Devnote Tuesday: QA focus

[Palaceviking]

On 8/31/2016 at 1:25 PM, SQUAD said:

Hello everyone!

 We’re very near the end of our QA testing period for update 1.2, and all developers have switched to polish and bugfixing tasks so that we can enter the next phase of testing soon™. We’ll be making a more expanded announcement when we get there, because there will be some important information to share!

I don't see how learning a new language so close to a major release date could possibly help?!?!     :~/

[0111narwhalz]

3 hours ago, Palaceviking said:

I don't see how learning a new language so close to a major release date could possibly help?!?!     :~/

I fail to see the connection between the quoted content and a new language.

[Andem]

3 minutes ago, 0111narwhalz said:

I fail to see the connection between the quoted content and a new language.

Quote

and all developers have switched to polish and bugfixing

But I still don't see any correlation @Palaceviking.

[FullMetalMachinist]

13 minutes ago, 0111narwhalz said:

I fail to see the connection between the quoted content and a new language.

8 minutes ago, Andem said:

But I still don't see any correlation @Palaceviking.

"Polish" the language versus "polish" the verb (to make something smooth or shiny). 

It's one of those double entendre things. 

[Andem]

Just now, FullMetalMachinist said:

"Polish" the language versus "polish" the verb (to make something smooth or shiny). 

It's one of those double entendre things. 

Dammit. 

[Pthigrivi]

7 hours ago, KerbMav said:

If one would rather get the equipment one wants to use to go beyond Mun or Duna sooner, the science gain slider in the difficulty settings comes handy.

This is an important misperception so I'm glad you brought it up. Grind isn't actually caused by a resource being too scarce. In fact most grind in games happens around resources that are all too common. It's not that science is hard to get, it's that getting it requires lots and lots of very repetitive tasks. Adjusting the settings doesn't change that. If you're sending a one-way probe, of course, just transmit. But if you're doing a crewed mission and don't feel like throwing away free science you're forced to log each experiment each time you pass into a new biome and then climb out to manually gather and store each peice of data individually. This is what is causing frustration. These repetitive tasks are the real cost we are paying to run reasonably efficient missions. This isn't caused by a lack of available science, it's caused by the basic drive to maximize return on investment. 

So the question I have is, with all that going on, what exactly is the value of repeating identical experiments as a game mechanic? What is it producing besides more clicking? Doesn't this seem like an easy opportunity to streamline things?

Edited September 3, 2016 by Pthigrivi

[Andem]

15 minutes ago, Pthigrivi said:

This is an important misperception so I'm glad you brought it up. Grind isn't actually caused by a resource being too scarce. In fact most grind in games happens around resources that are all too common. It's not that science is hard to get, it's that getting it requires lots and lots of very repetitive tasks. Adjusting the settings doesn't change that. If you're sending a one-way probe of course, just transmit. But if you're doing a crewed mission and don't feel like throwing away free science you're forced to log each experiment each time you pass into a new biome and then climb out to manually gather and store each peice of data individually. This is what people are complaining about. These repetitive tasks are the real cost we are paying to run reasonably efficient missions. This isn't caused by a lack of available science, it's caused by the basic drive to maximize return on investment. 

So the question I have is, with all that going on, what exactly is the value of repeating experiments as a game mechanic? What is it producing besides more clicking? Doesn't this seem like an easy opportunity to streamline things?

^^THIS! A THOUSAND TIMES THIS!

[tntristan12]

1 hour ago, FullMetalMachinist said:

"Polish" the language versus "polish" the verb (to make something smooth or shiny). 

It's one of those double entendre things. 

At first I was like -_-, and then I was like >_O, and now I'm like ^_^. Good work!

[taniwha]

8 hours ago, FullMetalMachinist said:

"Polish" the language versus "polish" the verb (to make something smooth or shiny). 

Actually, in this context, "polish" is a noun, but the act of polishing (or maybe the result of polishing). English is a bit of a mess with its words that vary their part of speech with the only indication being word position or even just context.

Edited September 3, 2016 by taniwha

[StahnAileron]

3 hours ago, taniwha said:

Actually, in this context, "polish" is a noun, but the act of polishing (or maybe the result of polishing). English is a bit of a mess with its words that vary their part of speech with the only indication being word position or even just context.

No worries. English is a horrible language to master because of all the weirdness and exceptions (I'm a native English-speaker), but Japanese (side-effect of a hobby) can be just as bad. Lots of homonyms and near-homonyms in speech while in text/print, you need to make sure you use the proper characters or else you risk saying completely different. (And that's assuming you use common words. Use an obscure one might send your reader/listener in puzzlement.)

It doesn't help that Japanese is usual used in a very context-sensitive manner. Subjects and even direct objects may very well be left out, being implied from prior dialogue. There's also the tendency in print to not alwaysspaceoutwords. (Classic example in English is the website for the company "Pen Island": remove the space and make it all the same case.)

Okay, way off-topic now... I'll shut up. (I'm pretty sure every language has its own problems in some form, though.)

Edited September 3, 2016 by StahnAileron

[Superfluous J]

10 hours ago, Pthigrivi said:

So the question I have is, with all that going on, what exactly is the value of repeating identical experiments as a game mechanic? What is it producing besides more clicking? Doesn't this seem like an easy opportunity to streamline things?

I ask a different question: Why do we have to click at all? I brought the Kerbals along. I threw in the cash for the extra fuel. Why can't THEY run the experiments themselves? And isn't there some underling on the ground who can tell the probe to do these experiments as well? I play this game to build and fly rocket ships, not hunt for 4-pixel-wide doodads stuck to my ship so I can right click them.

I won't play anymore without some sort of automated science gathering mod, ever since the time I realized I did not ONCE look at the planet I was flying by on close approach, so engrossed was I with "clicking all the things." And that was one click each (I consider that "maximum science") so no, reducing it to once won't help.

Edited September 3, 2016 by 5thHorseman

[StahnAileron]

It'd be nice if the science system just recorded data whenever new data could be taken (situation/biome transitions) once the craft is in a stable state. (Landed, level flight, fixed orbital orientation, etc. for X seconds.) Think Science Alert, but actually doing the data gathering for you.

If data can be transferred like a resource now, I wouldn't mind having "data storage" limitations so you have to engineer for recoveries, higher data storage, or transmissions. I also wouldn't mind the option to gather science over time from sensors in a scan mode. (The data would just accumulate for a while until you transmit it or shut down the sensors.) I know the game abstracts a lot for gameplay reasons, but I never liked how science was usually a one-shot instantaneous things versus real life where is takes some amount of time to acquire it.

(Then again, I hate the research system in KSP as well... I should be able to spend funds over time to research a topic, with sciences providing a boost. I still don't quite get how a dirt sample can directly help me develop a dual-mode rocket engine...)

I launch and set-up a probe into a nice orbit, but it's essentially useless in five minutes or less once all the science is collected. It's now debris I can either attempt to recover (if I bother to engineer that way), leave it to clutter my map view, or destroy it. (Well, I play with RemoteTech, so sometimes they end up as local relays, depending on what I need/could use.) It's fine for fly-by missions, but orbital missions feel short-changed on the experience.

Well, I'm the one harping on KSP being more Simulation that actual Game due to the lack of true gameplay mechanics that make sense and feedback into one another well and meaningfully. One day, I hope, KSP will be an ACTUAL game and not a sandbox sim masquerading as a "game". (Obviously I have lots of other ideas for my idealized version of KSP. Who doesn't here?)

[Pthigrivi]

4 hours ago, 5thHorseman said:

I ask a different question: Why do we have to click at all? I brought the Kerbals along. I threw in the cash for the extra fuel. Why can't THEY run the experiments themselves? And isn't there some underling on the ground who can tell the probe to do these experiments as well? I play this game to build and fly rocket ships, not hunt for 4-pixel-wide doodads stuck to my ship so I can right click them.

Yeah I mean this is the deeper issue, but I'm hoping this could be addressed in the future as part of a broad look at the science system in general. I was just pointing out this small part of the problem since they're looking at the transmission vs recovery balance right now.

Edited September 3, 2016 by Pthigrivi

[Rodhern]

6 hours ago, 5thHorseman said:

I ask a different question: Why do we have to click at all?

You mean "Kerbal the Movie", the epic eight hour motion picture?

[ my comment is a joke - in case you didn't guess already  ]

[Superfluous J]

Just now, Rodhern said:

You mean "Kerbal the Movie", the epic eight hour motion picture?

[ my comment is a joke - in case you didn't guess already  ]

Haha okay maybe "at all" is a bit much, but never clicking again in the flight screen to generate or move science around is a worthy development goal IMO

[razark]

6 hours ago, 5thHorseman said:

I ask a different question: Why do we have to click at all? I brought the Kerbals along. I threw in the cash for the extra fuel. Why can't THEY run the experiments themselves? And isn't there some underling on the ground who can tell the probe to do these experiments as well?

Wouldn't this go against Squad's "no automation" policy?

You might miss out on some of the fun like "click the thermometer again".  It's the same reason pilots can't execute maneuver nodes.  Who would want to miss the excitement of "trans-munar injection burn #69,105"?

[Pthigrivi]

Well I've written a number of really annoyingly long posts on this in the past haha but yeah, its a bit much. I do think players should still be involved in the act of experimentation it just shouldn't draw us away from the act of flying and navigating. Even just a "gather all data" button would do wonders. Your story about flying by a planet without even looking at it is a great example. If anything it should feed into it, driving us to execute our missions differently in order to reach the most valuable locations. 

Edited September 3, 2016 by Pthigrivi

[Curveball Anders]

5 hours ago, StahnAileron said:

It'd be nice if the science system just recorded data whenever new data could be taken (situation/biome transitions) once the craft is in a stable state. (Landed, level flight, fixed orbital orientation, etc. for X seconds.) Think Science Alert, but actually doing the data gathering for you.

Like this?

 

[Superfluous J]

24 minutes ago, razark said:

Wouldn't this go against Squad's "no automation" policy?

Sure, but so do the "aim at..." buttons when SAS is on. And fairings go against their "no procedural parts" rule. And several things, including transmitting science and drilling resources, go against their "no time-based mechanics" rule.

They're more guidelines than rules

[StahnAileron]

2 hours ago, Curveball Anders said:

Like this?

 

Why, yes. Though I still feel like that is treating the symptom instead of the cause. I believe the gameplay in KSP as it stands is fundamentally flawed (as a game, not a sandbox). "Science as a resource" for aerospace research seems kinda dumb in most cases. (Only atmospheric analysis makes some sense to me. Well, gravity logs as well in relation to rocket science.)

If KSP had procedural planets (via seeds, so you can share solar systems/planets), I think science as a discovery mechanic would be a bit better. Half the charm of space exploration in general is the discovery of new things. I personally would love it if my "science" endeavors in KSP did something like slowly filled out a system survey encyclopedia with the data I recorded. I know the stock science system doesn't really have it, but getting photos of planets as part of the science would be nice. (I know there was a mod that attempted this. I think maintenance was discontinued several times for different reasons... The last I think being an unreasonable user-base.)

Anyway, in short, I kinda want a complete overhaul of the current gameplay mechanics instead of trying to patch systems that don't make sense overall. Though at least some of the new features do work towards an overhaul still being feasible rather than an all-in investment into the current systems. (Though at the same time, you can't really do much to improve the current mechanics anyhow without an overhaul, IMHO. That's how poorly thought out there are.)

Hopefully with 1.2+, they fart out a diamond instead of trying to polish (...) a turd. (I'm not expecting TOO much out of 1.2 initially since I know it a code refactoring. Features and gameplay can wait while they un-screw the foundation.)

[Curveball Anders]

5 minutes ago, StahnAileron said:

I kinda want a complete overhaul of the current gameplay mechanics instead of trying to patch systems that don't make sense overall.

In other words, you want a different game.

[StrandedonEarth]

9 hours ago, 5thHorseman said:

Sure, but so do the "aim at..." buttons when SAS is on. And fairings go against their "no procedural parts" rule. And several things, including transmitting science and drilling resources, go against their "no time-based mechanics" rule.

They're more guidelines than rules

I think many of their previous rules and guidelines may have been Harvester's (Felipe's), or at least inspired by him, Now that he's moved on, old guidelines may be getting re-evaluated.

When rules don't work, they tend to get broken

[steve_v]

Another devnotes, another conspicuous non-mention of random CTDs.
Error in QA: double free or corruption (out)

[GorillaZilla]

On 03/09/2016 at 1:59 PM, StahnAileron said:

It'd be nice if the science system just recorded data whenever new data could be taken (situation/biome transitions) once the craft is in a stable state. (Landed, level flight, fixed orbital orientation, etc. for X seconds.) Think Science Alert, but actually doing the data gathering for you.

If data can be transferred like a resource now, I wouldn't mind having "data storage" limitations so you have to engineer for recoveries, higher data storage, or transmissions. I also wouldn't mind the option to gather science over time from sensors in a scan mode. (The data would just accumulate for a while until you transmit it or shut down the sensors.) I know the game abstracts a lot for gameplay reasons, but I never liked how science was usually a one-shot instantaneous things versus real life where is takes some amount of time to acquire it.

(Then again, I hate the research system in KSP as well... I should be able to spend funds over time to research a topic, with sciences providing a boost. I still don't quite get how a dirt sample can directly help me develop a dual-mode rocket engine...)

I launch and set-up a probe into a nice orbit, but it's essentially useless in five minutes or less once all the science is collected. It's now debris I can either attempt to recover (if I bother to engineer that way), leave it to clutter my map view, or destroy it. (Well, I play with RemoteTech, so sometimes they end up as local relays, depending on what I need/could use.) It's fine for fly-by missions, but orbital missions feel short-changed on the experience.

Well, I'm the one harping on KSP being more Simulation that actual Game due to the lack of true gameplay mechanics that make sense and feedback into one another well and meaningfully. One day, I hope, KSP will be an ACTUAL game and not a sandbox sim masquerading as a "game". (Obviously I have lots of other ideas for my idealized version of KSP. Who doesn't here?)

Absobloodylutely Brilliant,  This is the science/advancement system I want in the game :-)

 

I've hated the current science system from the start because of how nonsensical and arbitrary it is,  might has well have been:

"collect samples  of all the different coloured and flavoured jelly beans that are scattered throughout the solar system, ,  bring them home and funnel them into the magical jelly bean grinder that then poops out completely unrelated knowledge and blueprints of new technologies and parts,  and special new jelly bean scoops that can scoop up the special harder to scoop beans you couldn't scoop before......."

 

And its true for me too, in a way, the current Science system killed the game for me.    It went from having fun seeing what I could make from rocket parts,  and seeing where I could get to,  and what I could achieve and enjoying flying around moons and planets,  and generally enjoying the physicalness and visuals of spaceflight,  and challenge of it all.....     The things that made Kerbal Space Program great in the first place....

 

To not being able to just relax and enjoy the flights and sights,  due to being too distracted worrying about constantly checking science parts to see if I was orbiting over a new biome,  or just obsessively repeating experiments to make sure I caught the window when passing over a little biome crater or ravine....

 

And now I can't really go back to sandbox because the science system has instilled me with having to have a reason to go places,  and it's hard to shake off.    In fact I wish Science in it's current guise had never happened.   And instead they had used the effort to make the planets more alive and worth visiting just for explorations sake alone...

 

The current System is a completely Awful skinners box,  everything from the Pigeon in the box and on wards is a direct analogue to the science system in the game:   

 

 

The current science and technological advancement system as it is needs ripping out wholesale and rebuilt to be more in analogue with the real world,  and just happen and advance tech in the background as you play the game,  so you can relax and concentrate on playing the fun part of game,  which is ultimately Designing, building and flying (crashing)  rockets,     and the goal being to simply see what's out there,  By getting there :-)

 

The basics of the new system would be:

The people of the world want to know what is out there and how the universe works....   

So they pay and donate you cash in return for experiment data and accounts of first hand kerbal experiences on what you find out there,   you use that cash + time + feedback from real world flight testing    to further develop the technology you need to get out further and faster to find out more about that's out there,  which of course,  you and the world want to know....  

 

And as for Science Data and Transmission,  That stuff just needs to happen on it's own as you do whatever it is you feel like,  or have set out to achieve,   and you can check up on science instruments and experiments just for the flavour text,  or curiosity on what a sensor is reading,  Like "I wonder how cold it is on Minmus in the night?"   So you can just play the game and enjoy the experience of spaceflight while knowing that just by playing the game and pushing yourself ever further you are advancing yourself and your Kerbal people further technologically, culturally and philosophically,  without being skinner boxxed into micromanaging a clicker game to maximize results and not actually having any fun,  as 5th Horsemans Point gets across perfectly.

On 03/09/2016 at 0:44 PM, 5thHorseman said:

I ask a different question: Why do we have to click at all? I brought the Kerbals along. I threw in the cash for the extra fuel. Why can't THEY run the experiments themselves? And isn't there some underling on the ground who can tell the probe to do these experiments as well? I play this game to build and fly rocket ships, not hunt for 4-pixel-wide doodads stuck to my ship so I can right click them.

I won't play anymore without some sort of automated science gathering mod, ever since the time I realized I did not ONCE look at the planet I was flying by on close approach, so engrossed was I with "clicking all the things." And that was one click each (I consider that "maximum science") so no, reducing it to once won't help.

 

And for example If you want better rocket engines,   Flying with what you have at the start,  and running them in different conditions,  speeds and atmospheric compositions, pressures and temperatures while attempting to achieve your own self set goals of higher and faster....

Will build up flight data on rocket engine dynamics...  which your engineers and scientists can use alongside cash and time to shed light on ways to improve understanding of rocket engine technology which opens up ideas for new rocket engine designs,  which when flown themselves with their differences and application of new ideas shed more light on how rocketry works when run in different conditions...

And this should manage itself as much as possible,  and just notify you when you have enough flight data on the current rockets you have been using to begin research into understanding that data,  and designing new engines based on that understanding...  And you just have to "Pull the trigger"  pay the cash,  and then allow for the allotted time to pass to have your new rocket engines available for their first flight testing in the real world (With obvious danger warning symbol on their thumbnail indicating they they are potentially dangerous prototype)

They may explode first flight as they have been built purely on theory and yet to be real world tested X-D.....

But that's OK because such a failure will result in a big chunk of flight data advancement,  and future engines of the same type won't suffer the same failure again*  

*Just maybe another different failure later on under specific conditions preferably based on real life rocket engine failures to spice things up and add mystery,  and some fun flavour text and more Flight data advancement ;-)

 

 

So in short,  you get better.....  At whatever it is you are doing...    As it should be :-)

Which in turn means you can get further "Out there"  to get the thing you're really after,   the thing you're actually developing and building rockets for....    The Exploration of the unknown!    Unlocking the Secrets of the universe,  Seeing sights no one has seen before,  and bringing it all home for the world to enjoy,   and in return you get gifted with more cash and support to keep doing what you are doing :-)

 

 

 

 

 

 

 

Edited September 4, 2016 by GorillaZilla

[Superfluous J]

23 hours ago, StahnAileron said:

If data can be transferred like a resource now, I wouldn't mind having "data storage" limitations so you have to engineer for recoveries, higher data storage, or transmissions. I also wouldn't mind the option to gather science over time from sensors in a scan mode. (The data would just accumulate for a while until you transmit it or shut down the sensors.) I know the game abstracts a lot for gameplay reasons, but I never liked how science was usually a one-shot instantaneous things versus real life where is takes some amount of time to acquire it.

I'm not smart enough to do it, but I think all this could be done in a mod. Use the current numbers for science values per biome, add "science storage" in each pod and science equipment piece (and make it BIG in pods), and then have science accrue in those places (and transfer to pods automatically) over time, up to 100% for each biome/experiment/situation (perhaps with diminishing returns so the first 50% takes 10% of the time, etc).

You don't store "experiments" in your pod but raw science, and if you transmit science you get half its value back home, and "raw science" turns into "transmitted science" which is also stored on the craft. This science can be recovered for half value as well, so that surface sample you took and transmitted still exists on the craft. You transmit what you can, and what you can't transmit you can still bring home. You can also set antennas to automatically transmit all raw science all the time.

This would serve to do multiple things. It would simplify the collection of science, add a time-based mechanic (Oh noes! But it would also benefit getting into orbit and/or landing instead of doing a flyby or a dive into an atmosphere. FLYING in Jool's atmosphere would yield far more science than burning up in it for 5 seconds and then exploding), and abstract science while still keeping it present.

EVA reports and surface samples may need some love, but they could possibly still "just work" with the EVA'd Kerbal doing them automatically (and fairly instantly) when idle. Land on a planet, get out, he does the infamous "Surface Sample Animation" and bam, you just got 100% surface sample science for that biome. He pulls out a pad and scribbles on it, instantly you get an EVA report.

I'd use this mod.

Edited September 4, 2016 by 5thHorseman