I quit.

[SpacedCowboy]

Maybe you should submit your resume to Squad. ( Seriously )

I like your ideas

[Xyphos]

29 minutes ago, SpacedCowboy said:

Maybe you should submit your resume to Squad. ( Seriously )

I like your ideas

No thanks, I quit Software Engineering 3 years ago.
I still do a bit of code here and there as a hobby, primarily data hacking and reverse-engineering, but nothing serious.

 

2 hours ago, Slam_Jones said:

 

"Now that I'm done raging, allow me to continue raging:"

"I don't understand the development process of video games, nor can I comprehend the time and resources required to make a proper game engine.  However, I, a random person off the internet, obviously am better qualified to direct the development of a video game than someone who does so professionally."

oh I understand. it's not that freaking hard at all. just requires proper planning and outlining of the game's needs; one that supports LUA scripting for mods would be ideal, but hey.. what do I know? according to you, I know nothing, so it must be true.

Edited June 3, 2016 by Xyphos

[Superfluous J]

4. The user has realistic expectations and they are met or exceeded.

[regex]

1 hour ago, Xyphos said:

LUA scripting for mods

Why?  Modding KSP is already incredibly easy and C# is an easy language to pick up.

E: Not to mention that LUA isn't supported by Unity (AFAIK) so that's dev time to be spent integrating and ensuring it works.  Meanwhile, C# comes for free and works just fine.

Edited June 3, 2016 by regex

[steve_v]

3 hours ago, Xyphos said:

I've come up with several hypothesis's:

1. The user is in denial. KSP always works, even when it's not.
2. The user has low expectations and is overwhelmed with euphoric joy when something actually works.
3. Unity. building a half-assed shell around a rotten, un-maintainable core expecting a one-size-fits-all approach to actually work, instead of writing a custom-tailored game engine designed specifically for the tasks required.

Oh, almost certainly #3, at least from what I have been able to discover. However...

This is irrelevant. KSP is sold as a complete game, and it's not at all an unreasonable expectation that it should work properly as shipped. Which game engine it's built on is not something that should concern the end-user, that's between Squad and whoever supplies the tools they use.

As I see it, the situation is very simple:
1.0.5 was stable (32bit address space limits are not instability).
1.1.x is not stable for a significant number of people, on the same system. As far as I can tell Squad has no clue how to fix this.

This is all the end-user should need to know, expecting players to research game-engines, submit bug-reports to Unity3d, or cut Squad some slack for their poor choice of tools is ridiculous.

 

2 minutes ago, 5thHorseman said:

4. The user has realistic expectations and they are met or exceeded.

Aren't you lucky.
I have realistic expectations too, the same expectations I have of any commercial software... Expectation: it works as advertised and doesn't crash.
 
If it has stability problems on certain system configurations you advertise this fact so customers can make an informed purchase decision. If it has stability problems on all configurations, or the cause is still unknown, you advertise that too... or better yet, stay in beta until it's fixed or compatibility issues have been identified.
The relevant bug-report has more information, crash reports, logs etc. than any other I have seen... yet so far I see zero progress. Perhaps there is more on the private tracker, but that does nothing to assure that Squad know what they're doing or that a fix is even in the works.

Had I purchased KSP at 1.1 release I would have filed for a refund somewhere around the third time it crashed to desktop, or the first time I heard "It's a Unity problem, no fix."
I am still here because I know how good the game can be when it runs properly. Sadly, right now it most certainly does not.

As far as this topic goes, I'm not at all sure what you're trying to achieve - none of the points made are "answers" - the only acceptable answer is fixing the instability.

[samstarman5]

Gonna say, you lost me at "If I was on the development team things would have been different...." You aren't the first to say that, and sadly you aren't going to be the last.

If you honestly were so motivated you would already be working on your own project instead of trying to sound so wise here. That alone just makes me take any of your other arguments not seriously at all.

End of line.

[Bill Phil]

Developing an engine is really hard to do. It'd probably be a good idea to use an already existing rendering engine and just* write the physics aspect.

* as in only do that part. Still wicked hard

Edited June 3, 2016 by Bill Phil

[magnemoe]

6 hours ago, Xyphos said:

After taking some time off to cool down from my KSP ragequit, I've spent some time thinking about why KSP appears to work well for some people and not others and I've come up with several hypothesis's:

1. The user is in denial. KSP always works, even when it's not.
2. The user has low expectations and is overwhelmed with euphoric joy when something actually works.
3. Unity. building a half-assed shell around a rotten, un-maintainable core expecting a one-size-fits-all approach to actually work, instead of writing a custom-tailored game engine designed specifically for the tasks required.

My money's on Unity. I'm not saying it's a bad system, but clearly, it's not working out. if I was on the dev team, I would have pulled Unity's plug 5 years ago.
in fact, 5 years would have been plenty of time to write a custom game engine three times over, from scratch.

Well my KSP 1.12 is more stable than previous versions, mostly because of no 4gb limit I think.
Its likely to be an conflict with something else, it might be hardware or software, it might be multiple conflicts too. This is why it only happens to some players
And yes this is likely to be unity's fault 

[Xyphos]

3 minutes ago, Bill Phil said:

Developing an engine is really hard to do. It'd probably be a good idea to use an already existing rendering engine and just write the physics aspect.

no, it's not.
graphics engine is easily ran in a looping stand-alone thread, OpenGL or DirectX API calls are made here and the graphics card handles the rest.
physics engine can be broken down into multiple, multi-threaded sub-engines; collision, trajectory, velocity, angular velocity, etc...
combine all that together and you've got a well-oiled game engine that fires on all cylinders.

[Hary R]

43 minutes ago, Xyphos said:

no, it's not.
graphics engine is easily ran in a looping stand-alone thread, OpenGL or DirectX API calls are made here and the graphics card handles the rest.
physics engine can be broken down into multiple, multi-threaded sub-engines; collision, trajectory, velocity, angular velocity, etc...
combine all that together and you've got a well-oiled game engine that fires on all cylinders.

It certainly is not, from an engineering point of view but at least from a financial point of view it is... it's pretty expensive no matter what. 

That's why most indies dev don't create their own engine, and why Squad stick to Unity to begin with. I agree that due to the scoop of KSP it should have been done with custom engine but we have unity so we have to deal with it.

As for your OP I'm one of the lucky one that don't have any CTD in 1.1.2 stock so:

1- No it really just work... at least for me it does

2- No, if I put money on something, it should work as intended. don't mind few bugs but CTD no way.

3- In part but I think the CTD have something to do with KSP 1.1.2. something is wrong and I hope 1.1.3 solves it for you. The wheels on the other hand is fully on Unity.

Edited June 3, 2016 by Hary R

[Jovus]

Is it too late to go back to Alpha? Can we call this 0.27.2?

[Bill Phil]

19 minutes ago, Xyphos said:

no, it's not.
graphics engine is easily ran in a looping stand-alone thread, OpenGL or DirectX API calls are made here and the graphics card handles the rest.
physics engine can be broken down into multiple, multi-threaded sub-engines; collision, trajectory, velocity, angular velocity, etc...
combine all that together and you've got a well-oiled game engine that fires on all cylinders.

Game engines are hard to make. Why do you think Unity exists? Unreal? Source? Because it's hard to make your own engine. It's time intensive, resource intensive, and much more.

Engines have to be flexible enough to run well on most computers, but that necessitates compromise.

Physics engines can be broken down into smaller components, yes. But there are a LOT of them.

If it was so easy to make your own game engine, Squad would've done it. They're not stupid. It's just hard to make game engines. Otherwise everyone would have their own. And if you're making a new game you could always make a new engine. But here's the truth: it's hard to do. It may not seem hard to you. It's all a matter of perspective. But it is hard. It takes time. Time to code. Time to test. Time to improve. And time is money.

[r4pt0r]

35 minutes ago, Xyphos said:

no, it's not.
graphics engine is easily ran in a looping stand-alone thread, OpenGL or DirectX API calls are made here and the graphics card handles the rest.
physics engine can be broken down into multiple, multi-threaded sub-engines; collision, trajectory, velocity, angular velocity, etc...
combine all that together and you've got a well-oiled game engine that fires on all cylinders.

DUDE! Thats awesome! When is your release date? probably before the end of the weekend with how easy it is. Cant wait to play your game XD XD XD XD XD XD

[Curveball Anders]

4. Stability issues are limited to certain software/hardware in combination with use of certain features of the game. Which means that a large group of users don't have any crashes what so ever.

[Superfluous J]

1 hour ago, Xyphos said:

no, it's not.
graphics engine is easily ran in a looping stand-alone thread, OpenGL or DirectX API calls are made here and the graphics card handles the rest.
physics engine can be broken down into multiple, multi-threaded sub-engines; collision, trajectory, velocity, angular velocity, etc...
combine all that together and you've got a well-oiled game engine that fires on all cylinders.

 

Is it? Sweet. I can't wait to see yours. All those posers who've made them up to this point will be so red faced.

[Majorjim!]

Un-modded KSP for me now seems stable 90% of the time. Yes I get random crashes when staging sometimes or other game events but they are not constant. Having said that of ALL the games I own KSP is by far the most unstable.

[DChurchill]

15 hours ago, Xyphos said:

After taking some time off to cool down from my KSP ragequit, I've spent some time thinking about why KSP appears to work well for some people and not others and I've come up with several hypothesis's:

1. The user is in denial. KSP always works, even when it's not.
2. The user has low expectations and is overwhelmed with euphoric joy when something actually works.
3. Unity. building a half-assed shell around a rotten, un-maintainable core expecting a one-size-fits-all approach to actually work, instead of writing a custom-tailored game engine designed specifically for the tasks required.

1. Nope. It really works. It crashed last night when I alt-tabbed out and started Ventrilo. KSP's fault? I'm not even having any real wheel problems. The small rover wheels want to spin in circles, but that's a known issue that's being addressed.

2. Hardly. I've bought games on Steam, played them for 2 minutes and said "Whelp, this is a piece of excrement." and promptly deleted it.

3. See #1. Although, I suspect the cause of most peoples problems IS Unity. Correct me if I'm wrong, someone, but isn't 1.1.2 built on an outdated Unity version that is going to be updated?

Edited June 3, 2016 by DChurchill

[cantab]

16 hours ago, Xyphos said:

if I was on the dev team, I would have pulled Unity's plug 5 years ago.
in fact, 5 years would have been plenty of time to write a custom game engine three times over

With a large and well-funded development team that would be the case. But that's something KSP has never had. When it started of course it *couldn't* have it, an off-the-shelf engine was the only way. And based on the claims by ex-developers last month, management never really moved out of that small team cheap mindset even once KSP started making the megabucks. And there's probably too much "put into" Unity now for a custom engine to look attractive.

It's widely thought that technically speaking, the crashes may be caused by Unity's code, but I still hold Squad responsible for their choice of engine. They might not be able to directly fix the Unity code, but Squad are Unity's customer and can act accordingly - push Unity to get things fixed, choose the appropriate version of Unity (and it looks like Squad made an error in that respect for KSP 1.1), or if need be find a workaround.

I feel that a KSP 2 would do a lot better with a custom engine. Unfortunately I fear Squad could end up with KSP 1 much like Microsoft did with WIndows XP - they've got so much work to do fixing the current release that they can't put effort into the sequel (which after all Squad haven't even publicly mooted yet). Also space is very "in" in gaming at the moment, and KSP was quite early on that wave, so I think there's a strong chance that KSP 2 could end up missing it.

Edited June 3, 2016 by cantab

[p1t1o]

1. Offensive.

2. Offensive.

3. Offensive.

 

 

[The_Rocketeer]

Note: this post kinda circles round to topic.

Years ago I played a lot of World of Warcraft. Loved the lore, eventually tired of the grind. Anyway, I went to see Warcraft: The Beginning at the cinema this week (pretty good I thought, for a first out - at least it wasn't a LotR clone). Well, I got home and realised that the movie had left me with many questions and confusions about the Warcraft lore (which when I played WoW I knew pretty well). So I hopped on youtube and started watching videos.

One I came across was this one: The History of Warcraft (Documentary). Basically it tracks the progress of Blizzard the company in parallel with their videogame releases, from a tiny indy company (hello SQUAD!) to a massive MMO heavyweight.

What I didn't realise is it basically took 2 iterations of the Warcraft line (Warcraft and Warcraft 2) before Blizzard actually became solvent enough to start pacing their own development (from 1994-1999, the company was doing what Squad is doing, pushing out the best they could and hoping for returns further down the line). The result of this pacing (when it came) was Warcraft 3, a game that took the RTS scene by storm and paved the way for World of Warcraft (which was based on a more developed version of the Warcraft 3).

So where am I going with all this? Well, I think this speaks volumes about our expectations from SQUAD. Seems to me they're only just getting to the same place Blizzard was in in 1998/1999. Squad is also an indy company, but one without 20 years in the industry and multiple titles and multiple lines. Let's cut them some slack in our 'expectations' and stop assuming that they're doing it wrong because their vision hasn't fully coalesced yet.

[Vanamonde]

Overlapping threads have been merged. And since the second one is turning into personal attacks just like the first one did, why don't we all just move on to other discussions now, please.