Devnote Tuesday: UI, more UI, and answers to frequently asked questions!

[Orb8Ter]

This is awesome the two gripes I have with KSP, fixing the UI and making use of multi-core processing, this update is worth the wait.

[Hektos]

That's the rub.

Squad thought the game complete enough to be a full release. You know, a finished game. You cannot release a product and when someone spends time using it, take something away form the player. Name one game that has done that and survived unscathed. It's simply unacceptable after release. If the game is not finished don't release it!

Balancing is fine and dandy. Breaking the game or changing it enough to significantly change gameplay after a 1.0 release is not acceptable.

World of Warcraft has had numerous patches which has changed the game significantly enough to impact playstyle (some of which annoyed many users).

But, when there is significant movement forward you sometimes have to make some compromises. Does it suck that you may potentially lose some progress, or craft? Of course! But if it's that worrisome to you, you can always make a backup of your KSP folder (outside of Steam if you use that) so you have an old version to fall back on to continue your game.

Do you not thing it a bit odd to complain when developers add new features for FREE, when many others would be happy to charge for the extra stuff?

[Superfluous J]

So what you're telling me is that Cities: Skylines is still in beta too?

Acutally, the day/night thing was not part of the original plan of the game. They gave it away for free so modders didn't have to maintain two sets of assets - those with night textures (windows lit, neon signs, etc) and those without. As they had to change the main game to SUPPORT night textures, why not modify it to acutally USE them?

[regex]

Acutally, the day/night thing was not part of the original plan of the game. They gave it away for free so modders didn't have to maintain two sets of assets - those with night textures (windows lit, neon signs, etc) and those without. As they had to change the main game to SUPPORT night textures, why not modify it to acutally USE them?

Oh sure, makes perfect sense.

I'm not one of those who was fooled by Squad's silly versioning scheme. As RIC points out above the game has changed significantly with the delivery method and everyone has their own ideas of what the terms are. The problem for Squad is that they believe their own hype about the software's development stage and therefore their users believe it too. Far better, in my mind, to just keep incrementing releases and call 0.27.x a "Major Release". Or whatever it takes for Steam to be happy. No one is confused, no hype necessary, continue delivering the product as usual, stuff happens.

vOv

[Superfluous J]

I'm not one of those who was fooled by Squad's silly versioning scheme.

Yup.

And for me, it's even simpler. Out of one side of their mouth, Squad said "Version numbers, alpha/beta/release status, all that doesn't matter, really. You shouldn't even be thinking in those terms," and out the other side they're screaming "WOO HOO 1.0 WE'RE OUT OF BETA AND FULLY RELEASED WOO HOO!"

[regex]

And for me, it's even simpler. Out of one side of their mouth, Squad said "Version numbers, alpha/beta/release status, all that doesn't matter, really. You shouldn't even be thinking in those terms," and out the other side they're screaming "WOO HOO 1.0 WE'RE OUT OF BETA AND FULLY RELEASED WOO HOO!"

And that's why Hype®™ is stupid.

[softweir]

... and out the other side they're screaming "WOO HOO 1.0 WE'RE OUT OF BETA AND FULLY RELEASED WOO HOO!"

They had to do that at some point in order to get sensible reviews. Most magazines and websites still don't review Early Release/Beta whatever games no matter how close they get to completion, but will review games that are "released" even if they need a massive patch to make them workable. In addition, a lot of players won't even look at Early Release games - won't read reviews, won't look at forums, they just ignore them and thus miss a lot of excellence. A lot of people were missing out on KSP quite unfairly, and Squad had to make the announcement in order to allow them into the fandom.

KSP has been playable - for all its faults and occasional game-breaking update - for years. Why keep it down for the lack of that label "Released"?

[regex]

Why keep it down for the lack of that label "Released"?

Neither 5th nor myself have been disputing that. The point was that Squad hyped that up and juggled version numbers, and confused the existing community, and set other such expectations through said hype, instead of just saying "This particular release, 0.27, can be considered a Release game." If the labels "Alpha" and "Beta", and version numbers as well, were meaningless, why not just say "You've all been playing a beta-quality game and with this release we consider it a "Release" product. KSP is live." and keep the same numbering scheme? Do these publications need a "1.0" version to consider something "Released"?

[Red Iron Crown]

Do these publications need a "1.0" version to consider something "Released"?

I truly believe that some do.

Version numbers need to die, IMO. Date of release, ascending alphabetical names, or even Donald Knuth's thumb to the eye of version numbers are better. People get way, way too hung up on numbers (I have heard someone say a while back that OSX is better than Windows because 10>7).

[Hektos]

neither 5th nor myself have been disputing that. The point was that squad hyped that up and juggled version numbers, and confused some people in the existing community, and set other such expectations through said hype, instead of just saying "this particular release, 0.27, can be considered a release game." if the labels "alpha" and "beta", and version numbers as well, were meaningless, why not just say "you've all been playing a beta-quality game and with this release we consider it a "release" product. Ksp is live." and keep the same numbering scheme? Do these publications need a "1.0" version to consider something "released"?

Fixed that for you

As for version 1.0, it's been fairly standard practice (in my experience) that the release versions start at version 1.0. Squad was more like than not just following that tradition.

[regex]

Version numbers need to die, IMO.

No, the expectation that version numbers mean anything other than "The available product has been incremented to a new version" needs to die. Version numbers have utility to the programmer.

As for version 1.0, it's been fairly standard practice (in my experience) that the release versions start at version 1.0. Squad was more like than not just following that tradition.

It varies entirely by who's doing the versioning.

[BNSF1995]

This game isn't in beta anymore. Beta was 0.90. 1.0 was the final product.

[Red Iron Crown]

No, the expectation that version numbers mean anything other than "The available product has been incremented to a new version" needs to die. Version numbers have utility to the programmer.

I would argue that those numbers should be used internally, keep it simple for end users. Though I suppose that might cause some difficulty on open source projects.

[r4pt0r]

This game isn't in beta anymore. Beta was 0.90. 1.0 was the final product.

Final product? That's a damn shame, I was hoping they could upgrade to unity 5....

In all seriousness, 1.1 or the update after it depending on what those updates have, might be worthy(in my eyes) of calling it 1.0.

mentally I still call the most recent version .28

[stibbons]

Oh good. This dead horse again.

[juanml82]

Regarding the whole beta/release, here's the deal:

If you sign up for a beta, you know you're paying to become a game tester. If a game is a release, you are a customer, and you expect a final product, mostly free of bugs. Optimizations, DLC, updates, all those thing can continue to happen. But you don't get stuff like "Can't take off in a spaceplane/Reentry is meanignless/Can't dip into the tip of the atmosphere coming at interplanetary speeds/maybe the next update will fix it" or "Cargo and service bays continue to be bugged", let alone the ridiculously high system requirement (an overclocked i7 processor) because the game is, to be gentle, unoptimized.

[peachoftree]

Just to clarify, Harvs processor is so that the game can build faster, not just so he can run the game. Build times will almost always be an issue with a project this size.

[Ted]

Oh sure, makes perfect sense.

I'm not one of those who was fooled by Squad's silly versioning scheme. As RIC points out above the game has changed significantly with the delivery method and everyone has their own ideas of what the terms are. The problem for Squad is that they believe their own hype about the software's development stage and therefore their users believe it too. Far better, in my mind, to just keep incrementing releases and call 0.27.x a "Major Release". Or whatever it takes for Steam to be happy. No one is confused, no hype necessary, continue delivering the product as usual, stuff happens.

And for me, it's even simpler. Out of one side of their mouth, Squad said "Version numbers, alpha/beta/release status, all that doesn't matter, really. You shouldn't even be thinking in those terms," and out the other side they're screaming "WOO HOO 1.0 WE'RE OUT OF BETA AND FULLY RELEASED WOO HOO!"

I'm going to entertain this train of thought, just while the cleaners scrape together the dead horse guts that's being so gleefully smeared everywhere after yet another dead horse beatfest.

You're completely right that we didn't put much weight on the alpha/beta status as we never followed the development model that those terms apply to. We didn't do it out of some sort of PR move that was aimed at creating fog between the player and the development process. We follow an iterative and incremental development methodology with each update to KSP and the overall methodology for KSP borrows from Agile, but not by much. We've always done that and it's something I've discussed before.

I urge you to read up on iterative/incremental and you'll find that there are few, if any, places where Alpha and Beta are the canon terms to use to internally discuss a product's integrity with that methodology. So where do those words come from and why did we use them? Because they're largely used to inform consumers (and at times non-developers) about the integrity of a product from the developer's point-of-view. In other words, they're largely a marketing term now. While they are rooted in software development and still widely used there today, in games they're usually just a label for "We're not close to completing the vision we had, still laying foundations" and "We're close to release and working in a manner that reflects that".

I hope that explains the reasoning behind the decisions we made. Furthermore, I hope that the transparency helps alleviate your anxieties about KSP and its development.

At the end of the day, you speak absolutely about what me and my coworkers believe about our own project. A project we endear with all the passion, energy and skill that we have in us. A project that we want you to all enjoy. A project that we will continue to work on and refine until we feel that it is done, in an order that we deem to be the most efficient with the resources we have. If the biggest issue that you have with that project is that the version number on it is different from the one you would put on it and the way you choose to discuss that is by making sweeping statements about what I think, then so be it. But do try to take another tone, that one is getting old and it doesn't make me feel as though I can discuss KSP as candidly as I once did.

... let alone the ridiculously high system requirement (an overclocked i7 processor) because the game is, to be gentle, unoptimized.

As peachoftree pointed out, that is to compile code and build assets, not to play the game. Development machines are rarely an accurate reflection of the system requirements to play it.

Edited October 2, 2015 by Ted

[juanml82]

Just to clarify, Harvs processor is so that the game can build faster, not just so he can run the game. Build times will almost always be an issue with a project this size.

As peachoftree pointed out, that is to compile code and build assets, not to play the game. Development machines are rarely an accurate reflection of the system requirements to play it.

But I am talking about playing the game. When Fine Print first came out, I decided to keep expanding a Minmus base/Kethane rig. Whenever I had a base contract for Minmus, I'd send a module to expand into an already existing base. Here's a screenshot.

Besides the gameplay issues (with docking over the surface, for instance), whenever I came within 2.4 km from the base, the game turned into a slideshow. And it did because the game was calculating physics for every single part despite the fact nothing was moving so every frame the physics calculation results were coming to the same results as the previous frames. It would be like Excel checking, and recalculating, every single cell in a spreadsheet even when there is no user input.

As a result, I no longer expand bases. Instead, I try to make them as minimalist as possible and I only design space stations for refueling, not for role playing value.

And I have an quad core i5 processor running above 3.4ghz, so it's not like my computer is underpowered.

I'm not a software developer, but I believe skipping physics calculations if there are no forces applied to a vessel is a rather simple optimization which, over the years, has never been introduced.

[Ted]

but I believe skipping physics calculations if there are no forces applied to a vessel is a rather simple optimization which, over the years, has never been introduced.

And there's one of the problems that one has to work around with an optimisation like this. How are you to skip "physics calculations" if you are calculating whether any forces are applied to the vessel to evaluate skipping them? Those themselves are the physics calculations you mention.

It's not an impossible process to optimise, there are just many factors and unexpected outcomes that one encounters that might not be immediately obvious which makes it significantly harder than you might think.

Edited October 2, 2015 by Ted

[juanml82]

And there's one of the problems that one has to work around with an optimisation like this. How are you to skip "physics calculations" if you are calculating whether any forces are applied to the vessel to evaluate skipping them? Those themselves are the physics calculations you mention.

It's not an impossible process to optimise, there are just many factors and unexpected outcomes that one encounters that might not be immediately obvious which makes it significantly harder than you might think.

What I think, and I'm not a developer, is something like this:

If the vessel is not flying in an atmosphere, does not have engines, wheels or rcs firing, or it's not falling (ie, in a suborbital trajectory), skip all physics calculations for that vessel. Basically, this means that a vessel (such as a space station or a fixed base) which is in a stable orbit or landed gets ignored when running physics calculation and you can approach to a base like the one I've posted without a lowering your fps, as physics would only be calculated for the rover/rocket you're using to approach that base.

[monstah]

I believe I brought the current dead horse for beating this time, and I apologise if it derailed the conversation that much, but my point was lost when all people could read in my post was "beta" and "released", so I'll reiterate.

Break the game as much as you want in order to make it better. We'll start over and have fun.

[regex]

I believe I brought the current dead horse for beating this time

It wasn't you, it was others who are hung up on those labels as if they mean anything in a world where an update is a few minutes of waiting away (if it's made available, which most companies do, of course) instead of involving a change in media and a possible trip to a store or post office (if an update is even made available, which most companies didn't). In that sort of world games can be in flux for a very long time, even past "release" (another label that might be rightly ignored in said world) as issues are addressed. I watch an EU IV series on Youtube and the guy is constantly talking about what got rebalanced by the current update and how that is affecting the course of his campaign.

E: That's why I beat the horse, at least...

Edited October 2, 2015 by regex

[Friedrich Nietzsche]

What I think, and I'm not a developer, is something like this:

If the vessel is not flying in an atmosphere, does not have engines, wheels or rcs firing, or it's not falling (ie, in a suborbital trajectory), skip all physics calculations for that vessel. Basically, this means that a vessel (such as a space station or a fixed base) which is in a stable orbit or landed gets ignored when running physics calculation and you can approach to a base like the one I've posted without a lowering your fps, as physics would only be calculated for the rover/rocket you're using to approach that base.

Silly thought. If you try to bimp your vessel without engines, rcs etc. with another, you will floath thru, like time warping through a collision.

[Mitchz95]

I always start a new game with each new version so I have no problem with stuff getting broken between updates, but I understand why others would.