Jump to content

Devnote Wednesday: Tuesday Edition


SQUAD

Recommended Posts

2 minutes ago, Buster Charlie said:

Seriously how this is not stock?

 

Too many other things that are deemed higher priority.  It's not like they've got unlimited time to do everything they want right now, as much as we'd like them to have it.

 

Link to comment
Share on other sites

20 minutes ago, Eric S said:

 

Too many other things that are deemed higher priority.  It's not like they've got unlimited time to do everything they want right now, as much as we'd like them to have it.

 

Seriously? @Snark Already did it, and they wouldn't have to include the more advanced features of it, just the burn time. It's license doesn't even prohibit commercial use as far as I can tell.

 

So exactly how is this some unbelievable burden?  

Link to comment
Share on other sites

4 hours ago, maceemiller said:

For me this is the best Dev note yet. Loved reading every word and just made me smile. Felt good :)

I wouldn't yet say they have jumped anything yet, but ending an already vague dev notes with a poem did have me looking for a shark.

Edited by Sandworm
Link to comment
Share on other sites

14 minutes ago, Buster Charlie said:


@Snark

So exactly how is this some unbelievable burden?  

There is a vast ocean of ideas that fall between "We've got better stuff to do" and "unbelievable burden." 99 percent (give or take 99 percent) of all issues ever brought up fall into that ocean, including this one.

I tried to delete the beginning of the quote, and it deleted all the text except the tag. Sorry Snark if this tags you. If I wasn't forced to use a wysiwyg editor it'd not have happened.

Link to comment
Share on other sites

7 hours ago, Mitchz95 said:

Does that mean it's impossible to get KSP to work properly on my laptop? Can't I just get a new card?

Actually, there is a solution that might work for you. Had I not been planning on building a gaming rig anyway, I'd probably have tried this myself.

It's called an External Independent Video Card Dock. How it works is, you remove the bottom cover of your laptop, the one that lets you access the HD, RAM, etc, and unplug the internal wifi card. You then plug in the video card dock instead, and insert a dedicated graphics card of the type normally intended for a desktop machine. You will also need to plug a monitor into the external graphics card, though I believe certain models of laptops do allow you to connect the built-in LCD screen to the card.

Of course, this means your laptop has no wifi capability, but that can be overcome using a wifi dongle or even a cat5 network cable, if necessary.

It's a very cool solution to a long-standing problem.

http://www.ebay.com/itm/181943864835?rmvSB=true

 

Link to comment
Share on other sites

38 minutes ago, Buster Charlie said:

Seriously? @Snark Already did it, and they wouldn't have to include the more advanced features of it, just the burn time. It's license doesn't even prohibit commercial use as far as I can tell.

So exactly how is this some unbelievable burden?  

Actually, "just the burn time" is the most advanced feature in the BetterBurnTime mod you're referring to.  :) It just doesn't look like it 'coz there's not much to look at.

Not being privy to Squad's internal scheduling and deadline issues, I'm in no position to say what they "could" or "couldn't" easily do.  But there are few points I'm reasonably confident about.

First and foremost:  I seriously doubt that Squad would have any interest whatsoever in incorporating any actual code from my mod in the stock game.  I certainly hope they wouldn't.  I know that I sure wouldn't, if I worked there.  Because it would be a bad idea.  (And I say that, speaking as the author of the code in question, and I have a reasonably high regard for my own work.)

The reason I assert that it would be a bad idea is not due to licensing/legal issues, just that it's a mod.  Mod code isn't the same as stock code.  (Here's a recent post by RoverDude on the topic.)  Mod authors are limited to the public API. If you're working on the stock game, you have access to all the internals.  If I worked at Squad, I might have produced this same feature (I say might, I don't know what else they have on their plate)... but if I did, the code wouldn't have been written the same way.  Speaking as a software engineer:  it's a bad idea to grab some code from place A and just sort of cram it in place B.

Furthermore, my mod makes lots of assumptions that it can get away with because it's a mod and doesn't have to work for everyone under all conditions, including people running other mods.  Here's my recent rant on the subject, the tl;dr of which is basically "there are good reasons why Squad hasn't done what BetterBurnTime does."  I know the problem looks simple.  But it's harder and more complex than you think.

Maybe there are some small improvements that could be done without requiring a ton of work-- for example, persisting the state (i.e. memory of what acceleration you can do) beyond the current flight session?  But for anything much more than that, this isn't a "bug fix", it's a feature implementation, and not a small one.

The real takeaway, though, for anyone who's never worked as a software engineer on production code:  As a consumer of a piece of software, you really don't know how much work would be involved to "just do this one little thing."  It might be quick and easy.  Or it might be a lot harder than you realize.  Chances are very high that it's significantly harder than you think.  So by all means give Squad your feedback of "I would love for the game to do X instead of Y", that's useful for them to know-- but please don't be judgmental about it, or hector them with "why the dickens haven't you fixed this?", because unless you work there you really don't know what it is you're asking.

Edited by Snark
Link to comment
Share on other sites

2 minutes ago, Snark said:

I seriously doubt that Squad would have any interest whatsoever in incorporating any actual code from my mod in the stock game.  I certainly hope they wouldn't.

I felt exactly the same way about my bug fixes. They are hacky and workaround, built specifically to use and abuse stock code to make it fix itself. They are/were far from the correct solution in nearly all cases, and (quite frankly) sometimes very ugly to look at. But they were built specifically to not break people's saves by using existing stock code, vs. overriding everything and replacing it with my own.

Have some high-fives. :D

Link to comment
Share on other sites

39 minutes ago, Buster Charlie said:

Seriously? @Snark Already did it, and they wouldn't have to include the more advanced features of it, just the burn time. It's license doesn't even prohibit commercial use as far as I can tell.

 

So exactly how is this some unbelievable burden?  

SQUAD isn't going to just put someone's mod in. The way mods are made and the way code in the game is made are not the same thing. It has nothing to do with it being hard, they just haven't had the time to work on that particular thing. They're probably working on, say, making the game functional in the new unity version for the next release.  Despite licensing "appearing" legal to use in a product being sold, realities of the legal system and use are usually more complicated than that, and that alone would rule it out I'd say.

Link to comment
Share on other sites

15 hours ago, Arsonide said:

Data doesn't flow in partial amounts between the labs either. If the experiment data will not fit in the current lab, it will try the next. If it will not fit in any labs, it will tell you that now. The only functional difference here is that storing data in the lab does not produce a copy of said data. Once it goes to the lab, it disappears, so if you want to fill multiple labs you will need to get multiple sets of data. This was done for balance purposes. If the copy behavior remained, then you could take one surface sample, send copies to twenty labs, and still keep the sample when you were done. :confused: This was a bit overpowered to say the least, but it wasn't a problem prior to multiple labs being fixed.

Please reconsider this. The reason being, all it accomplishes is to make science even worse of a grind than already is. 

Here's what I mean. Right now, I collect a bunch of science during an interplanetary mission, store it in a command pod, then process it through a lab somewhere (maybe two or three, depending on the circumstances) to get maximum value out of it before I finally bring it back to Kerbin. A bit grindy, but not too bad.

With the system as you're describing, I'll have to collect all that same science repeatedly and carefully store each copy in a different part in order to process it through my labs. I can still do it, in fact the way I play the game I'll still have to do it, but it'll be a ridiculously slow, boring grind.

Realistically, it's reasonable to assume that the science collected by experiments either can be duplicated (e. g. temperature or gravity readings, which are just numbers, or crew and EVA reports, which are just words) or are collected in sufficient quantity to be split up between several labs (e. g. surface samples, which could as easily be several rocks as just one--certainly the Apollo landings brought back more than a single, indivisible rock per mission).

I get that you think the current mechanic is overpowered, and I don't entirely disagree. But please find a better way to nerf the science labs than just making the grind even worse!

Edited by Hotaru
Link to comment
Share on other sites

1 hour ago, Snark said:

but please don't be judgmental about it, or hector them with "why the dickens haven't you fixed this?"

With this in mind I do apologize for all the times I did this in all the devenote threads of the past!  Thank you Snark for the reality check.

Link to comment
Share on other sites

8 hours ago, Perry Apsis said:

I wonder if the docking change would fix the problems with Infernal Robotics's moving parts and docking ports. That would make me *very* happy.

Unfortunately, no. The problem IR has with docking ports is all about how the child docking port becomes the root of the child vessel when docked. If the docking port and root node are on the same side of all IR joints (even better: the docking port starts as the root node), then IR does not have a problem with docking. It's only when the hierarchy around the joint switches that the joint breaks because the joint messes with the hierarchy in its vicinity.

Link to comment
Share on other sites

At this point is pretty clear that SQUAD needs more people  and more experts on the team if they want to make things better.  The game without mods lacks on many many many areas and they don't even give a propper mod support. At some point the community will grow tired of this things

Link to comment
Share on other sites

8 hours ago, Hotaru said:

I get that you think the current mechanic is overpowered, and I don't entirely disagree. But please find a better way to nerf the science labs than just making the grind even worse!

Let me clarify: nothing is being nerfed. As of 1.0.5, if you have multiple labs on a single vessel, data will only be sent to a single lab anyway. Once a single experiment goes to that lab, it cannot accept another copy of that experiment. The new behavior is that it will recognize the first lab is processing that experiment, and automatically choose the next lab. In that regard it was buffed as you can now process a lot more at once, provided you have the right infrastructure. The data duplication exploit that was fixed is something that requires multiple labs running to experience, so it is not something that was present in the last patch to be nerfed.

Trust me, the lab is still quite a scientific powerhouse. :P

Edited by Arsonide
Link to comment
Share on other sites

@Arsonide

Hi,

I think the fact that it was mentioned that the data is moved, not copied, led to some misunderstanding (at least for me, and perhaps the question u answered to has the same source).

I misunderstood the "ADD DATA" button that is now present with a lab nd the fact that u MOVE it, that u actually move the science / the experiment out from its origin to the lab, so that the actual science points that u would get "raw" by sending it to kerbin are disappearing.

after some reading i now understood, that each experiment seems to have a "data" value for the science lab that u can use just once with 1.1, but u will keep the "raw" science andd can retrieve any experiment back on kerbin after storing the "DATA" in a lab before.

Is that right, or will an experiment vanish by sending its data to a lab (as i understood it before from the devnotes)?

Link to comment
Share on other sites

9 hours ago, Snark said:

Actually, "just the burn time" is the most advanced feature in the BetterBurnTime mod you're referring to.  :) It just doesn't look like it 'coz there's not much to look at.

I was not referring to the complexity of the code, I was just talking about people (squad) seem scared of making too much info stock. So the Landing prediction, and other other 'new' features would probably be right out the window. So really it was just me being too specific, I'm directly responding to the burn time comment.

Also I know how hard coding can be, I've dipped my toe in some super basic stuff and honestly I don't have the patience for it. It's not a comment on squads competence or their character, it was simply the concept of once one programmer has done this, you know it's possible; even if (by your admission) it's a bit of a hack. As a user, I find an inaccurate burn time more useful than NO burn time, because i've never found no time when your prediction was off so much as to cause trouble.

 

BTW thanks for that mod, it's super useful.

Edited by Buster Charlie
Link to comment
Share on other sites

8 hours ago, Hotaru said:

Please reconsider this. The reason being, all it accomplishes is to make science even worse of a grind than already is. 

*snip*

Not wanting to step on any toes, but have you considered whether your selected difficulty settings are right for you?

What feels "grindy" to someone will be different from person to person. But considering that the stock tech tree can be completely unlocked with exploring just Kerbin and the Mun (without using the MPL), and that there are at least two other sources of infinite science points in the game (next to the MPL, which is also one), I'm a bit surprised that you feel the need to massively grind science points with multiple MPLs during interplanetary trips. Perhaps you are running a modded tech tree, or perhaps you are playing on hard mode...?

Whichever it is: you sound like a prime candidate for the advice of "move that science yield slider a little higher when you start a game". It will reduce the amount of constant effort you need to put into science collection, ideally to the point where the required effort to progress feels "just right": neither trivial nor grindy. I'd recommend starting with +20% or thereabouts.

Edited by Streetwind
Link to comment
Share on other sites

9 hours ago, CloudlessEchoes said:

SQUAD isn't going to just put someone's mod in. The way mods are made and the way code in the game is made are not the same thing. It has nothing to do with it being hard, they just haven't had the time to work on that particular thing. They're probably working on, say, making the game functional in the new unity version for the next release.  Despite licensing "appearing" legal to use in a product being sold, realities of the legal system and use are usually more complicated than that, and that alone would rule it out I'd say.

To repeat myself, I wasn't advocating this, I'm just saying it's not impossible to add burn time predictions, even if it's a self confessed hack.

Link to comment
Share on other sites

4 hours ago, Arsonide said:

multiple labs on a single vessel

I'm actually talking about the effect this will have on using multiple labs on different vessels. What it sounds like you're saying as I understand it is that processing an experiment through a lab will now destroy the experiment, so if I want to return it to Kerbin or run it through another lab somewhere else I'll need another, separate copy of that experiment. If that's not the case, then I have no complaints!

 

3 hours ago, Streetwind said:

Not wanting to step on any toes, but have you considered whether your selected difficulty settings are right for you?.

Yes, I have. If anything my current career mode has science returns way too high. I finished the tech tree before even sending Kerbals out of the Kerbin system and now use science labs as my main source of revenue (I've been periodically editing my save file to reduce fund returns on contracts to keep things balanced). My concern is that, if I understand Arsonide's changes correctly, keeping those labs stocked with data will get a lot grindier in 1.1 even with a high science return setting. (Again, I might be misunderstanding, in which case, no worries.)

I still say the main balance issue in KSP career mode is the disparity between the early game and the late game. I've never been able to find a setting that makes the early game doable without a crazy grind, but also keeps the late game interesting without funds and/or science becoming virtually unlimited. The best I've come up with so far is tech-tree-extending mods for science, and periodic budget-cutting save edits for funds. My main career is actually still in 1.0.2, but I have done a couple of quick career runs in 1.0.5 (my last one got as far as a Kermanned mission to Duna) and the balance situation doesn't seem to have changed much: on 100% science and fund returns things were a little tough at the start, but I was pretty near to unlocking the whole tech tree and funds were practically unlimited after a single interplanetary mission even though I hadn't done much science-farming on the Mun or Minmus and hadn't even started running a lab yet.

 

PS.

Quote

at least two other sources of infinite science points in the game (next to the MPL, which is also one)

What's the second one? The only two I can think of are asteroid samples and the MPL, but it sounds like you mean two in addition to the MPL.

Edited by Hotaru
Link to comment
Share on other sites

10 minutes ago, Hotaru said:

What's the second one? The only two I can think of are asteroid samples and the MPL, but it sounds like you mean two in addition to the MPL.

Contracts and strategies. Basically the opposite thing you've been doing - converting as much of your reputation and funds income into science as you can afford to lose. Since there is an infinite amount of contracts available, the payout of all resources is infinite, and the vast majority can be turned into science without impacting your ability to continue to execute contracts forever.

Link to comment
Share on other sites

21 minutes ago, Streetwind said:

Contracts and strategies. Basically the opposite thing you've been doing - converting as much of your reputation and funds income into science as you can afford to lose. Since there is an infinite amount of contracts available, the payout of all resources is infinite, and the vast majority can be turned into science without impacting your ability to continue to execute contracts forever.

Right, didn't think of that one! I've gotten kind of into the habit of thinking of science in terms of converting to funds, forgot about the option to do it the other way round.

Link to comment
Share on other sites

4 hours ago, Arsonide said:

Let me clarify: nothing is being nerfed. As of 1.0.5, if you have multiple labs on a single vessel, data will only be sent to a single lab anyway. Once a single experiment goes to that lab, it cannot accept another copy of that experiment. The new behavior is that it will recognize the first lab is processing that experiment, and automatically choose the next lab. In that regard it was buffed as you can now process a lot more at once, provided you have the right infrastructure. The data duplication exploit that was fixed is something that requires multiple labs running to experience, so it is not something that was present in the last patch to be nerfed.

Trust me, the lab is still quite a scientific powerhouse. :P

 

45 minutes ago, Hotaru said:

I'm actually talking about the effect this will have on using multiple labs on different vessels. What it sounds like you're saying as I understand it is that processing an experiment through a lab will now destroy the experiment, so if I want to return it to Kerbin or run it through another lab somewhere else I'll need another, separate copy of that experiment. If that's not the case, then I have no complaints!

In the first iteration of the lab an experiment could be processed to buff its transmission value.
If in the upcoming iteration experiments that have already been processed could be marked as such, making them unable to be processed in another lab (without getting a fresh copy), but still be returned/transmitted back to Kerbin, it might be a good middleground?

Yes, we can build ships with labs and fly around the system, but sometimes all you want is to bring the experiments back home, upload them to a science station with one or more labs for further studies and return the experiment after processing for its original science value.
Since regular pods can only hold one instance of an experiment and stock KSP does not come with a handy science container, it is a bit difficult to bring enough copies of an experiment to do both.

Edited by KerbMav
Link to comment
Share on other sites

Well Kasper, good luck with your exams and the thesis writing stuff. And if you're reading this, then get back to work, and stop being distracted!

Ah, the update anticipation. The useless waiting, not starting a new save game, because what if it released tomorrow? Me, not anymore. Nowadays I've got things to do. Many things. Wife, kid, work, responsethingummies. Cowhammit...

Link to comment
Share on other sites

9 hours ago, sieve said:

At this point is pretty clear that SQUAD needs more people  and more experts on the team if they want to make things better.

Actually, they have to be careful. One of the axioms of software development management is "adding manpower to a late software project makes it later" (see Brooks' Mythical Man-Month)

That's because the old hands have to train the newbs, the newbs make stupid mistakes that the old hands have to take time to fix, and it's more time-consuming to communicate vital things to everyone.

I see Squad's been careful to hire mod devs that already know a good chunk of KSP development tools and practices, so there's not so much of a learning curve. They're also apparently careful not to make the team too large too.

I would not want to be in Dr. Turkey's shoes. He's got a difficult row to hoe.

Edited by GeneCash
Link to comment
Share on other sites

This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...