KSP Weekly: Refactoring Kerbal

[sal_vager]

6 hours ago, Redneck said:

And may I ask how do you/squad know this? "welcomed by the majority of other players" How is this conclusion reached? Because the majority is us and in that case already own KSP

The KSP forums are a small part of the KSP playerbase, there are many other communities of players, especially for other languages who currently play KSP in English or with the excellent language patches project.

[John FX]

On 06/03/2017 at 10:02 PM, regex said:

I think it would end up like how I imagine they'll do delta-V; it'll have none of the features that make Engineer or MechJeb great (local TWR indicators, time of burn, that sort of thing because "it's confusing"). They'd add all the controls to the right-click menu and forget to make them action groupable, and then some modder would have to come along to make sense of it all.

I mean, I suppose it would be stable like you're asking, but I also don't think it'd be any good.

It would not need to be any good, just stable.

Like you say, some modder would come along and make sense of it, the current problem is that any mod implementation of moving parts is so far from what is possible stock that they are unstable.

So even if they only implemented a single powered hinge, no other features, all controls on right click menu, no grouping, quite unusable. That would be enough for a modder to make a perfectly acceptable catalogue of powered parts capable of anything.

If the APi was capable of allowing them to do so of course, which AFAIK it is.

It does not matter that Squads implementation is lacking, what matters is that it is there. Then modders have a stable base for that function which they can use to make something proper.

[Guest]

KSP 0.23.5 : Asteroid Redirect Mission

KSP 0.24    : First Contract

KSP 0.25    : Economic Boom

KSP 0.90    : Beta Than Ever

KSP 1.0      : We Have Liftoff

KSP 1.1      : Turbo Charged

KSP 1.2      : Loud & Clear

KSP 1.3      : The Most Disappointing Update Of All Time

 

Edited March 10, 2017 by Guest

[Red Iron Crown]

I remember way back in the day, when Linux support first started to be worked on. Very few of the existing players were interested in Linux support, but I don't recall much gnashing of teeth about Squad working to bring the game to new players even if that meant they had fewer resources in that update cycle to devote to making improvements to the core game (improvements that were much more badly needed than anything these days). The release came and went, KSP grew into a new market, the community gained a bunch of new people, and work refocused on making the core game better for the next cycle.

For some reason I started thinking about this when reading the comments here.

[Superfluous J]

8 minutes ago, Red Iron Crown said:

I remember way back in the day, when Linux support first started to be worked on. Very few of the existing players were interested in Linux support, but I don't recall much gnashing of teeth about Squad working to bring the game to new players even if that meant they had fewer resources in that update cycle to devote to making improvements to the core game (improvements that were much more badly needed than anything these days). The release came and went, KSP grew into a new market, the community gained a bunch of new people, and work refocused on making the core game better for the next cycle.

For some reason I started thinking about this when reading the comments here.

Conclusion: it's the Linux people complaining about internationalization.

Amusingly I was thinking along similar lines just today. Not about Linux, but about how having more people able to play the game benefits everybody because that means more potential modders.

17 hours ago, John FX said:

[A stock hinge] would not need to be any good, just stable.

You are presuming that the problems modders are having would not also befall any stock attempt.

I posit that the reason we don't have stock hinges is because every stock attempt has had the same problems modders have had.

[StrandedonEarth]

40 minutes ago, Aperture Science said:

KSP 0.23.5 : Asteroid Redirect Mission

KSP 0.24    : First Contract

KSP 0.25    : Economic Boom

KSP 0.90    : Beta Than Ever

KSP 1.0      : We Have Liftoff

KSP 1.1      : Turbo Charged

KSP 1.2      : Loud & Clear

KSP 1.3      : The Most Disappointing Update Of All Time

 

KSP 1.3: Localized Loco

[FullMetalMachinist]

20 hours ago, cfds said:

KSP is not really a text heavy game

I disagree. Have you seen how much info is in the KSPedia? Then there's the part descriptions, science result descriptions, contract requirements, and tutorial instructions.

20 hours ago, cfds said:

the only point where texts are important are the arbitrary module requirements for satellites/space stations/bases in contract descriptions. And there the better solution is to open the contract list in the VAB and look when green markers appear for the current craft [this feature exists, right?] than hoping the localized text is any more clear than the English one.

And how would they know that they have selected a sat/station/base contract if they don't speak english? You're saying that people are better off selecting a contract at random, and then randomly throwing parts together in the VAB until it magically meets the requirements? Seriously? Also, you cannot see if the craft you're building in the VAB will give you a green check on the contract window while in the VAB. You have to go to the launchpad for that.

20 hours ago, cfds said:

[rather] than hoping the localized text is any more clear than the English one.

That's why they have recruited native speakers from those localized languages to help qc those languages. 

[Mako]

47 minutes ago, Red Iron Crown said:

I remember way back in the day, when Linux support first started to be worked on. Very few of the existing players were interested in Linux support, but I don't recall much gnashing of teeth about Squad working to bring the game to new players even if that meant they had fewer resources in that update cycle to devote to making improvements to the core game (improvements that were much more badly needed than anything these days). The release came and went, KSP grew into a new market, the community gained a bunch of new people, and work refocused on making the core game better for the next cycle.

For some reason I started thinking about this when reading the comments here.

I am not endorsing any comments (except my own of course), but I'm guessing the community was smaller and closer to the devs then. I think a lot of the frustration would be alleviated if the current community had a better understanding of Squad's intentions for KSP. Right now all we have is (paraphrased): we're going to keep working on it. That is such a nebulous commitment and it's coming from a team with whom we are now mostly unfamiliar. It remains to be seen what kind of work this team will do and what goals they have for KSP. I believe most of the apprehensions that people are expressing are directly related to the doubts that are born out of the currently unknown KSP plans and leadership.

Right now the community is being asked to trust Squad, and it seems that some people in the community do. But we're being asked to trust Squad by people who have less interaction with the community than the past team, who have been working on KSP for a relatively short amount of time compared to the past team, and who have not yet released any work on which the community can form opinions. The only work we can judge at this time is the Weeklies and the Contests, and I don't believe either of those are what brought most of the community together in the first place or are responsible for keeping it together now. I don't mean this as an attack of any kind; it is simply the truth.

I hope for good things to come, but right now I just have no way to assess the likelihood of that happening. Time will tell, of course.

Edited March 10, 2017 by Mako

[klgraham1013]

1 hour ago, Aperture Science said:

KSP 1.3      : The Most Disappointing Update Of All Time

KSP 1.3 : Tower of BABEEEEEL [guitar riff]

[Vanamonde]

Shall we assume that those of you disappointed with 1.3 have somehow acquired early copies upon which you are basing these pronouncements? 

[Mako]

1 hour ago, Vanamonde said:

Shall we assume that those of you disappointed with 1.3 have somehow acquired early copies upon which you are basing these pronouncements? 

If people choose to be disappointed then no amount or rhetoric is going to change that; it'll probably just exacerbate it. I get what you're saying, but if the upcoming update only includes the already announced features and nothing else it seems like that's enough for some people to choose to be disappointed.

However, your point is not invalid: we don't know exactly what we're getting until we get it. We really don't know much of of what the future holds (especially following this update) because we have too little information to go on. It's not a complaint; it's just the facts.

[Malah]

22 hours ago, sal_vager said:

The KSP forums are a small part of the KSP playerbase, there are many other communities of players, especially for other languages who currently play KSP in English or with the excellent language patches project.

Yes, or as the french community which needs a translation ... As we are to many, we need 3 forums.

• http://forum.kerbalspaceprogram.fr/index.php
• http://www.jeuxvideo.com/forums/0-28031-0-1-0-1-0-kerbal-space-program.htm
• http://canardpc.com/forums/threads/107811-Kerbal-Space-Program-1-2-2-Topic-V3-!

Please, can you respond, when the french translation is planned ?

[Kenny Kerman]

I think you would help modders by rouding return values from functions like speed: I see some mods ("smart parts" for example) measuring speeds around 10⁻³⁵ m/s never reaching zero... I guess 10⁻³ would be enough precision........

Edited March 10, 2017 by Kenny Kerman
precision (!)

[sal_vager]

2 hours ago, Malah said:

Please, can you respond, when the French translation is planned ?

Squad plans to release further language packs after this one, and will let us all know via the announcements and weeklies, I have no idea what languages will be next but I see no reason why Europe wouldn't be a focus

[John FX]

7 hours ago, 5thHorseman said:

You are presuming that the problems modders are having would not also befall any stock attempt.

I posit that the reason we don't have stock hinges is because every stock attempt has had the same problems modders have had.

No, I presume we would not need to have a full, well rounded, implementation of moving parts, a single stable powered hinge would be enough for modders to work with. I presume nothing about the problems of implementing a stock hinge. It may well be that moving parts are unstable due to Unity but Unity has been very much upgraded since that information was new. It might be time to revisit the rockface and have another chip at it.

EDIT :

45 minutes ago, Kenny Kerman said:

I think you would help modders by rounding return values from functions like speed: I see some mods ("smart parts" for example) measuring speeds around 10⁻³⁵ m/s never reaching zero... I guess 10⁻³ would be enough precision........

Someone will always want the extra precision. Can`t you limit the decimal places in your own code?

Edited March 10, 2017 by John FX

[Superfluous J]

31 minutes ago, John FX said:

No, I presume we would not need to have a full, well rounded, implementation of moving parts, a single stable powered hinge would be enough for modders to work with. I presume nothing about the problems of implementing a stock hinge.

Well we do agree that if it's doable then Squad should do it. I just think that if it was doable, Squad or a modder would have done it by now.

[Kenny Kerman]

Quote

Someone will always want the extra precision. Can`t you limit the decimal places in your own code?

Actually I'm not coding these parts, but physically 10⁻¹⁰ is already measurable only around 0°K more than 5 decimals is maybe useful for calculs, not for physical usability. The propagating inexactitude is a realistic (real, actually) effect, + needs less computing power. I've talked about this to the authors of "smart sparts" too. Their speedometer can't detect a zero speed (plus it seems that the speed is sometimes < 0). But well, great job, I just LOVE the mix of realism/surrealism you made, and I owe you some points of mana. I'm not complaining, just trying to help (to help me too... A speedometer able to measure a zero speed is quiet a reasonnable technology).

Edited March 10, 2017 by Kenny Kerman
Difficulties in english :) nothing urgent. Thx. Ok finished I won't edit more :) never speed to answer me, buddy

[John FX]

2 hours ago, 5thHorseman said:

Well we do agree that if it's doable then Squad should do it. I just think that if it was doable, Squad or a modder would have done it by now.

It may well be that it used to not be possible but might be now if the idea was revisited. Since the first attempts there has been Unity upgrades. It may be the KSP code needs revisiting and altering to allow a modder to be able to make a stable hinge that was not possible before.

Of course it still may well not be possible in which case all this is moot. Would be interesting to see though.

Robotics would be a game changer for KSP.

[Kenny Kerman]

I think it's reasonnable to have "cascading event-firers" allowing a vessel to behave automatically from drop to deploy. It's fully doable with the "smart parts" mod (if it detects 0m/s), I asked them about a landing-strut-weight-sensor A zero m/s able speedometer would be fine too... Anyway I have fun. Dropping telecom relays around the planet from a plane... Trying to make "fire&forget" things.

Edited March 10, 2017 by Kenny Kerman

[klgraham1013]

9 hours ago, Vanamonde said:

Shall we assume that those of you disappointed with 1.3 have somehow acquired early copies upon which you are basing these pronouncements? 

They're disappointed at the feature list currently presented.  I'm not really among them, but I can see their reasoning.

[Guest]

11 hours ago, Vanamonde said:

Shall we assume that those of you disappointed with 1.3 have somehow acquired early copies upon which you are basing these pronouncements? 

I am disappointed at the fact that this update is going to be basically the addition of a localization feature, instead of being something we've been wanting for a long while (such as the new gas giants, proper weather, multiplayer, rocket part revamp (yes I know it was shelved), new science experiments, life support, in-situ part building, upgradeable engines, tweakables with text input, TWR readings, Delta-V readings, G effects on Kerbals, precise fuel transfers, more interactive IVAs, etc).

We've been requesting these for a LONG time now, and seeing that none of these have been accepted in favor of localization, the sole "major" feature in the next update, makes me get disappointed.

Edited March 10, 2017 by Guest

[DoctorDavinci]

3 minutes ago, Aperture Science said:

I am disappointed at the fact that this update is going to be basically the addition of a localization feature, instead of being something we've been wanting for a long while (such as the new gas giants, proper weather, multiplayer, rocket part revamp (yes I know it was shelved), new science experiments, life support, in-situ part building, upgradeable engines, etc).

We've been requesting these for a LONG time now, and seeing that none of these have been accepted in favor of localization, the sole "major" feature in the next update. That makes me get disappointed.

There's a whole lotta Murphy's Oil Soap Law being spread around here (those who have a tendancy to complain will continue to complain regardless of what you do)

So after reading through this thread, I thought I would save people some time by providing a sort of cliff notes concerning the gist of the thread so far ........

-----------------------------------------------------------------

@SQUAD should stop doing what makes good business sense and only do what the vocal dissenters on this forum tell them to do, especially since these dissenters now own you due to them purchasing a product from your company.

Once you accepted money for something that you created, you have essentially made a life debt pact with each and every person who uses KSP (@Elon Musk, we know where you got your testing software ) and now you and all your employees are indentured to the internet as a whole

We, the internet, demand you do what we tell you to do or else we will fart flames and spew salty tears ... and if you do what we tell you then we will still fart flames and spew salty tears 'cause that's how we roll son 

--------------------------------------------------------------

End of note .............

[Guest]

40 minutes ago, DoctorDavinci said:

There's a whole lotta Murphy's Oil Soap Law being spread around here (those who have a tendancy to complain will continue to complain regardless of what you do)

So after reading through this thread, I thought I would save people some time by providing a sort of cliff notes concerning the gist of the thread so far ........

-----------------------------------------------------------------

@SQUAD should stop doing what makes good business sense and only do what the vocal dissenters on this forum tell them to do, especially since these dissenters now own you due to them purchasing a product from your company.

Once you accepted money for something that you created, you have essentially made a life debt pact with each and every person who uses KSP (@Elon Musk, we know where you got your testing software ) and now you and all your employees are indentured to the internet as a whole

We, the internet, demand you do what we tell you to do or else we will fart flames and spew salty tears ... and if you do what we tell you then we will still fart flames and spew salty tears 'cause that's how we roll son 

--------------------------------------------------------------

End of note .............

Oh wow, this has become the Kerbal Godwin Law: the longer a discussion about KSP updates go, the higher the chance of someone saying we demand SQUAD to completely stop being a company since we supposedly think we're the owners of SQUAD

 

Look, I'm disapproving of the 1.3 update having basically only localization as a new feature, not the fact that SQUAD is trying to appeal to wider audiences, since we've been requesting so many more features for so much longer and they've had no work done on.

Although it does kind of upset me how the dev notes have turned into a spambot feed as noted by someone in the first page, it just worsens things by having shoddy public relations.

[Joal ban Kluane]

Localisation is a good feature to have, both for Squad and users.

Only users with some understanding of English can play KSP at the moment, and from those, only users with some mastery of the English language will spend any time on this (English-speaking) forum. There is thus a bias against the need for localisation. However, localisation will open the game to many new users, of which quite a few might become modders (however, I doubt we will see the development of a significant non-English speaking modding community, but I could be wrong).

For Squad, there is a clear interest in a multi-lingual game: it will open a lot of new markets, of people that haven't bought KSP because they do not understand it. This, in return, should provide some dVs (err... funds) to Squad to do further development of actual new features in our interest (it's really like a second stage ). Given the lack of DLCs at the moment, the only way to get that sweet dV is by  expanding the user base.

I myself selfishly look forwards to a French localisation support, not for me (I can play in English just fine), but for my 8yo son, which wants to play, but finds it difficult at the moment (we use the language patch for his install, but it isn't quite the same). It may also help to spread the KSP virus to family and friends.

On a more constructive point, I would hope that the localisation support will also include tools for an easy localisation of mods, which would allow modders to add localised content in their mods should users provide them with the proper localisation strings.

[Snark]

1 hour ago, Aperture Science said:

I am disappointed at the fact that this update is going to be basically the addition of a localization feature, instead of being something we've been wanting for a long while (such as the new gas giants, proper weather, multiplayer, rocket part revamp (yes I know it was shelved), new science experiments, life support, in-situ part building, upgradeable engines, tweakables with text input, TWR readings, Delta-V readings, G effects on Kerbals, precise fuel transfers, more interactive IVAs, etc).

We've been requesting these for a LONG time now, and seeing that none of these have been accepted in favor of localization, the sole "major" feature in the next update, makes me get disappointed.

Well, sure.  I'm "disappointed", too, in the sense of "sad that I didn't get a shiny fun thing I would enjoy".

Just because Squad (or anyone else) doesn't give me the thing I'd like, though, doesn't necessarily mean that they're blameworthy for it, or even that they did anything wrong.

I'm of two minds here, since on the one hand I'm a KSP player (which means I want all the same shiny stuff that you do, and I'm just as sad as you are when I don't get what I want) ... but on the other hand, I've been a professional software engineer for over 20 years, at a lot of companies, and I've got a pretty good idea of what goes into the production of commercial software, and therefore I know that sometimes difficult decisions have to be made.

Disclaimer: To be clear, I don't work for Squad and therefore don't have any more actual knowledge of their internal workings than you do; I'm not privy to their decision-making process.  But I'll go ahead and speculate here (that's all this is, speculation), because I know the software industry well enough, from the inside, to make what I believe to be a pretty good guess about how any software company needs to operate.

Here's the thing:

Squad, like any software company, is first and foremost a business.  It has to be.  It has to make sound business decisions for sound business reasons, or else it'll lose money and have to close up shop, in which case everybody loses, including you and me.

So, if I understand the complaints here, it's people saying that they think "Squad should have done <all this other stuff> instead of localization", yes?

Well, okay... so.  I think it's safe to assume that Squad knows lots of users want these features.  But they went ahead and did localization anyway, instead.  So.  There must be a reason, right?  Because everyone has a reason for everything?  Well, if Squad knowingly did a thing that doesn't give lots of users what they want, and did localization instead, I can think of only two conceivable reasons:

1. They're blithering incompetents who don't know what's good for their own business.
2. They actually have a sound business reason for what they did, and it's the right move for them at this time.

Both of these are, in principle, possible, of course.  But in the absence of direct evidence otherwise, #2 seems like a more likely explanation to me than #1.  What evidence is there for #1 that I'm missing, here?

I'd think that even if I knew nothing whatsoever about the software industry.  But, given that I am a software industry professional... I have good reason to lend credence to #2.

There are two important things to understand, here.

Point #1:  Localization is a VERY big-ticket item in terms of development effort.  It's frickin' huge.  To take a large body of software that wasn't originally designed to be localizable, and then retrofit it to become localizable... holy smokin' mackerel, that's a big job.  Trust me on this one.  I've done it.  If you've never done it yourself, you have no clue just how huge a job this can be.  It's not "difficult" in the sense of being rocket science, but it's incredibly laborious and tedious and requires rewriting and refactoring the heck out of so much stuff.  So... if they've decided that for business reasons "we need to do localization", I'm not at all surprised that it could pretty much crowd out just about all other feature work over a major version release cycle.

The above point addresses this comment:

41 minutes ago, Aperture Science said:

Look, I'm disapproving of the 1.3 update having basically only localization as a new feature

 

Point #2:  Localization is also a very big-ticket item in terms of business impact.  In terms of "dollars of new revenue per hour of engineer time spent," it's huge.  When you restrict KSP (or any software product) to just one language-- even a language as widely-known as English-- you're excluding many, many people from being able to use the product.  Localizing the product into a new language opens up the market to a lot of new users who otherwise wouldn't have been paying customers.  That, in turn, generates cash flow that the company can use for all kinds of purposes... including new feature development, which is what we all want.  Money is the oxygen supply of a business.  Choke that off, and the body shuts down.

(To take just one example:  One of the new supported languages is Chinese.  Do you have any idea how big the Chinese market is?  It's gargantuan.  And an awful lot of those folks don't speak English at all and have little inclination to try something that's not in their own language, given how many other products there are that they can get in Chinese.  Any software company would have to be nuts to ignore the Chinese market.)

Both of the above points are pretty important.  #1 explains "here's why we're not getting much in the way of new features, because of localization."  And #2 explains "here's why localization may be the right call for Squad, and is likely more important than delivering new features."

Yes, I'm sad (like a lot of us) that I'm not seeing a long list of shiny new features at no additional cost to me.  And, since I'm a native English speaker, I don't directly benefit from the fact that they're localizing.

<Begin reasonable but non-authoritative speculation>

However, I'm glad to see that Squad's (presumably) keeping their eye on the ball, business-wise.  If it's good for the business, it generates revenue that Squad could use to develop stuff that I will benefit from.

So, localization ends up being an indirect win for me, even if it means I have to be patient for the shiny toys.

<End reasonable but non-authoritative speculation>

I'm not saying that I know that Squad is doing the correct, smart thing by doing localization now, or by releasing a 1.3 that's basically just localization.  I'm just saying this:  based on a couple of decades of doing this stuff for a living, that seems to me to be a very plausible explanation.

That said, of course, much of the angst that's being expressed in this thread isn't so much about the issue of localization and new features per se, but rather about Squad's communication about them (KSP Weekly, etc.)  I'll bow out of taking on that discussion, since PR is something I'm definitely not an expert at.