Page 1 of 6

Problem Povray Planets

Posted: Fri Mar 02, 2012 1:44 am
by submersible
Please record any utterly wrong or incongruous planets here.
El Viejo wrote:
submersible wrote:
that the resulting texture is totally inappropriate for the description.
As with Ribilebi, you mean?
Image
Just like this one. Where possible , please quote the galaxy also.

Code: Select all

Galaxy 1
164,"Ribilebi"," (252","11)",6," Human Colonials"," Communist"," The planet Ribilebi is most famous for its vast oceans and its fabulous goat soup. ",

Re: Problem Povray Planets

Posted: Fri Mar 02, 2012 9:00 am
by Eric Walch
Probably it also will help when looking at the planetinfo.plist for the planet overrides. It contains all kind of systems were the description describes the planet. Internally the plist is used to modify the generated planets to match the descriptions.

For Ribilebi there is an entry but it is commented out. I don't know why.

EDIT:
Looking at the default generated planet texture for this system, it was already >95% oceans, (green oceans to be exactly) so there was no reason to make an override here. The entry was only a future reminder.

Re: Problem Povray Planets

Posted: Fri Mar 02, 2012 11:29 am
by submersible
Eric Walch wrote:
Probably it also will help when looking at the planetinfo.plist for the planet overrides. It contains all kind of systems were the description describes the planet. Internally the plist is used to modify the generated planets to match the descriptions.
I have been feeding the planet rendering macros using data from a spreadsheet I downloaded from the wiki. It includes information like description and tech level which I had planned to apply heuristics to for determining some planet features.

eg; m/(\w+) oceans/;
where the capture will contain 'vast' or 'pink' and can then be used to alter the default povray parameters.

Re: Problem Povray Planets

Posted: Sat Mar 03, 2012 10:29 pm
by Greyth
Galaxy 1,
Teaatis, Looks like gas giant or possibly a frozen planet but is an average agricultural.
Legees, Desert texture but is rich agricultural
Laeden, Desert texture but is average agricultural

Re: Problem Povray Planets - Heuristics

Posted: Sun Mar 04, 2012 12:19 am
by submersible
I have started building a list of key features which could be used to make a more informed choice of planet texture.

Criteria Description Keywords:
For the most part - presence of these imply the 'Continental/Blue Marble' planet type.
  • oceans . This is seen with a modifier pink/vast/unusual
  • mountains.
  • volcanoes
  • earthquakes
  • plantations
  • tropical
  • forests
Criteria Production Type:
I would say these contribute to the likelyhood of a planet being Rocky/Barren vs Continental. Also to impact the choice of colour maps - Industrial implies smog.
  • Industrial
  • Agricultural
  • Mainly Industrial
  • Mainly Agricultural
I will update this list as required. Still looking .
Eric Walch wrote:
Probably it also will help when looking at the planetinfo.plist for the planet overrides. It contains all kind of systems were the description describes the planet. Internally the plist is used to modify the generated planets to match the descriptions.
Those are a very useful starting point - thanks! I have also noticed that the spreadsheet I have been using does not detail the Industrial/Agricultural nature of the planets, only "Planet #","Name","X","Y","Tech Level","Inhabitants","Government","Description"

It is claimed that
If you press '!' on a long range chart, it'll dump the galaxy information to files in the same folder as Oolite. That'll give you a list of the planet numbers, and some other information about each planet for the given galaxy.

If anyone has managed to make this work or find the file being dumped - please let me know , or post an example of that data.

Re: Problem Povray Planets - Heuristics

Posted: Sun Mar 04, 2012 8:16 am
by Eric Walch
submersible wrote:
It is claimed that
If you press '!' on a long range chart, it'll dump the galaxy information to files in the same folder as Oolite. That'll give you a list of the planet numbers, and some other information about each planet for the given galaxy.

If anyone has managed to make this work or find the file being dumped - please let me know , or post an example of that data.
That function is removed after version 1.65. The info provided is the same as now is available on the wiki. I will remove that claim as it is no longer valid.

EDIT: Updated wiki + converted all XML example code to open step on that page.

Re: Problem Povray Planets - Heuristics

Posted: Sun Mar 04, 2012 9:59 am
by cim
submersible wrote:
I have started building a list of key features which could be used to make a more informed choice of planet texture.

Criteria Description Keywords:
For the most part - presence of these imply the 'Continental/Blue Marble' planet type.
  • oceans . This is seen with a modifier pink/vast/unusual
  • mountains.
  • volcanoes
  • earthquakes
  • plantations
  • tropical
  • forests
Criteria Production Type:
I would say these contribute to the likelyhood of a planet being Rocky/Barren vs Continental. Also to impact the choice of colour maps - Industrial implies smog.
  • Industrial
  • Agricultural
  • Mainly Industrial
  • Mainly Agricultural
I will update this list as required. Still looking .
I've been doing similar heuristic work for New Cargoes. In addition to the above, and possibly visible from orbit, I've been using:
{population} has adjective "Furry" AND {solar radius} < 4000km to indicate an icy world
and:
{population} has adjective "Fierce" AND {hub count} >= 10 to indicate a world with a harsh climate (high winds, acid rain, etc.)
and:
{hub count} >= 13 to indicate a system with a large asteroid belt (and therefore maybe more crater impacts on the planet below?)

Frog worlds probably have a greater chance of liquid water, though not necessarily oceanic.

"weird rock formations" is a rare keyword in planet descriptions.

"forests" may be "dense", "exuberant", "tropical" or "rain"

"night life" might well be visible from orbit, but would need shader support to implement.

The six major galactic sports (Zero-G cricket, Zero-G hockey, mud tennis, mud hockey, vacuum karate, vacuum cricket) might imply something about the planet: I would guess "gas giant", "swamp" and "airless rockball" respectively, in the absence of any other hints.

"civil war" could easily be visible from orbit (especially the "deadly" or "frequent" varieties)

Re: Problem Povray Planets

Posted: Sun Mar 04, 2012 10:29 am
by Mauiby de Fug
submersible wrote:
I have also noticed that the spreadsheet I have been using does not detail the Industrial/Agricultural nature of the planets, only "Planet #","Name","X","Y","Tech Level","Inhabitants","Government","Description"
There may be some useful things on the [wiki]Oolite planet list[/wiki] page. There are a few spreadsheets there, although looking at them, I think you're using the first, and the second, PhantorGorth's, while it has ind/agri data and most other stuff, doesn't have the planet descriptions.

But the individual galaxy pages linked from there contain plenty of information:
# 0. Tibedied (2,90), {19,71,75} within 7.0 LY. Radius 4610 km.

Feudal, Poor Ind. Pop. 3.6 B, Prod. 11520 MCr. HC: 3, TL: 9, Human Colonials.
This planet is most notable for Tibediedian Arma brandy but scourged by deadly edible grubs.

# 1. Qube (152,205), {40,74,159,187,194,240,242,251} within 7.0 LY. Radius 5528 km.

Corporate State, Rich Agri. Pop. 3.7 B, Prod. 16280 MCr. HC: 8, TL: 7, Human Colonials.
Qube is reasonably well known for its great dense forests but scourged by deadly civil war.

# 2. Leleer (77,243), {6,94,128,132,146,153,196} within 7.0 LY. Radius 3149 km.

Dictatorship, Mainly Ind. Pop. 3.5 B, Prod. 13720 MCr. HC: 7, TL: 8, Human Colonials.
The world Leleer is very noted for its pink Leleerian Er plant plantations but beset by frequent civil war.
It's not too difficult a job to copy this data and reformat it into something that can be used.
Alternatively, I wrote an oxp that wrote this sort of information to the log file when I was having a look at planet data myself. An example output:
22:35:48.250 []: 0, Tibedied, Poor Industrial, Feudal, 9, 3.6, Human Colonials, 11520, 4610, 3, This planet is most notable for Tibediedian Arma brandy but scourged by deadly edible grubs.
22:35:48.265 []: 1, Qube, Rich Agricultural, Corporate State, 7, 3.7, Human Colonials, 16280, 5528, 8, Qube is reasonably well known for its great dense forests but scourged by deadly civil war.
22:35:48.265 []: 2, Leleer, Mainly Industrial, Dictatorship, 8, 3.5, Human Colonials, 13720, 3149, 7, The world Leleer is very noted for its pink Leleerian Er plant plantations but beset by frequent civil war.
Again, with a bit of regex matching to remove the prefix, this data (and anything else that can be accessed/calculated from script) can be written into whichever format is necessary, potentially more quickly than processing the data on the wiki page, as I can determine how I want the output to be printed.

Interestingly, if you have a look at Oolite's default descriptions.plist, there are various potential natural features that just don't show up in game. For example, there are no "dust clouds" or "ice bergs" (which should be one word, not two, but as it doesn't actually appear in-game I'm not too fussed), and the oceans could theoretically be "funny", "weird", "strange", "peculiar", "ancient" and "great" as well. I'm also not aware of any "exuberant" forests".

Anyway, having had a very quick glance through the aforementioned file, I think you've missed "rock formations", which are in game, and should be added to your oceans, mountains etc list. Also, the mentions of night-life could well be visible from space. I can have a proper look through and do some searches to see which modifiers are actually used and on what at some point to get a more complete idea of what's mentioned in the descriptions could affect appearance if you like!

Also, I agree that the economy type should impact upon the texture. For example, I would have thought that a "Poor Agricultural" planet would be rather rocky and barren (as could, potentially, be some of the "unremarkable"/"boring"/"dull"/"tedious" worlds), whereas a "Rich Agricultural" one would be rather green. Similarly a world being "Industrial" would affect smog/cities, as could Tech Level.

I'm not sure how "solar activity" would affect a world, or how one could make a "revolting" world (indeed, it could well be the inhabitants which make the world revolting, rather than the actual planet), but they might be things to consider...

Edited to add: Ooh, I like your thinking, cim! I'm clearly going to need to have a look at the New Cargoes oxp. Does it include the food/drinks mentioned in the planet descriptions?

Re: Problem Povray Planets

Posted: Sun Mar 04, 2012 10:48 am
by cim
Mauiby de Fug wrote:
I'm also not aware of any "exuberant" forests".
They're very rare (which suggests they deserve special treatment, perhaps?).
gal3.list- # 64. Xesoon
gal3.list- #140. Esmaonbe
gal5.list- #104. Teusatve
gal8.list- # 64. Onveat
gal8.list- #166. Tile
Mauiby de Fug wrote:
Does it include the food/drinks mentioned in the planet descriptions?
Yes. They were among the easier cargoes to derive from planetinfo :)

Re: Problem Povray Planets - Heuristics

Posted: Sun Mar 04, 2012 11:41 am
by submersible
cim wrote:
"night life" might well be visible from orbit, but would need shader support to implement.
Given the current work I see happening on trunk to support auto-shader generation , I am hopeful that planets too will get some first class
shader lovin'.

I am experimenting with using diffuse+specular+emission+parallax (by spawning a spherical ship). Currently very broken due to texture co-ordinate problems and my generally feeble grasp of OXP writing.

Re: Problem Povray Planets

Posted: Sun Mar 04, 2012 12:10 pm
by Smivs
Just allowing emission_maps to be applied to planets would be a big step forward, and would be more universally useful.

Re: Problem Povray Planets

Posted: Sun Mar 04, 2012 2:27 pm
by Amaranth
WOuld it be possible to add reflectance maps too, to allow ocean reflectance?


White cloudy planets (look like Venus) are not too much of a problem as unless it is a desert there is no reason why a habitable planet would not be perpetually covered in clouds, for example any 'wet' planets or any with a description involving 'known for heavy rain'

Re: Problem Povray Planets

Posted: Mon Mar 05, 2012 2:20 am
by submersible
Smivs wrote:
Just allowing emission_maps to be applied to planets would be a big step forward, and would be more universally useful.
My first impression was certainly WOW - it looks really nice to see city lights on the dark side of a planet.
Amaranth wrote:
WOuld it be possible to add reflectance maps too, to allow ocean reflectance?
If you mean specular highlights - then yes and it does appear to give a good effect, bright shiny ocean and rivers while the land is matte.

The effect of parallax mapping on clouds was quite astounding , parallax mapping is my new favorite toy. :D

Re: Problem Povray Planets

Posted: Mon Mar 05, 2012 10:38 pm
by Greyth
Galaxy1, Ceinzala. Gas giant but is mainly agricultural.
Galaxy1, Arusqudi. Gas giant but is average agricultural.
Galaxy1, Rebia. Gas giant but is rich agricultural
Galaxy1, Riiser. Gas giant but is poor industrial
Galaxy1, Zarece. Gas giant, poor agricultural
Galaxy1, Rilace. Gas giant, rich agricultural (rich and fertile due to terraforming)

Re: Problem Povray Planets

Posted: Tue Mar 06, 2012 2:46 pm
by Cody
Gas giants as primary planets are a little problematical, I'm afraid. Great eye-candy, but hard to explain.