Devnote Tuesday: Turning Up The Heat!

[SQUAD]

<figure data-orig-width="480" data-orig-height="92" class="tmblr-full"></figure>Hello everyone,

It’s still a bit of a mess at the office: the cabling that was installed turned out to not meet the contracted specifications and so the renovation crew came back in to tear down previously completed work. By now the renovation is starting to feel almost natural, but we’re positive these final glitches will soon be worked out and we can resume working normally soon™.

The PAX pox Max caught last week seems to have spread around the office, and while he’s been catching up on the tasks he had to put aside while he was at Pax both Daniel (DanRosas) and Andrea (Badie) have fallen sick. Thanks Max!

Andrea and Kasper (KasperVld) have been working on the number-crunching side of our Facebook and Twitter accounts to determine which content resonates the most with you, so we can share the most interesting things the community does. They’re also busy organizing a contest that should go live this week, and are preparing to take in applications for smaller channels to join the Media Group next week. For those of you who aren’t aware, the Media Group consists of YouTubers and livestreamers who collaborate and also get early access to new versions of Kerbal Space Program to showcase the new features of an update. The Media Group consists of both small and large media channels, but are united in the high quality content they put out.

On the development side things have been progressing a bit more slowly than we anticipated, but we’re optimistic we’ll be moving the Unity 5 update to QA testing very soon. Felipe (HarvesteR) ended up having to rewrite the backend code for the orbit icons because it turned out to be largely incompatible with Unity 5’s new user interface system. If you’re wondering what the orbit icons are, think of the apoapsis and periapsis icons, as well as the ascending and descending nodes. Of course, rewriting the code provides an excellent opportunity to improve upon these sometimes hard-to-use interface elements.

The new implementation should be far more robust and flexible, and should make the map icons much more responsive in the UI too. The ‘pinning’ system – allowing you to keep the apoapsis or periapsis captions visible by clicking on them – is one area that has our specific attention here. Each icon now has full control over the input events they receive. They are also all tied to a MapObject, which is a representation of a point in the map view. That means every icon can potentially be a camera target. Not all of them will be camera-focusable, but this should be a good addition to improve usability in the map.

Meanwile Jim’s (Romfarer) work on the staging interface is progressing steadily. The Staging interface is a core part of the game, and is used in many different scenes such as the Flight scene and the Editor scenes, not to mention many mods. The code behind it then is also some of the oldest found in the game, and indeed the staging list can be found in some the oldest versions of the game. The interface used an OnGUI() method, which makes transition from the old system to Unity’s new event driven interface system a bit trickier than many parts of the interface that were updated previously. The system is being completely redesigned, rather than just rewritten.

As we mentioned earlier we’re still gunning for QA testing of the main development branches soon, and no person is more involved with that part of development than Ted. He’s been chasing us around, making sure any and all internal requirements for the update are swiftly being met. Part of that work was finalizing an upgrade to the web servers that host the website, forums, wiki, bug tracker and other systems. Staff changes over the past few months have delayed this task, but now that the paperwork is being finalized we should see a lot more reliability on the servers, including better peak-load handling capabilities. Once the upgrades take effect various glitches on the bug tracker and wiki should be remedied, including issues with uploading images and embedding mathematical formulas.

Currently the QA team is focusing on their first task for update 1.1: testing a series of thermal fixes Nathanael (NathanKell) has been working on. Most areas of the thermal code ran solidly, but we’ve seen many bug reports of specific configurations that would cause issues. Ultimately we decided that the conduction code needed to be rewritten to account for these cases as there were two core issues here, both causing large amounts of energy to be transferred.

The first core issue revolves around the constraints on the amount of heat that could be conducted in multi-part configurations. Individual part-part conduction are constrained so that a source part can’t transfer so much energy from itself that the receiving part would become hotter than it. This system worked perfectly in situations where parts are connected to each other serially, but unfortunately a bunch of hot parts attached to a cold part would each transfer a lot of energy, causing the receiving part to potentially overheat very quickly. To overcome this problem the new system analyzes groups of parts as a whole and uses the resting temperature of the entire ‘cluster’ to work out the heat transfers.

The second core issue boils down to the fact that the heating system didn’t have a particular order in which it would conduct heat between parts, causing previous constraints to be undone when a hot part was being processed after a cold part. The heating system now processes all parts in order of hottest to coldest to overcome this. These two changes seem to have made the thermal spikes a thing of the past.

Heat convection has received some attention as well, or rather the input data of the system have. The game will now support convective shock temperature multipliers for different celestial bodies to simulate different types of atmospheres such as the low-shock-temperature hydrogen atmosphere of a gas giant, and the upper atmospheres of the atmospheric planets have been revised to have a more exponential falloff.

Finally, the analytic thermodynamic code has been improved to give it a more gradual transition between temperatures, which can be configured globally and on a per-part basis. All this is a fancy way of saying that the game will support (for example) refrigerated interiors.

Marco (Samssonart) has been busy getting used to living in the Netherlands and has been getting started with his studies. He’s already sharing knowledge gained from his optimization and vectorization course with the team. This course deals with high- and low level optimization and parallelization of code. Although the Unity engine takes care of the code at the low level, there’s certainly possibilities to further optimize the high level code in the game.

That’s it for this week, be sure to post any questions you have below!

Edited September 8, 2015 by KasperVld
Typo!

[HafCoJoe]

A contest? Oh boy!

Sorry to hear you all got sick . I hope it wasn't too bad!

I also hope you can all come up to Seattle again sometime in the future. It makes me sad I wasn't able to shake a few hands while I was at PAX. Alas I was there on Monday, not Saturday. At least I won a new mouse in the Rocket League tournament!

It's always fun hearing about the information and announcement pipeline! I hope it's more fun then pain to sift through all the responses... It's also good to hear you guys talking about the heating systems. I'm not sure if it's me over-imagining or not, but it appears as though there is considerably more lag in high part count ships with the addition of heating and mining functions/processes. Regardless it's nice to hear you guys talking about any related information on that. And good luck Samssonart with your studies! Optimization and simplification is the only thing that keeps me from going mad when updating my own mods.

Good news to hear as always.

- Joe

Edited September 9, 2015 by Avera9eJoe

[Endersmens]

You know, for the first time I actually appreciated this format. It was actually easier to follow, because the information was organized into separate paragraphs. You can still kinda look for a specific dev name to see what they've been up to, but you can also just read it through smoothly.

I hope they continue to be that easy to read.

[Moesly_Armlis]

"The ‘pinning’ system – allowing you to keep the apoapsis or periapsis captions visible by clicking on them."

Does this mean we no longer need to hold the mouse over the marker to display the value?

Such improvements make the game much friendlier.

[Virtualgenius]

How long are you going to be testing this new format as I really DONT LIKE IT, please go back to the old format it was easier to skim read the bits I was interested in

[nightingale]

[...] and the upper atmospheres of the atmospheric planets have been revised to have a more exponential falloff.

Interesting - does this mean that the atmosphere (and atmospheric drag) will extend beyond the "atmosphere depth" value of 70,000 km?

Of course, to make this meaningful, we'd need to have drag (even if simplified) applied to unloaded craft...

[hermes47]

Does this mean we no longer need to hold the mouse over the marker to display the value?

The pinning system already exists for the Ap and Pe markers; came in around 0.20 IIRC. I think this new systems means that it could be applied to all the other markers as well, like closest approach, AN/DN, SOI transitions

[nightingale]

"The ‘pinning’ system – allowing you to keep the apoapsis or periapsis captions visible by clicking on them."

Does this mean we no longer need to hold the mouse over the marker to display the value?

Such improvements make the game much friendlier.

It actually already does this. What is needed is a way to communicate this ability to new player (and old ones that don't know about it). No idea if this is mentioned in the new tutorials, that might be one place for it.

Edited September 8, 2015 by nightingale

[sumghai]

Excellent dev notes, especially the detailed look into NathanKell's work in tweaking the thermal system.

[magico13]

Does this mean we no longer need to hold the mouse over the marker to display the value?

As mentioned, the pinning system is already in place. You can click on any marker to force it to stay up. But if you have several markers near each other (or overlapping, like you might with small maneuvers on a maneuver node) then they frequently overlap and it selects all of them and it makes it hard to see much of anything. I imagine that's one of the things they're addressing in particular this time around.

Also, like last week I liked the new layout this time around.

[FleetAdmiralJ]

Finally, the analytic thermodynamic code has been improved to give it a more gradual transition between temperatures, which can be configured globally and on a per-part basis. All this is a fancy way of saying that the game will support (for example) refrigerated interiors.

I have no idea how this might be applied in the game, but for some reason this sounds awesome

[NathanKell]

nightingale: Nope, it's just referring to changing the pressure curves and other existing atmospheric properties like molar mass (in addition to adding the new ones regarding shock temp multipliers).

[RocketPilot573]

Again some very juicy and well written dev notes! This new format feels less awkward now.

[Moesly_Armlis]

The pinning of any marker is not functional at the beginning tier tracking.

I have to use this technique of grouping and mouseover to view the information simultaneously or mod it with KER.

[ShadowZone]

The PAX pox Max caught last week seems to have spread around the office, and while he’s been catching up on the tasks he had to put aside while he was at Pax both Daniel (DanRosas) and Andrea (Badie) have fallen sick. Thanks Max!

Get well soon you guys!

They’re also busy organizing a contest that should go live this week, and are preparing to take in applications for smaller channels to join the Media Group next week. For those of you who aren’t aware, the Media Group consists of YouTubers and livestreamers who collaborate and also get early access to new versions of Kerbal Space Program to showcase the new features of an update. The Media Group consists of both small and large media channels, but are united in the high quality content they put out.

Speaking as a KSP-YouTuber, color me intrigued

Meanwile Jim’s (Romfarer) work on the staging interface is progressing steadily. The Staging interface is a core part of the game, and is used in many different scenes such as the Flight scene and the Editor scenes, not to mention many mods. The code behind it then is also some of the oldest found in the game, and indeed the staging list can be found in some the oldest versions of the game. The interface used an OnGUI() method, which makes transition from the old system to Unity’s new event driven interface system a bit trickier than many parts of the interface that were updated previously. The system is being completely redesigned, rather than just rewritten.

So does that mean that a lot of mods have to be rewritten as well? In general: What can modders expect in regards to how much work they are going to have to put into their products to respond to the update?

Is there sort of a "modder's early access" so modders can get a bit of a time advantage so we don't have to wait weeks for our favorite mods to be available?

Heat convection has received some attention as well, or rather the input data of the system have. The game will now support convective shock temperature multipliers for different celestial bodies to simulate different types of atmospheres such as the low-shock-temperature hydrogen atmosphere of a gas giant, and the upper atmospheres of the atmospheric planets have been revised to have a more exponential falloff.

Finally, the analytic thermodynamic code has been improved to give it a more gradual transition between temperatures, which can be configured globally and on a per-part basis. All this is a fancy way of saying that the game will support (for example) refrigerated interiors.

I can't say I understood everything, but... frozen snacks, yay!

Keep up the good work and thanks for the update!

[Starhawk]

Thanks for the devnotes!

Very detailed and informative.

improve upon these sometimes hard-to-use interface elements.

The new implementation should be far more robust and flexible, and should make the map icons much more responsive in the UI too. The ‘pinning’ system – allowing you to keep the apoapsis or periapsis captions visible by clicking on them – is one area that has our specific attention here.

Meanwile Jim’s (Romfarer) work on the staging interface is progressing steadily. ... The system is being completely redesigned, rather than just rewritten.

I'm very happy to hear some details on this work. This has the potential to add an enormous amount of polish to the look and feel of the game.

Happy landings!

[nismobg]

This is a pleasant surprise , thank you for finally getting work done on the map nodes, and very good to hear about the fix on the thermal system. Right now it is very wonky as is. I cant express how much everybody is looking towards this update. Maybe much more then the 1.0 one !

[smjjames]

It actually already does this. What is needed is a way to communicate this ability to new player (and old ones that don't know about it). No idea if this is mentioned in the new tutorials, that might be one place for it.

Not consistently though and sometimes it doesn't want to turn off.

Cool to hear that you guys are taking the opportunity to, well, renovate the code.

[Mad Rocket Scientist]

but we’ve seen many bug reports of specific configurations that would cause issues. Ultimately we decided that the conduction code needed to be rewritten to account for these cases as there were two core issues here, both causing large amounts of energy to be transferred.

YYYYYYYYYEEEEEEEEEEEEEEEEESSSSSSSSSSSSSSSSSSSSSSSSS!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Thanks NathanKell!!!!!!!!!!!!!!

[smjjames]

Get well soon you guys!

Speaking as a KSP-YouTuber, color me intrigued

So does that mean that a lot of mods have to be rewritten as well? In general: What can modders expect in regards to how much work they are going to have to put into their products to respond to the update?

Is there sort of a "modder's early access" so modders can get a bit of a time advantage so we don't have to wait weeks for our favorite mods to be available?

I can't say I understood everything, but... frozen snacks, yay!

Keep up the good work and thanks for the update!

Some of the major modders already have early access, but yeah, some detailed information that modders would want would be useful, or maybe just a guide in general what modders can expect will change, etc.

[Lukaszenko]

nightingale: Nope, it's just referring to changing the pressure curves and other existing atmospheric properties like molar mass (in addition to adding the new ones regarding shock temp multipliers).

I'm hoping this has something to do with the instant exploding from the slightest touch of the Joolian atmosphere?

[TheMoonRover]

...preparing to take in applications for smaller channels to join the Media Group next week...

How small? I currently have a whole 5 subscribers. Although I have almost finished a fairly long video about the game's easter eggs. (That Magic Boulder is proving most elusive. Yes, I know it's no longer there.) But hopefully that'll get some attention.

[Cydonian Monk]

Not consistently though and sometimes it doesn't want to turn off.

Cool to hear that you guys are taking the opportunity to, well, renovate the code.

I'm also not sure it worked at all the last time I tried to pin Apo/Peri nodes on OS-X. But I agree - the way it works now, when it works, is a bit frustrating. Glad to hear this is being retooled.

[ArmchairGravy]

Get well, everybody! Remember, the green men should be inside the game, not working on it.

[HafCoJoe]

Get well, everybody! Remember, the green men should be inside the game, not working on it.

This. lol