Procedural Gen of systems question

[Wolf]

That model you describe for planet formation is now NOT the accepted model. With the confirmation of nearly two thousand exo planets (so far), and the fact that most systems are not like our own, there is now no accepted universal model for planet formation.

That reminds me of a very old interview with DBOBE, which of course I can't find at the moment, where he said the PG in FE2 produced some systems astronomers later found to be surprisingly close to real data.

It would be so great if the PG in ED, with all it's fine tuning to match what we know so far would turn out to be so accurate that it will be used in real science as a model for predicting systems and to create a new model for system formation.

 

[cpbtrading]

Thinking about this PG, if/when we have planet landing and so on. Should damage happen to a PG planet i.e. a building gets blown up or whatever, does the damage remain, or when you return is the building generated as good again. Or indeed do other players see it as damaged...:S

If the answer is no and the building will appear to everyone damaged, doesn't that mean they have to put this data in for all areas, i.e. a damaged version for all. But would that not be PG then.

Anyway if that were the case, it will take forever, if, as we know, the universe is to be soo large in ED here.
??

 

[Hellhawk666]

I was under the impression it was randomly generated, or near random, i.e. if you return to a place it may alter in the generation slightly.
...

Well, that's sort of understandable, because procedural generation does entail random generation to an extent, it's just that the randomness is pre-determined

Think of it like this. Imagine you rolled a dice and noted the number that came up. Then you roll again, and again etc. 400 billion times. That would be absurd to do, but it would give you a list of 400 billion random numbers. Then you assign a number to each star in a game to determine the type of star, and note the serial number of the star.

Now you have random access to your random number list, so it doesn't matter if no-one has ever been to star number 654,387 (which we'll say has random number "5"), the fact is that when anyone gets there the random number will come into play and be used to generate the type of star etc. Every time anyone comes there, it will always have the same random number because it was pre-assigned

Now, procedural generation is just a way to do what I've described above but by using a mathematical formula to simulate all the dice rolls. Computers do this sort of stuff very well, so suddenly the absurd scenario of all those dice rolls is practical ad happens in seconds.

That's all oversimplified, of course, and it involves much more, but basically that's why everything, even down to the last blade of grass, that is modelled in a pg game is always the same for every player.

No magic, just math

 

[Hellhawk666]

Thinking about this PG, if/when we have planet landing and so on. Should damage happen to a PG planet i.e. a building gets blown up or whatever, does the damage remain, or when you return is the building generated as good again. Or indeed do other players see it as damaged...:S
...

I imagine that alongside the algorithms and tables that hold all the PG data there are places for specifying persistent user-placed and/or influenced items and events that would over-ride the pg data. We know that some systems etc in the game will be hand-crafted to a certain degree, so there must be something like this happening.

PG is a simple concept, but I bet the implementation in this game is astonishingly complex.

 

[Robbie]

Well we don't have PG at the moment in ED, just 'skybox' to suit, in different cubes.
Hopefully it will come properly and we can see the back of this, rather gimmicky 'load coverup' called Hyperdive or jump or whatever.

I would like to fly forever without a break, i.e. one system to another in great speed but not broken to load the next section, or at least without a really noticeable break. Unlike the one we have now in ED here, terrible, obvious load level type thing they have going on...

To get that kind of seamless immersion you need to play SpaceEngine, if you haven't done so already.

 

[Harrimon Hoxx]

PG is a simple concept, but I bet the implementation in this game is astonishingly complex.

Very simple to demonstrate, amazingly complex in scope.

A while back I was attempting to explain PG to my wife who wasn't having any of it. "How can something that complex [Elite] be based on such minimal information?"

In response I knocked up a small Excel sheet that looked up a predetermined table of star attributes such as min & max luminosity, min & max diameter, for 6 star colour types. I then showed her how a massive amount of "new" stars could be generated by simply referencing that table with an algorithm.
In my case a simple random number generator.

Yes I know it's not strictly PG but it served to demonstrate that from a small set of data (my attribute table for 6 star types) I could generate and endless list of variants that conformed to the attributes expected for each star colour and I hadn't had to program each star individually.

I think she was convinced. Either that or she'd gone off to watch DBOBE's talk on PG.

 

[Rookie Ray]

This isn't the case. All star positions on the skybox are dynamically created from the actual galaxy data.

You do need to jump between systems to generate the new positions though.

Michael

Michael, will you ever answer my question about actual star names according to the flamsteed designation?

Like 23 Boötis being called 23 Boötis instead of Asselus Primus and all the other weird traditional names?

 

[Michael Brookes]

Michael, will you ever answer my question about actual star names according to the flamsteed designation?

Like 23 Boötis being called 23 Boötis instead of Asselus Primus and all the other weird traditional names?

I don't recall the question but we are using proper names for the stars except for low population or unexplored systems which use a variety of catalogue designations depending on the source catalogue.

Michael

 

[Rookie Ray]

I don't recall the question but we are using proper names for the stars except for low population or unexplored systems which use a variety of catalogue designations depending on the source catalogue.

Michael

Thanks for the answer.

Will there ever be an option to turn off the traditional names in favor of the scientific ones so that I would see 44 Boötis in place of I Boötis?

I know that Flamsteed and Bayer designations are based off of constellations and those constellations become warped the further from earth you move but I find it jarring that we would still call a star by it's old traditional name in 3000 AD.
It's even more jarring when some stars have traditional names and some have scientific names, it's a huge disconnect. Is there any plan to use one system for names or will it stay the way it currently is?

Thanks in advance.

 

[NeilF]

I don't recall the question but we are using proper names for the stars except for low population or unexplored systems which use a variety of catalogue designations depending on the source catalogue.

Michael

Maybe Michael could reply here with a list of all 400,000,000,000, so folks can go through them to verify?

Maybe split the list into two for easier digestion?

 

[Michael Brookes]

Thanks for the answer.

Will there ever be an option to turn off the traditional names in favor of the scientific ones so that I would see 44 Boötis in place of I Boötis?

I know that Flamsteed and Bayer designations are based off of constellations and those constellations become warped the further from earth you move but I find it jarring that we would still call a star by it's old traditional name in 3000 AD.
It's even more jarring when some stars have traditional names and some have scientific names, it's a huge disconnect. Is there any plan to use one system for names or will it stay the way it currently is?

Thanks in advance.

No as they are the names used by the population in the game. Also using Flamsteed is even more out of place as it only covers stars visible from Britain and doesn't include most of the stars in the database.

We may include a search function that allows you to search for stars by common catalog ids like Hipparcos and Gliese, but that's not confirmed yet.

Michael

 

[Skizomeuh]

Isn't it a good way to find at a glance systems that have been colonized, or not?

I don't see the point of displaying London as "51°30.5118′ N / 0°7.5444′ W"

 

[Michael Brookes]

Isn't it a good way to find at a glance systems that have been colonized, or not?

I don't see the point of displaying London as "51°30.5118′ N / 0°7.5444′ W"

That's the main reason they are named - because enough people live there.

Michael

 

[Rookie Ray]

No as they are the names used by the population in the game. Also using Flamsteed is even more out of place as it only covers stars visible from Britain and doesn't include most of the stars in the database.


Michael

Why did I ever bother learning Flamsteed designation.. damn education.

We may include a search function that allows you to search for stars by common catalog ids like Hipparcos and Gliese, but that's not confirmed yet.

Michael

THE HYPE IS REAL

 

[Rookie Ray]

Isn't it a good way to find at a glance systems that have been colonized, or not?

I don't see the point of displaying London as "51°30.5118′ N / 0°7.5444′ W"

The second I read this I got Michaels point, London is London because we have called it London.

Abstract locations/random places, we know their locations through GPS coordinates.

It makes sense.

 

[Skizomeuh]

We may include a search function that allows you to search for stars by common catalog ids like Hipparcos and Gliese, but that's not confirmed yet.

Would be a great feature, I understand the OP from this point of view: help pinpointing the relative position of a named system lost deep in the galaxy.

 

[Herne]

I know this is a slightly off topic flippant comment but does anyone else picture Michael Brookes standing at the stellar forge with a big hammer, beating it into shape whilst sparks fly everywhere, occasionally dousing ED in a bucket before returning it to the anvil?

I do now lol

 

[Mephane]

We may include a search function that allows you to search for stars by common catalog ids like Hipparcos and Gliese, but that's not confirmed yet.

Michael

This is a feature I've been sorely missing.:smilie:

 

[EidLeWeise]

No as they are the names used by the population in the game. Also using Flamsteed is even more out of place as it only covers stars visible from Britain and doesn't include most of the stars in the database.

We may include a search function that allows you to search for stars by common catalog ids like Hipparcos and Gliese, but that's not confirmed yet.

Michael

Please also give us 'fuzzy matching too'.
Or even a filtering as in LHS**** would give us a load of LHS systems, LHS 3*** would narrow it down again and so on...
Currently needing to know the exact name is 'awkward' especially for some of the more obscure systems

 

[Thunderchief]

Please also give us 'fuzzy matching too'.
Or even a filtering as in LHS**** would give us a load of LHS systems, LHS 3*** would narrow it down again and so on...
Currently needing to know the exact name is 'awkward' especially for some of the more obscure systems

Yes this please! The number of times I have gone into the galaxy map only to have forgotten the exact numbers or spelling of a system, is nearly the same as the number of stars in E!

or perhaps the ability to view your mission locations in the map.

cheers