

53 minutes ago, SQUAD said:

Meanwhile Andie (Badie) and Kasper (KasperVld) are working to set up the Media Group and KSPTV group for showing you previews of KSP 1.1 close to the release of the public pre-release testing branch. Aside from that a new group of forum moderators may appear amidst you soon, and planning for the console releases has already started.

@Badie, @KasperVld- can I get some information about this? I might want to apply to be a Mod, but I don't know what the job description would be.

[Glaran K'erman]

Enjoy the break guys, well deserved I am sure!
Dr.Turkey, sorry not this time....but hey great poem!

[Kerbart]

Why not use regex for the part search filters? It adds a lot more flexibility and hardly requires any extra effort to implement?

[Dman979]

6 minutes ago, Kerbart said:

Why not use regex for the part search filters? It adds a lot more flexibility and hardly requires any extra effort to implement?

Because I don't think @regex wants to become a part of the game.

[godefroi]

1 hour ago, Claw said:

Yes, radially mounted things hidden behind the heat shields (or any other parts) should be more protected from direct heating.

Ah, good. I just ran a test, and this seems to be the case with the existing non-inflatable parts as well? Or, at least, larger heat shields seem to have more thermal mass, and therefore can survive harsher reentry profiles, even for non-radially mounted parts? Forgive me if this information is available somewhere, but I haven't seen it.

[Claw]

Just now, godefroi said:

Or, at least, larger heat shields seem to have more thermal mass,

Slightly larger thermal mass, but also can ablate faster. Meaning that the larger heat shields can give off heat faster than the smaller ones.

And yes, current mechanics in game is that parts out front shield ones further behind from reentry heat. Surface and stack mounted.

[godefroi]

3 minutes ago, Claw said:

And yes, current mechanics in game is that parts out front shield ones further behind from reentry heat. Surface and stack mounted.

Does this take into account part size? I.e. if something "looks" like it should shield a surface-mounted part, then it probably will? Something that sticks out further than the parts above it will be less shielded (or unshielded)?

Edited March 9, 2016 by godefroi
s/part/parts

[Neutrinovore]

2 hours ago, SQUAD said:

Previously the deploy direction was dependent on the control surface position relative to the control part, now it will deflect based on its own orientation. This also means that control surfaces can now be deflected based on their symmetry attachment method. Mirror attached parts will deflect in a mirrored way and radial attached parts will deflect radially.

What?  That's how the parts SHOULD deploy.  So you're saying that, if I build a rocket in the VAB that has 4 fins, radially attached, and then I add a control surface to each fin, then when I launch the rocket and then 'pull up' on the stick, the rocket won't pitch positive but will instead just roll around on its axis?  That makes no damn sense...