|
A different list of suggestions
|
|
| LevArris | Date: Sunday, 22.04.2012, 08:42 | Message # 1 |
 Observer
Group: Users
Germany
Messages: 12
Status: Offline
| Hi, I am new to Space Engine and this forum. I used the software a couple of days already and first of all I'd like to send a lot of thanks and appreciations to the author. Space Engine is just amazing and exactly what I was looking for (after using Celestia for years). I guess a small donation is the smallest thing we can do for Vladimir! 
I saw the TODO list for the next versions and there are already some good ideas and bugfixes coming up. It's not my purpose to repeat them again here, so if I still do, it's maybe because I don't remember all of them. But in general, my sugestions are new and as far I can see, they are not in this forum already:
* Features *
1. A user "description" field in the object's .sc would be awesome, maybe around 512 chars. This should appear in the celestial body's info. Would make lots of RPG fans happy, I guess.
2. Possibility to extend/change specific procedural objects properties. In my vision, this could work by creating an .sc file having the procedural object name (RSXXXXXXXXX.sc). Not all properties should be able to be overridden for procedural objects, only the ones which are not conflicting with the celestial position and gravitational stability of the body.
3. Introducing atmosphere gases class (made from combinations of gases), showing them in the info and linking them to procedural generation of atmosphere colors. I guess it's a little bit of a struggle to find out all logical possibilities in the Internet, but hey: everything what NASA doesn't tell you, you can still find on planets lists given to us by good SF games like MassEffect ;). For instance: Nitrogen-O2-Oxygen, Methan-Argon, Hydrogen-Nitrogen, CO2-Nitrogen, Helium-Hydrogen etc. etc. I guess algorithms to create density and size of the atmospheres are already in place. Now they only need some extra conditions on what gases are possible depending on: gravitation, distance to the star, type of star, vulcanism of the body (additional property?). The combination should also be logical, so an 80% O2 atmosphere on a vulcanic planet is probably a bad result 
4. I am missing a way to influence the celestial body generation formulas via configuration. I do understand that this violates the vision of everyone having the same universe, but what speaks against introducing an universe_user_gen.sc file where users can influence their _own_ version of it? Personally, I would love to play around with some constants and probabilities affecting the generation formulas. They don't need to be very detailed, most users would be happy to change "simple" things like type of body probability per system, size of habitable zone, total mass of the system...
5. A feature to gradually define the export would be nice, for instance possibility to export the selected celestial body only, or the actual position data. I know about the navigation places, but I am missing a way to export the position data of the camera as celestial coordinates (RA, Decl, Dist).
6. Maybe it exists and I couldn't find it, but is there a way to search for the identifier of a procedural object, so one can select and navigate to it?
* Bugs / Issues *
1. I don't know if this is more like a bug, but in general I see lots of systems with small stars (orange, yellow dwarfs with masses around 0.8 sol) and LOTS of gas giants orbiting them, very often more than 4 and also frequently systems where a gas giant is the "last" planet orbiting the star near to the system's boundary. I didn't check the mass balance of those systems yet, but it just doesn't feel "right". Even if the balance is correct, giant planets orbiting so far away from the main star (on a stable orbit) are an exception, not the common.
2. It seems like the algorithm creates way more moons as "Terras with life" than normal "Terra" bodies oribiting the main star. While this is still scientifically possible when the gas giant orbits the star within the habitable zone, the probability should be IMHO lower. In general, it seems like something is currently broken with the "life" algorithm. I found lots of Terras (without life) orbiting in the habitable zone of stable and old enough stars, having a normal atmosphere and gravity. In the same system I found "Terra with life" orbiting a gas giant, but the air pressure was minimal (0.16 for instance) and the gravity also pretty low. How comes that this combination can be found so often?
I hope you find some time to read my thoughts. I love your software and the community around it is very promising!
Any comments and critics are appreciated!
Regards
Lev
|
| |
| |
| SpaceEngineer | Date: Sunday, 22.04.2012, 14:14 | Message # 2 |
 Author of Space Engine
Group: Administrators
Russian Federation
Messages: 4800
Status: Offline
| Quote (LevArris) 1. A user "description" field in the object's .sc would be awesome, maybe around 512 chars. This should appear in the celestial body's info. Would make lots of RPG fans happy, I guess.
Once the online game is started, players will have the ability to write thier own descriptions for every body discovered, and it will be saved on the server. But for manually-made worlds it may be useful at the current stage.
Quote (LevArris) 2. Possibility to extend/change specific procedural objects properties. In my vision, this could work by creating an .sc file having the procedural object name (RSXXXXXXXXX.sc). Not all properties should be able to be overridden for procedural objects, only the ones which are not conflicting with the celestial position and gravitational stability of the body.
This is possible, but hard. Once engine creates a new procedural object, it should start a search through .sc files to see if the object is described there. It will reduce generation speed a lot, even if all .sc files were preloaded and parsed at the start of the program.
Quote (LevArris) 3. Introducing atmosphere gases class (made from combinations of gases), showing them in the info and linking them to procedural generation of atmosphere colors.
This needs a modelling of a chemical composition of a planet, that is not implemented yet (but in the 'todo' list of course). Custom coloring of the atmosphere is not supported by the used atmosphere rendering algorithm - it needs a lot of calculations to generate the atmosphere look-up textures, so for now SE simply uses a set of precomputed atmosphere models and use one of them for every planet procedurally. I have an idea how to bypass this limitation, but it needs some tests.
Quote (LevArris) 4. I am missing a way to influence the celestial body generation formulas via configuration. I do understand that this violates the vision of everyone having the same universe, but what speaks against introducing an universe_user_gen.sc file where users can influence their _own_ version of it? Personally, I would love to play around with some constants and probabilities affecting the generation formulas. They don't need to be very detailed, most users would be happy to change "simple" things like type of body probability per system, size of habitable zone, total mass of the system...
Yes, this is done because in the online game all players should have the same universe. Even if you simply post some nice screenshot, and other players want to look at them ingame, then you should give them not just a location name, but your universe config file too. However, you can change many things in the universe.cfg - stars density in galaxies and clusters, sizes and luminosities of procedural clusters and nebulas, disable procedural making of catalog stars binary, and the most fun feature - probability of terra and oceania planet to have life.
Quote (LevArris) 5. A feature to gradually define the export would be nice, for instance possibility to export the selected celestial body only, or the actual position data. I know about the navigation places, but I am missing a way to export the position data of the camera as celestial coordinates (RA, Decl, Dist).
Exporting of the procedural system is already implemented - look for it in the File menu, and for exported script in the cache/export folder. BTW, the script file has a name RS-XXXX.sc, but placing it in the catalogs folder and universe.cfg file will not work. RA/Dec/Dist can be found in the object's info just by looking at the object from the Solar system in debug or edit mode (press * to change the mode).
Quote (LevArris) 6. Maybe it exists and I couldn't find it, but is there a way to search for the identifier of a procedural object, so one can select and navigate to it?
The object's ID is it's name - RS XXXX-XX-... It can be typed in the search dialog (F3), but it's not shown in the table below the edit field.
Quote (LevArris) 1. I don't know if this is more like a bug, but in general I see lots of systems with small stars (orange, yellow dwarfs with masses around 0.8 sol) and LOTS of gas giants orbiting them, very often more than 4 and also frequently systems where a gas giant is the "last" planet orbiting the star near to the system's boundary. I didn't check the mass balance of those systems yet, but it just doesn't feel "right". Even if the balance is correct, giant planets orbiting so far away from the main star (on a stable orbit) are an exception, not the common.
Science doesn't know if this is the exception or common, because we can discover only close to a star and massive planets. So our observations are not full and we cannot use those statistics for now. However, some corellations are detected - gas giants that close to their parent star is very rare for dwarf stars, and it is not taken into account in SE for now.
Quote (LevArris) 2. It seems like the algorithm creates way more moons as "Terras with life" than normal "Terra" bodies oribiting the main star. While this is still scientifically possible when the gas giant orbits the star within the habitable zone, the probability should be IMHO lower. In general, it seems like something is currently broken with the "life" algorithm. I found lots of Terras (without life) orbiting in the habitable zone of stable and old enough stars, having a normal atmosphere and gravity. In the same system I found "Terra with life" orbiting a gas giant, but the air pressure was minimal (0.16 for instance) and the gravity also pretty low. How comes that this combination can be found so often?
Yes, I've noticed that too. Also, I often find systems with 5-10 planets or moons with almost equal radius, and terras with atmosphere and water with mass lower than a moon's. The planetary system generator has a lot of such bugs - because it generates all randomly, it does not perform true modelling.
*
|
| |
| |
| LevArris | Date: Sunday, 22.04.2012, 16:41 | Message # 3 |
|
Observer
Group: Users
Germany
Messages: 12
Status: Offline
| Thank you for your detailed answer. I see that some features are indeed difficult to implement due to limitations of procedural generation.
I also realized today the search edit field works also with non-suggested items, but I never managed to get it working with RS objects for some reason. I'll try again. Besides - a quick and nice bugfix would be to allow Paste operations into that field. Currently, pasting a copied string via CTRL+V doesn't seem to work and I need to type the whole long name manually.
Quote (LevArris) Yes, I've noticed that too. Also, I often find systems with 5-10 planets or moons with almost equal radius, and terras with atmosphere and water with mass lower than a moon's. The planetary system generator has a lot of such bugs - because it generates all randomly, it does not perform true modelling.
True, the low-density atmosphere Terras are annoying, I forgot to specify them in my first posting :-)
I didn't know though that the generation is randomized so much, I thought there must be some sort of modelling behind the generation. I don't know enough about the procedural code used, is the logical workflow described anywhere?
|
| |
| |
ENG
New site
Germany 
Russian Federation