SPOILER-FREE: Data-mining yields good news

[RocketRockington]

1 hour ago, Periple said:

No, my theory is that Star Theory failed to build on top of KSP1's code base even though they tried, and that attempt wasn't halted until development was transferred to IG. If they had succeeded on building on it, we would be in a very different situation!

Gotcha ok.  Sorry I get a little confused debating between you and K^2 who seems to believe everything went fine with Stary Theory's engineering.  This makes more sense.  My apologies.

So yes, the one counter-arguement there is then why announce a ship date in 2021.  Your counter-arguements to that have seemingly been  'they knew it was a fake date, it was just because (business reasons) that they announced a public fake date so they could string along both investors and the fans.  (Please correct me if I'm wrong here about your position)  My response to that is 'why wouldn't they just have said 'we have to take longer, we just restarted the game' and then we're at an impasse, because I don't see a good reason for them to do that, and a lot of my distrust of KSP2 stems from their routine willingness to fabricate lies to their audience to make themselves look good in the moment.

Once you can believe that Nate et. Co are willing to just outright lie about the development of the game just to look good in the moment - even if its gonna bite them in the butt down the road - you can throw out a lot of what you guys seem to be hopeful about - for instance, that the devs actually were playing a fun game internally, that they were playing a version of multiplayer internally, etc.   And very little of the dataminining has shown any working code, afaict.  Bits and pieces of tuning files and namespaces, sure, but no swaths of seemingly functional code that just needs to be switched on.

Edited March 11, 2023 by RocketRockington

[Periple]

46 minutes ago, RocketRockington said:

So yes, the one counter-arguement there is then why announce a ship date in 2021.  Your counter-arguements to that have seemingly been  'they knew it was a fake date, it was just because (business reasons) that they announced a public fake date so they could string along both investors and the fans.  (Please correct me if I'm wrong here about your position)  My response to that is 'why wouldn't they just have said 'we have to take longer, we just restarted the game' and then we're at an impasse, because I don't see a good reason for them to do that, and a lot of my distrust of KSP2 stems from their routine willingness to fabricate lies to their audience to make themselves look good in the moment.

I don't remember having said much about announcing a 2021 ship date at all, but now that you ask, my theory is that maybe the creatives who moved from ST to IG just didn't have the technical understanding to know just how screwed the ST codebase was and thereby wildly underestimated the difficulty? I haven't managed to trace a single engineer who did make the move, and it would've taken the new ones a while – like a month or two, minimum? – before they felt confident enough to say hey guys, this isn't going to work, we need to start over?

And sorry man, but I'm not willing to believe that Nate is willing to outright lie to us. I am willing to believe he can be unrealistically or unreasonably optimistic, and I'm certainly willing to believe he leaves things unsaid.   I do know that I have a tendency to believe the best about people until proven otherwise though!

Thing is, in my experience this works most of the time. I mean, there ARE extremely ego-ridden people with a way too inflated sense of their own ability and importance in the industry, and they do a lot of damage, but these people are a pretty small minority. Most people are, by and large, what it says on the box. And nothing what Nate has said, or what I've heard about him from people who know him, suggests that he's remotely like that!

[RocketRockington]

4 minutes ago, Periple said:

I don't remember having said much about announcing a 2021 ship date at all, but now that you ask, my theory is that maybe the creatives who moved from ST to IG just didn't have the technical understanding to know just how screwed the ST codebase was and thereby wildly underestimated the difficulty? I haven't managed to trace a single engineer who did make the move, and it would've taken the new ones a while – like a month or two, minimum? – before they felt confident enough to say hey guys, this isn't going to work, we need to start over?

So your theory is now - they failed to do the job building off the codebase of KSP for 3 years.  They then lose all their engineers, and their codebase, and with the experience of failing to get the job done in 3 years, a bunch of non-coders with no engineering team decide that 'oh, obviously we'll be able to finish this by 2021'.    I realize Hanlon's razor is a thing, but at some point that level of rank 'optimism' by people who are paid, theoretically experienced developers slips into malice just through that much incompetence.

[Periple]

24 minutes ago, RocketRockington said:

So your theory is now - they failed to do the job building off the codebase of KSP for 3 years.  They then lose all their engineers, and their codebase, and with the experience of failing to get the job done in 3 years, a bunch of non-coders with no engineering team decide that 'oh, obviously we'll be able to finish this by 2021'.    I realize Hanlon's razor is a thing, but at some point that level of rank 'optimism' by people who are paid, theoretically experienced developers slips into malice just through that much incompetence.

My theory is roughly this:

1. T2/PD starts to negotiate buying the KSP license somewhere in 2016. This is finalized in 2017. Then PD starts looking for a suitable work-for-hire studio to develop KSP2 while SQUAD carries on with KSP1, and settles on Uber, perhaps because of Nate's pitch.
2. Uber is renamed to Star Theory and starts work on KSP2 in mid to late 2017, with a really optimistic 2-year development cycle. They believe they can do this by building on the KSP1 codebase. 
3. By mid 2019 it becomes obvious that they can't deliver. PD starts negotiations with them about how to get the project back on track. These negotiations break down completely, and PD decides to cut bait. However, they still really dig Nate's pitch/concept and the creative approach, so instead of looking for another work-for-hire studio, they propose to found an internal studio and hire them as the core team. The engineering team doesn't come with them, either because PD doesn't want them, or because they preferred to stay with ST. 
4.  Early 2020, IG starts building a new engineering team. Some way into that process, but after the regrettable 2021 release date announcement, the new engineering team makes it clear that the codebase is a dead end and they need to start over. 

Clearly the 2021 release date announcement was a bad call. I don't know what was behind that, but I will point out that at that time it wasn't Nate or IG calling the shots, it was PD. Maybe PD felt they had to announce it for 2021 for internal corporate reasons – to get the funding to found IG from T2. That kind of thing isn't unheard-of!

Clearly there were other similarly bad calls: the EA was botched – at the very least in expectations management. So on those points we're in agreement.

But that's also where we part ways. I don't think KSP2 is built on the KSP1 codebase, and I think it is fundamentally sound – not exceptionally competently managed, produced, or executed, but not exceptionally badly either: nobody wants a buggy alpha, yet somehow non-buggy alphas are the exception in the industry, not the rule. It's just not finished. I'm also pointing out that this is IG's first release, and even if you have experienced people on board, that's always going to be difficult and messy. Releasing is a team sport and if a team hasn't trained together for it – by, you know, making releases – it's not going to go perfectly smoothly. The studio I'm at has been doing this for a few years now, and just last week we passed certification for all of our platforms (Sony, XBox, Switch) on the first try, a week ahead of schedule – and we had never managed that before. It took us a lot of practice to get there! 

So I remain quite optimistic that the project is fundamentally sound, the EA will turn around, we'll start to see tangible process, and over the next six months to a year they'll get their release cadence straightened out too. They will learn as they train together. I do recognize that your arguments aren't entirely without merit and you might be proven right in the end – but we'll all see how that plays out. If KSP2 is still a bug-ridden mess treading water this time next year, I will concede that I was completely mistaken about all of this!

Edited March 11, 2023 by Periple

[shdwlrd]

Why are ya'll (still) arguing about the past events? It's not like you can change them. There's nothing you can learn from them (because you weren't there.) Nothing about situation was released but just the broad strokes. (NDA's are good at suppressing relevant info about a situation. Plus it's generally bad form to publicly speak about companies goings on if nothing illegal or immoral is happening.)

So you're technically arguing about opinions. One thing you should know if you've been around for a few decades or more, arguing opinions without hard facts is a kin to arguing about religion. It becomes more emotional than factual based.

@RocketRockington I have to ask, is there any positives you see within KSP2? Almost all the posts I've seen from you are trashing the studio and the people that are employed by the studio by proxy. 

[Haustvindr]

6 hours ago, Chilkoot said:

But no evidence.  Apocrypha/heresay.

It's not just that we understand K^2 expertise, but also that there are a lot of people who mostly shares the same opinion. There's a suprising number of users in this forum related to the software development world, and even I, working in a field far away from games (but still development) agree that a restart happening is very probable.

But instead of dismissing my opinion right away, I can also offer a new insight and a different option for your train of though: legal issues, clauses or limits over the ownership of the code.

When people migrate to other workplace, they will absolutely bring their know-how, it's obvious since it is ingrained in their own memories and that's something you don't lose in a few weeks (except in rare cases I guess).

What is not always able to be migrated in the source code itself.  I'd even say "typically not migrated", but I don't know what common options contracts bring to the table in the US. This is due to companies covering their own bases against unexpected troubles.

When we e.g. develop a custom solution, the end result (software) may be property of the client (if they request it), meaning we cannot reproduce it for another client and a tad more constraints. What is never property of the client is the code itself, because in the code there is all the know-how, techniques, algorithms and so on, that we use as a company in a wide range of places. If we somehow relinquish the ownership of our code, we are opening a door (big and wide) for any client to attack us due to using fragments of "their" code in other projects.

As we don't know what specific clauses T2/PD had with Star Theory, we don't know if the source code could be witheld by ST. But, if companies over there are used to cover their bases, there is a chance that ST may witheld the source code. And if ST witheld the code, there's no code. And if there's no code, there's no way to keep working on it. And if you cannot keep working on it, you are left with nothing else than restarting.

So, speculation again, no evidence, apocrypha, heresay, yada yada. What I'm trying to convey to you is that there are possibilities as to why they WOULD NEED to restart from zero, and not just keep working on ST code. Very real possibilities.

[MarcAbaddon]

58 minutes ago, shdwlrd said:

Why are ya'll (still) arguing about the past events? It's not like you can change them. There's nothing you can learn from them (because you weren't there.) Nothing about situation was released but just the broad strokes. (NDA's are good at suppressing relevant info about a situation. Plus it's generally bad form to publicly speak about companies goings on if nothing illegal or immoral is happening.)

So you're technically arguing about opinions. One thing you should know if you've been around for a few decades or more, arguing opinions without hard facts is a kin to arguing about religion. It becomes more emotional than factual based.

@RocketRockington I have to ask, is there any positives you see within KSP2? Almost all the posts I've seen from you are trashing the studio and the people that are employed by the studio by proxy. 

Since we are still talking about the same actors, the past can give us an idea of what to expect in the future. That's why.

[Chilkoot]

7 hours ago, K^2 said:

We've never weighed an electron either, yet we can compute its mass indirectly.

Whoa there - let's not conflate the scientific method with forming a viewpoint based on layered tiers of speculation that can't be tested.  That's how fringies like Graham Hancock pull the wool over people's eyes.  Not suggesting you're being devious, but we need to be careful about what we believe and why, and not apply the same rules to very different ways of assessing certainty of knowledge.

Think about the burden of proof in science (like your example) or the justice system.  Even if "we weren't there", to have some kind of reasonable belief that development started afresh, we'd be looking for some kind of evidence in the form of documentation, access to the historical code repository, correspondence, etc., or even testimony from the people involved, which would be much weaker evidence.  A modern archaeologist wouldn't jump to the conclusion development was started fresh given the current evidence, neither would a prosecutor file charges based on the strength of the evidence we have.

7 hours ago, K^2 said:

Now, if you have reasons to doubt the model, that's fair.

There are two "real world" reasons we have to doubt that Intercept intended to throw out the code base from Star Theory, even though they may have eventually:

• The initially optimistic release window(s) communicated by Intercept shortly after they took over development of KSP 2 implied that development was continuing, not starting over.  These initial dates were completely unrealistic for a from-scratch dev cycle.  Were the developers/publishers hopelessly optimistic or even lying to save face?  Possibly.  Or they could have been relatively confident they could keep pushing forward with what they have.
• The fact that the code and creative assets produced by Star Theory under contract by PD would have legally been the property of PD/TT, and there was no investor guidance related to the write-off of that asset.

There's more than enough cause for doubt that we should be taking the position that we presume or suspect that development was rebooted somewhere along the line, not that we are sure it did on day one at Intercept.

6 hours ago, Periple said:

No, my theory is that Star Theory failed to build on top of KSP1's code base even though they tried, and that attempt wasn't halted until development was transferred to IG.

Based on what we do know, this sounds like the the most probable situation.  There were some off-handed remarks in interviews right around the initial announcement that alluded so *some* code or perhaps asset re-use from KSP 1.  This could have been the case with the code transported from Star Theory to Intercept as well - there was every intent (and timeline projection) to continue along with development, but eventually reality settled in and that code was binned in favour of a rewrite for the reasons a lot of people have cited here.  This is also supported to an extent by the significant, recurring delays and the state of the EA release several years into the process.

[K^2]

7 minutes ago, Chilkoot said:

Whoa there - let's not conflate the scientific method with forming a viewpoint based on layered tiers of speculation that can't be tested.

We can test it. We are testing it. This is what we're doing right now.

8 minutes ago, Chilkoot said:

hink about the burden of proof in science (like your example) or the justice system.  Even if "we weren't there", to have some kind of reasonable belief that development started afresh, we'd be looking for some kind of evidence in the form of documentation, access to the historical code repository, correspondence, etc., or even testimony from the people involved, which would be much weaker evidence.

For Big Bang, do you have a written evidence or a witness testimony?

Again, indirect evidence is evidence. You can use the effects something had as evidence if you understand the chain of causality that is applicable. And in cases where a critical experiment is impossible, that's the only way you test a theory. Instead of performing a critical experiment, you go over the existing observations and see if they make sense. If that wasn't the case, cosmology, archaeology, evolutionary biology, and many other disciplines would be impossible.

9 minutes ago, Chilkoot said:

The initially optimistic release window(s) communicated by Intercept shortly after they took over development of KSP 2 implied that development was continuing, not starting over.

In order for this to be an argument for a point you're making, you have to assume both good faith and competence from Intercept and Private Division. And if they are both competent and have been communicating with us in good faith, the rest of your argument does not follow. You can't have it both ways. You can't claim that we're seeing delays due to Intercept's incompetence and Private Division's negligence based on the claim that they were competent and diligent in producing the "initially optimistic release window". Unless you can tell me why they were competent in 2020 and stopped being so subsequently.

11 minutes ago, Chilkoot said:

The fact that the code and creative assets produced by Star Theory under contract by PD would have legally been the property of PD/TT, and there was no investor guidance related to the write-off of that asset.

They say you become a senior engineer when the amount of code you're deleting becomes greater than code you're adding. It's tongue in cheek, of course, but we rip out so much of the code all the time. Yeah, if you have a working product that you're expected to support, ownership and continued maintenance of the code is a must. But while the project is being built, stuff gets thrown out the window all the time.

Again, if the original version of KSP2 was based on KSP1 code and was supposed to be developed in shorter time, and what Intercept worked on was built on a new code base from scratch, then not throwing away most of ST's code would be silly. Yeah, I'm sure one could salvage something, but most of the code would be modification of KSP1's, which is getting thrown out. It'd be harder to adapt it to new code than to write it from scratch.

And we have strong evidence that Intercept's KSP2 is new code, because all of the craft and save files are JSON now with completely different structure. And the way the game loads modules is completely different. I haven't done as deep of a dive into modding as some people on Discord have, but what I've seen so far also indicates something completely new.

 

The only reason Intercept would be keeping Star Theory's code is if Star Theory was already writing KSP2 from scratch, without reusing any of the KSP1's code. And if that was the case, Star Theory did not have the time, given their limited engineering, to build a game by 2020 in the first place.

The only reasonable thing to assume based on all of this is that Star Theory was working off of KSP1 code, and Intercept wasn't. And if that's the case, then your argument to code reuse is invalid, and without it, the rest of the argument collapses as well. If Intercept only started on KSP2's code in 2020, there was no way to release the game in 2021. So I don't know why that release window was ever given to us, but there are a whole bunch of possibilities for this happening from mistakes to misunderstandings to lies, and all of them are more probable than the idea that Star Theory was going to write all of that by 2020 or that Intercept was going to get it done by 2021.

[shdwlrd]

8 hours ago, MarcAbaddon said:

Since we are still talking about the same actors, the past can give us an idea of what to expect in the future. That's why.

That's the same thinking as an ex-con will always be a criminal and the homeless will never stay in a home.

[Jacke]

21 minutes ago, shdwlrd said:

That's the same thinking as an ex-con will always be a criminal and the homeless will never stay in a home.

No it is not.  In both cases you mention, given appropriate support, change is possible and likely.

In the case of the leadership of Star Theory/Intercept Game and Private Division, as @MarcAbaddon said, we can expect their future behaviour to be similar to their past behaviour in similar conditions.  Because there would have to be reasons why it would be different.

[Alexoff]

19 hours ago, K^2 said:

This is a touch subjective, but I'm basing this on when people were being hired. Positions that got posted around Feb 2020 and filled summer-fall 2020 are consistent with game going from pre-production to full production.

I think it's right to consider the beginning of the development of the game as the beginning of funding for its development in T2 accounting. I'm sure they have data we'd rather not know. For example, why development started from scratch, although there were bases in the 2019 gameplay trailer, and after 2020 we saw some separate buildings of an unclear purpose only a couple of times. We know almost nothing about bases at all in 2023. About the same probability as you claim about a complete restart of development in 2020, I dare say that in 2019 we were shown just base models, and not functioning units.

[TackleMcClean]

Speaking of looking at the code, has there been any theorycrafting regarding the `ReplayFeature`?

[RocketRockington]

23 hours ago, shdwlrd said:

@RocketRockington I have to ask, is there any positives you see within KSP2? Almost all the posts I've seen from you are trashing the studio and the people that are employed by the studio by proxy. 

Sure.  There's some stuff I like.  Procedural wings are good.   Part recoloring.  The sound is miles away better (KSP1 never had a sound guy and used freeware music lol).  The VFX is way better than KSP1 (which didn't have a VFX artist) Orthographic view in the VAB.

None of those speak to the foundation of the game - they speak to having lots of developers and development time and budget and getting a few new ideas in.  And making a really good hire w/Howard Mostrom.  And there are lots of things KSP1 did right that they went backward on.  If they'd come out with a game that had a solid engine - reasonably performant, physics way better than KSP1 (no noodle rockets for instance) mod ready out of the gate and reasonably bug free, I wouldn't be trashing them even if they dropped the ball on getting any other features done - or even less features, though hard to imagine what even less might be - back to a single planet?

But they've done so badly in so many areas given so much time and money - my problem is what this does the the Kerbal franchise.  T2 is unlikely to ever do a KSP3 if this thing isn't profitable.  This is not a more solid foundation to build mods and other stuff on.  Could they fix it?  Sure.  Maybe.  Eventually.  But KSP1 never yoinked its physics engine or did a major refactor of that nature- games RARELY do that because you have to go dark for a long time and then rebuild/retune a lot of stuff that was built on top of it.  

So yeah - I'm critical because I think the wrong publisher and developer took my favorite franchise and mangled it.  Sorry if that offends you.

Edited March 12, 2023 by RocketRockington

[shdwlrd]

@RocketRockington No offense taken. I was just curious. There seems to be people griping about the game that haven't played it yet. I don't totally disagree with you. The game shouldn't have been released with show stopping bugs. I'm not playing it now because of the "revert to VAB bug." That should have been fixed prior release or patched soon afterwards since Intercept did know about it.

When the game is working, it's just as I would expect for an EA release. Lack luster performance, weird bugs popping up, some of the QoL stuff being broken. Nothing really to get upset about. 

Basically,  I'm saving my judgment until the first series of patches comes down. If all known game breaking bugs are fixed and some of the more annoying minor bugs are fixed, I'll be happy. If that doesn't happen, I'll be joining the chorus of "wth Intercept."

[Nicrose]

All I have to say is that they could have a working build with everything in it that runs at 5fps, but that by no means guarantees we will even get to see those changes ever. It's common in game dev to develop now, optimize later so while it may be being worked on, it probably isn't close to being released. I am hopefully but so far, I feel really burned by the company and the decisions, as well as the communication. Everything they say seems like a pr bandaid right now.