KSP Loading... A closer look into Update 1.6

[Gapone]

Let's move it all to the grand Discussion thread.

[KSK]

Some of the comments in this thread don't match up to my own experience.

KSP runs just fine on a Mac. Or at least it did when I last played it on my late 2013, mid-specced iMac. Which, incidentally plays (or again, it did the last time I tried) more graphically intensive games such as Elite: Dangerous acceptably well too. As a Mac gamer my definition of acceptable may be more forgiving than others though - I'm certainly not trying to pretend they're gaming powerhouses but as has been mentioned, with a proper port (or a game that's written with them in mind from the outset), they do okay.

I'm not familiar with the technicalities here but I play other Unity based games (Battletech most notably) and they seem fine. Graphically, they're more detailed than KSP but I'm betting that they're not running much in the way of detailed physics calculations behind the scenes. Performance isn't great but given that I'm running them on a machine that doesn't meet minimum spec, that's no surprise.

Limited experience to go on but I haven't had any joystick woes on my Mac. Plugged in my el cheapo HOTAS setup and everything just worked without so much as an ignored axis or failed keybinding.

The plural of anecdote isn't data but I thought I'd at least add my contrary anecdotes to the pile.

[Lisias]

On 1/1/2019 at 6:36 PM, T1mo98 said:

They're still a semi-indie developer, with nowhere near the resources of something like DICE, Blizzard or EPIC. 

Also keep in mind that if they were to switch game engines, they would essentialy have to rewrite the entire game or huge parts of it. The decision to use Unity was a very sensible one when the game was first developed, and there might just not be the resources they need to completely fix it.

It's very easy to criticize them for the decisions they made, but it isn't easy to actually make those decisions.

Being the reason I don't criticize Squad by choosing  Unity. At that time, no one could foresee what they are going to do.

I did some Orbiter development, and let me tell you; its "better', it's "faster" - but it's not for the faint of heart. KSP would not reached this huge user base and add'on ecosystem without a fairly decent game engine with an easy to use programming language.

I make my point again: I blame Unity Technologies for the current state of affairs. They are the ones responsible for the mess, the worst problems KSP had (or have) are exactly due Unity idiosyncrasies and absolute lack of proper support.

(give a look on the memory management of this krap and see by yourself)

2 hours ago, KSK said:

Some of the comments in this thread don't match up to my own experience.

KSP runs just fine on a Mac. Or at least it did when I last played it on my late 2013, mid-specced iMac. The plural of anecdote isn't data but I thought I'd at least add my contrary anecdotes to the pile.  [cut by me]

Being the reason I think the problem is Unity.

KSP itself is the same code-tree for all platforms - or pretty almost.

9 hours ago, steve_v said:

The "Don't switch to Unity 2017 until we have a solution for the utterly broken input stack" decision seems like a fairly straightforward one to me.

Unless the problem they are solving by switching is even worse than the consequences.

Sometimes, you just can't win - you must choose the less painful way to loose the battle.

[shdwlrd]

1 hour ago, Lisias said:

Sometimes, you just can't win - you must choose the less painful way to loose the battle.

I'm going to steal this statement. Too perfect for the no win BS in life.

[steve_v]

4 hours ago, Lisias said:

Unless the problem they are solving by switching is even worse than the consequences.

If you can spot anything worse than completely unusable joystick support that was fixed in 1.4.0, please enlighten me. As I recall it had a bunch of new bugs over 1.3.1, and precious little else except some DLC I'm not going to buy.

As far as I can see, the unity Unity 2017 update didn't fix any major bugs at all, or at least not any that I have encountered.
Frankly, I suspect it had something Squad wanted for the DLC, and Squad considered DLC more important than not shafting their Linux users.

[Lisias]

25 minutes ago, steve_v said:

Frankly, I suspect it had something Squad wanted for the DLC, and Squad considered DLC more important than not shafting their Linux users.

Nothing is more important than incoming. A bug-less software that is not selling have very little value as a product.

But, again, this is wild speculation. I don't have the smallest clue about what is really happening.

[Superfluous J]

37 minutes ago, steve_v said:

not shafting their Linux users. 

I assume you mean their Linux users who will not play the game without a joystick.

[klgraham1013]

6 hours ago, Lisias said:

Being the reason I don't criticize Squad by choosing  Unity. At that time, no one could foresee what they are going to do.

I certainly don't think we can blame them for choosing Unity.  We can, though, blame them for moving to another version of Unity with seemingly little gain to the player (so far), and a lot of new bugs.

[steve_v]

2 hours ago, 5thHorseman said:

I assume you mean their Linux users who will not play the game without a joystick.

How about "All their non-windows users who don't like having features removed without warning".
It's joysticks for now. In a space/flight/physics sandbox, one where people go so far as to build custom control hardware no less, input support is a pretty important feature.

Seeing as squad is fine with leaving this regression completely unaddressed, what will it be next time?
Sound completely stops working, and Squad does nothing? Wait, we've been there.
Maybe the game will crash every couple of hours, and they'll just leave that unfixed...  We've been there too.

If this is the level of effort Squad is willing to put into their Linux port, the single reason I ever considered buying their product, I'm done with this company.
Done and liquided off, because I love the  game when it works properly. If only Squad could get their excrements together long enough to release a version free of infuriating bugs, just one would be enough.
 

[Clonos]

I love the new Terrier.

[The Dressian Exploder]

Guys, who'd love to see some stock clouds  on kerbin? EVE is nice, but it can make your game a little LagGY, Kerbin would look way better with clouds as part of the stock game, though.

[Curveball Anders]

1 hour ago, Spacenerd Kerman said:

Guys, who'd love to see some stock clouds  on kerbin? EVE is nice, but it can make your game a little LagGY, Kerbin would look way better with clouds as part of the stock game, though.

Not me.

I see that as a complete waste of developer resources and cpu/gpu cycles.

But that's just me.

[Tyko]

2 hours ago, Curveball Anders said:

3 hours ago, Spacenerd Kerman said:

Guys, who'd love to see some stock clouds  on kerbin? EVE is nice, but it can make your game a little LagGY, Kerbin would look way better with clouds as part of the stock game, though.

Not me.

I see that as a complete waste of developer resources and cpu/gpu cycles.

But that's just me.

It would be interesting but as you point out it could slow down computers and there are much more important things I'd rather see developed first. Also, if they put out a compromised set of LQ clouds to keep frame rates up they're just going to get flack for not having an HQ experience and if they release HQ clouds they're going to liquid off people on lower spec machines.

I'd say "leave this to modders"

[Phreakish]

48 minutes ago, Tyko said:

It would be interesting but as you point out it could slow down computers and there are much more important things I'd rather see developed first. Also, if they put out a compromised set of LQ clouds to keep frame rates up they're just going to get flack for not having an HQ experience and if they release HQ clouds they're going to liquid off people on lower spec machines.

I'd say "leave this to modders"

Weather is inherent when an atmosphere is present, and increases immersion as well as game appearance. Adding features doesn't preclude development on other fronts, etiher. 

Settings changes are normal for people to vary depending on computer/machine capability. Those with lower spec machines can simply turn off clouds, those in the middle can get LQ versions and those with beastly rigs (or low part count ships) can run HQ. It's not rocket surgery... 

Mods are great, but shouldn't be relied on to fill in the gaps. Just because that's the way it's been, doesn't mean it should continue. I say bring on stock clouds, weather, etc and give us settings to vary it or leverage it for a more challenging experience. 

[Curveball Anders]

16 minutes ago, Phreakish said:

Adding features doesn't preclude development on other fronts. 

Settings changes are normal for people to vary depending on computer/machine capability.

Adding features doesn't preclude other development but does compete for resources.

Every single user selectable option adds to code complexity and muddles up QA (drop the settings in the current game and look at the gfx bugs that pop up).

[Phreakish]

1 minute ago, Curveball Anders said:

Adding features doesn't preclude other development but does compete for resources.

Every single user selectable option adds to code complexity and muddles up QA (drop the settings in the current game and look at the gfx bugs that pop up).

That's a Squad problem. Doesn't change my wishlist of features. 

Bugs are bugs - the focus on fixing/addressing them shouldn't change. Code complexity should not be a user concern - a good developer will figure it out. 

[Curveball Anders]

1 minute ago, Phreakish said:

That's a Squad problem. Doesn't change my wishlist of features. 

Bugs are bugs - the focus on fixing/addressing them shouldn't change. Code complexity should not be a user concern - a good developer will figure it out. 

I didn't challenge your wishlist.

I only noted that adding features does compete for developer resources.

[Phreakish]

2 minutes ago, Curveball Anders said:

I didn't challenge your wishlist.

I only noted that adding features does compete for developer resources.

Agreed, but it's not a reason preclude the addition of a feature. Just changes the priority of it. 

1.6 to me gives me hope that plenty of bandwidth will be available to not only tackle the existing issues, but to fill gaps that have long existed. Bring on 1.7

[Tyko]

Has anyone seen a mention of a 1.6.1 with a fix for the PartDatabase bug?

I'm really impressed with 1.6 so far, but this bug is driving me bonkers and I can only imagine how horrible it is for any parts modders who are trying to update/test their mods

Edited January 7, 2019 by Tyko

[klgraham1013]

22 minutes ago, Tyko said:

Has anyone seen a mention of a 1.6.1 with a fix for the PartDatabase bug?

I'm really impressed with 1.6 so far, but this bug is driving me bonkers and I can only imagine how horrible it is for any parts modders who are trying to update/test their mods

At this point, I think that, the Z fighting bug, etc. will probably be pushed to the 1.7 update.  Will just have to live with them tell then.  This may be the new post Take Two plan.

Edited January 7, 2019 by klgraham1013

[Damon]

18 hours ago, Tyko said:

Has anyone seen a mention of a 1.6.1 with a fix for the PartDatabase bug?

I'm really impressed with 1.6 so far, but this bug is driving me bonkers and I can only imagine how horrible it is for any parts modders who are trying to update/test their mods

I got to a point where I just test everything in 1.5 and call it a day. My install is just too big to launch ksp twice every time I change my Gamedata folder. It’s really annoying and I hope it will be fixed soon.

[OrbitsR4Sissies]

On 1/4/2019 at 9:23 AM, Spacenerd Kerman said:

Guys, who'd love to see some stock clouds  on kerbin? EVE is nice, but it can make your game a little LagGY, Kerbin would look way better with clouds as part of the stock game, though.

Better to keep the stock release simple and add these in through mods.

With 1.6's performance changes, I have begun happily running Sci-Fi Visual Environments on my 3-year old iMac with nVidia graphics very well with medium textures. It provides clouds and dust without EVE and related mod overhead.

As Macs draws everything, with my relatively low 8GB RAM, I would lighten the textures and resolution to strike a balance and SvE works very well. I'm a happy camper there.

Per a related message: Turn DOWN your KSP graphics settings with a Mac when trying out visual mods then carefully increase some game settings to check performance to your liking. Macs has superior drawing ability but lower RAM and other factors of the OS mean that you're telling your Mac to do too much. Be sure NOT to run multithreading apps like Google Chrome when using KSP.

[SpektrSoyuz]

Great to see all of the new engine textures, wishing for more in the future in 1.7 or so on.
 

On 1/4/2019 at 9:23 AM, Spacenerd Kerman said:

Guys, who'd love to see some stock clouds  on kerbin? EVE is nice, but it can make your game a little LagGY, Kerbin would look way better with clouds as part of the stock game, though.

I don't know, it would definitely need to come with an option to enable/disable the clouds. Would be great for people who don't like installing mods

[Murdabenne]

Maybe its time for the Devs to declare the next "point" release will be nothing other than bug fixes and Unity work-arounds?  No changes to the API, no new content, just fix the broken stuff, and push hard upstream to Unity for the things that cannot be fixed in code at the app level.  Every once in a while, a "bug sprint" followed by a "refactor" sprint (from problems identified during the bug sprint) was a handy thing to have when I was directing a product team.

[Curveball Anders]

18 hours ago, Murdabenne said:

Maybe its time for the Devs to declare the next "point" release will be nothing other than bug fixes and Unity work-arounds?  No changes to the API, no new content, just fix the broken stuff, and push hard upstream to Unity for the things that cannot be fixed in code at the app level.  Every once in a while, a "bug sprint" followed by a "refactor" sprint (from problems identified during the bug sprint) was a handy thing to have when I was directing a product team.

This is a good idea.

But not everyone in a dev team are coders or QA and you can't just replace people for a point release (at least not very many).

You can however change the focus of a release, concentrate on current issues, and maybe add some temporary hired guns (as they've done before) to clean up code and resolve some standing issues.

 

Edited January 13, 2019 by Curveball Anders