KSP Weekly: The night that turned into day

[Theysen]

1 hour ago, SupperRobin6394 said:

What will the impact of 1.4 be on mods, is there any possible estimate on that?

They're switching unity's engine version right? 

Anything with code in it - expect a minor to major breakdown probably

Anything without code in it - should maybe work okayish, unless they get rid or change some of the existing modules and functions. 

 

Let's hope for the best!

[razark]

2 hours ago, Theysen said:

Anything with code in it - expect a minor to major breakdown probably

Indeed.

Always assume things will break with an update.  Make backups, and be prepared to wait for a while to rebuild your particular mix of mods.  Especially with a major change.

Sometimes, you'll be surprised, though.

[DMagic]

On 12/2/2017 at 5:19 PM, Eric S said:

As a matter of fact, that's probably the main performance increase we'll see.  The new Unity is using a much more up to date version of Mono, which has much better garbage collection.  I can't say how much of an improvement the end user will see, but I'm hoping for something nice.

The garbage collection remains unchanged as far as I can find out. I can't find any official mention of it, but no one seems to think anything has changed as of Unity 2017.2: 

Upgrading the GC is listed on the roadmap (you have to scroll down a bit or just search for garbage) as still in development, which seemingly puts it beyond 2017.3, or 2018.1.

In any event, the solution to GC problems is in better code for stock (does stock KSP have much of a GC problem?) and non-garbage () GUIs for mods. Nothing loads up the system with garbage faster than a moderately complex UI, particularly one with lots of separate text fields. There are certainly lots of other sources of garbage, but the UI is the only one that you can't really do anything to fix short of entirely replacing it with the new system.

[linuxgurugamer]

8 hours ago, SupperRobin6394 said:

What will the impact of 1.4 be on mods, is there any possible estimate on that?

They will all need to be recompiled, except for pure parts mods

[Eric S]

1 hour ago, DMagic said:

In any event, the solution to GC problems is in better code for stock (does stock KSP have much of a GC problem?) and non-garbage () GUIs for mods. Nothing loads up the system with garbage faster than a moderately complex UI, particularly one with lots of separate text fields. There are certainly lots of other sources of garbage, but the UI is the only one that you can't really do anything to fix short of entirely replacing it with the new system.

I actually agree completely on this point, and my ongui-must-die attitude contributes to most of my "I should write a mod mostly duplicating that mod some day" thoughts, with misc efficiency concerns second, and just wanting to change something coming in a remote third place.

As I understand it, GC in newer versions of Mono is better, but not hugely so, more a matter of optimizations to the existing routines, so not anything revolutionary or astounding.  I expect writing code that minimizes the amount of garbage generated to be a much more significant factor even with the latest Unity.

[0111narwhalz]

15 hours ago, SupperRobin6394 said:

What will the impact of 1.4 be on mods, is there any possible estimate on that?

It could be quite severe, or it could just require a recompile. SQUAD may try to keep the impact down, but it's still likely to kill several mods with restrictive licenses and absentee authors off. Same as any update.

[goldenpeach]

So, we will be able to create asteroids with a set trajectory and measure the time it took the player to complete a mission...

 

Who's up for creating a 3D race track using asteroids to set the boundaries?

[Raptor9]

10 hours ago, goldenpeach said:

Who's up for creating a 3D race track using asteroids to set the boundaries?

Or a huge gauntlet of asteroids that the player has to maneuver through without getting hit.