• Welcome to the new COTI server. We've moved the Citizens to a new server. Please let us know in the COTI Website issue forum if you find any problems.

New version of TravellerWorlds is out!

OjnoTheRed

SOC-13
Marquis
I've been working away on TravellerWorlds - http://members.ozemail.com.au/~jonoreita/TravellerWorlds/

The latest changes are in the Change Log, but the big news is:
1) You can scrub and add celestial objects to your star system to customise it to your hearts content.
2) You can edit the details of any star. Altering its luminosity will automatically update all temperature calculations.
3) Background changes with the (very gratefully received) assistance from David Wendelken which will prepare TravellerWorlds to edit maps. At present, if you've read the documentation, you can edit a downloaded JSON file with an editor to add or subtract terrain codes to edit your maps.
 
1) You can scrub and add celestial objects to your star system to customise it to your hearts content.
2) You can edit the details of any star. Altering its luminosity will automatically update all temperature calculations.
Does this mean that things like "moons of planets" (and especially, moons of gas giants) will become possible using the site?

This recently became a point of concern for the Boughene system (Regina/Spinward Marches), where TravellerWorlds wants to put everything onto a mainworld rather than a space station complex up in orbit (around a moon orbiting a gas giant).

Then there's the Bowman system (District 268/Spinward Marches), where TravellerWorlds yields some decidedly odd results that might not align all that well with the CT Beltstrike boxed set info.

Being able to get that LBB6 extended star system "data dump" for locations would be exceptionally useful.
 
Does this mean that things like "moons of planets" (and especially, moons of gas giants) will become possible using the site?

This recently became a point of concern for the Boughene system (Regina/Spinward Marches), where TravellerWorlds wants to put everything onto a mainworld rather than a space station complex up in orbit (around a moon orbiting a gas giant).

Then there's the Bowman system (District 268/Spinward Marches), where TravellerWorlds yields some decidedly odd results that might not align all that well with the CT Beltstrike boxed set info.

Being able to get that LBB6 extended star system "data dump" for locations would be exceptionally useful.
The "mainworld" issue at Boughene isn't really a Travellerworlds issue -- it comes from a canonical decision to basically ignore the mainworld UWP.

As "rolled" (back in the Supp 3 days) it was a planet with a hostile atmosphere. Some writer decided that the planet was so uninhabitable that almost everyone there lived in a space station instead (and honestly, that's pretty sensible). The UWP should have been revised to reflect that, though.
 
The "mainworld" issue at Boughene isn't really a Travellerworlds issue -- it comes from a canonical decision to basically ignore the mainworld UWP.

As "rolled" (back in the Supp 3 days) it was a planet with a hostile atmosphere. Some writer decided that the planet was so uninhabitable that almost everyone there lived in a space station instead (and honestly, that's pretty sensible). The UWP should have been revised to reflect that, though.
Come to think of it, the TravellerWorlds result is RIGHT, but not quite.

In the Boughene system, the mainworld IS a satellite! The problem comes from TravellerWorlds not being told by Travellermap -- which also doesn't know it -- that the mainworld is Bougene Station (A000531-D) and not the planet Boughene (A8B3531-D, or really, A8B131-D). Therefore, it "thinks" Boughene needs to be in orbit around something bigger -- which means a gas giant.

I'm not sure you can get to a result through the worldgen process that places a defined planet (Boughene) into the necessary orbit for a defined mainworld (Boughene Station) to be in orbit around it. And at Size 0, the station would automatically just get plugged into whatever asteroid belt the process generated anyhow...

Is there a way to show population in orbit vs on surface in the UWP? I've always left that to notes/fluff in the planetary description.
That would solve the problem nicely, if it were a part of T5 worldgen.

The bit about the station being in orbit around one of the three moons of a gas giant is pretty wild and I love it, but I don't see a way to make random generation get you the space station.

It might be a nice-to-have thing to add notes in wiki entries that tell you what tweaks to the default TravellerWorlds results are needed to get the output to conform with the wiki entry. (Things like unchecking the "mainworld is a satellite" box, how much the greenhouse and albedo factors need to be adjusted, and so forth.) If it could be standardized -- either as a list of factor adjustments, or simply a list of values that require substitution -- this might be amenable to automation if the inputs were thoroughly sanitized.
 
Last edited:
The bit about the station being in orbit around one of the three moons of a gas giant is pretty wild and I love it
Same here, but there aren't any guarantees that Boughene Station needs to be in orbit around the moon Boughene. Instead, Boughene Station could be "sharing an orbit" at Lagrange-4 or 5 in a 60º separation from the moon Boughene in the same orbit. Close enough for access, not so close as to be in any danger of an orbital decay/collision in the event of a station keeping thrusters mishap.

Lagrangianpointsanimated.gif

L-4 and 5 are stable equilibrium points.
L-1, 2 and 3 are unstable and would require thrust control to maintain (not much, but above zero).
 
Great discussion! But I'll be boring and go back to address Spinward Flow's original question and take in some other points.

A limitation of TravellerWorlds (at present) is that it is entirely a random generator with no reference to canonical data. Navigate from Terra (1827 Solomani Rim) on Travellermap (https://travellermap.com/?p=14.289!-107!7) to TravellerWorlds and you'll get something that has only a passing resemblance to our beloved home star system. The way I figure it, if you've already got canonical data to work from, just use that, don't bother with TravellerWorlds.

But having said this, it is possible to regenerate the results if you don't like them, and force certain results that you do want.

To regenerate a map using a random seed, go to Map Actions > Regenerate Map (new random seed). The other action to regenerate using the current seed means you can alter user preferences or map-affecting statistics (such as UWP stats) and the map will be regenerated but resemble the current map broadly (distribution

You can also regenerate the system using System Actions > Regenerate System, again with either the current seed or a completely new seed. This will sometimes randomly place the main world as a satellite of a Gas Giant (or a Big World if there are no Gas Giants) if you flick through many seeds.

If you want to force a result such as the main world being a satellite, go to View > Your Preferences where you can set up various system and map generation preferences. If you save your preferences, TravellerWorlds will use your altered preferences whenever you use it, regardless of whether you have navigated from TravellerMap or generated your own details.

The new System Actions > Add a celestial object to the system option allows you to manually add satellites and so on.

One correspondent has suggested to me that because you can export the system as a JSON file, and you can edit (nearly!) every body in the system, that we have a library of edited JSON files with canonical data that are imported automatically when clicked on TravellerMap. It's possible, but the repository of canonical data might become an issue - each JSON file is around 100 - 200 kB and if the number of canonical systems becomes even moderate (50+), it's megabytes of storage. But I'm going to work on it.

One last note: I do test the code before I release it, but I find that the endless variety of both data on TravellerMap and randomly generated data means there are bugs I have not spotted. Please report them (either message me here on CoTI or email me - jonathan dot a dot sherlock at gmail).
 
Counter-proposal.

Some systems, like Regina/Regina/Spinward Marches, already have some LBB6 extended system generation info available on TravellerWiki. Wondering if it might be possible, in collaboration with the wiki, to "rig" TravellerWorlds to be able to import that information for display in TravellerWorlds format so as to reconcile the resources of TravellerMap, TravellerWiki and TravellerWorlds towards a unified cross-pollination of data?

Additionally, TravellerMap at least consistently presents System info (here is Regina's):
System: F7 V Dwarf BD Brown Dwarf M3 V Dwarf

3 Gas Giants — 0 Planetoid Belts — 4 Other Worlds

Naval Base Scout Base
There ought to be a way to "bias" TravellerWorlds to use that baseline info when formulating planetary systems (if it is present).

As for JSON canonical data files, that is a bit of a poser as to how to compress that info for storage. However, since the number of recorded systems with extended planetary orbits and UWP data is relatively sparse, you might be able to get away with that.

It also seems that you're in luck in that there's only 65 systems detailed on the TravellerWiki.
LINK
So more than 50 systems to code into JSON files for reference, but not that much more than 50. So very likely less than 2MB of storage space needed to get them all.

The hazard, of course, is if TravellerWorlds starts getting used to create additional randomized results that the continuity of the seed for those randomization of results will need to be stored. I know you're doing that to some extent already using the map hex location, but you might wind up needing a slightly more "robust" system in the future so as to be able to "reliably return to" (and of course, link to) a specific system result from either TravellerMap and/or TravellerWiki as cooperating handoffs for a relatively seamless browsing experience.
 
TravellerMap already passes the number of gas giants, belts, worlds and the stellar data to TravellerWorlds. TravellerWorlds uses this data to construct the system. If you see particular systems where this is not happening, please report them! it's a bug. Even with this general system data, total actual data transferred per world is around 100 bytes - negligible.

The problem with data from the Wiki is the need for rigid consistency. On the other hand, if the wiki had the capacity to host JSON files ...
 
The problem with data from the Wiki is the need for rigid consistency. On the other hand, if the wiki had the capacity to host JSON files ...
Set a flag in TravellerWorlds for "this world already has a canon star-system layout, refer to wiki"
Have the wiki entry contain a link that incorporates instructions to TravellerWorlds to make it generate a system that matches canon (that's the JSON stuff, right? I'm computer-illiterate).

The idea is to use Travellerworlds as an engine to render the pre-set parameters stored on the wiki page. (I think that might have been what you said, but I don't know enough to know that.)

Ideally, Travellerworlds would read the information directly from code in the wiki entry, but that'd be way too much work for everyone involved.
 
Last edited:
I'm happy to include a JSON blob for known systems and pass it along from TM to TW if it exists, either as query data or by passing a URL to fetch it from.

I'd probably just stick them on AWS S3 as publicly readable, with a "SECT 3240" filename scheme, just like I do for world images/maps today.
 
When interoperability is the name of the game, using commonly agreed upon data standards to make handoffs as seamless as possible is going to make everyone in the browsing space happy. (y)
 
I'm happy to include a JSON blob for known systems and pass it along from TM to TW if it exists, either as query data or by passing a URL to fetch it from.

I'd probably just stick them on AWS S3 as publicly readable, with a "SECT 3240" filename scheme, just like I do for world images/maps today.
Once again, I'm amazed at the responsiveness and creativity of the comminity of Traveller hobby-developers!

Consider providing instructions on how to generate the appropriate JSON blob, so non-skilled wiki editors can just plug stuff in.
(Note to self: need to figure out how to edit the wiki... which would require signing up with it to do so.)

The layout of a star system isn't a significant plot point in most canon works. Therefore, I'd bet the TW output is perfectly acceptable even for most worlds already described in canon (except where there are canonical world maps, for obvious reasons). And, since it's dead-on consistent, where it doesn't contradict canon it may as well be canon itself.

Weirdness kicks in when writers skew the interpretation of the UWP. Two examples are Boughene, as mentioned above (turning a mainworld of A8B3531-D into an A000531-D orbital station plus B8B3131-D stats for the planet it orbits), and Trexalon/District 268 (where a nominally shirtsleeve B561851-C world was turned into an inhospitable ice-ball that should be B541851-C (thin, tainted Atm), by declaring that much of the atmosphere froze out and is now a layer of oxygen-nitrogen snow instead).
 
I'm happy to include a JSON blob for known systems and pass it along from TM to TW if it exists, either as query data or by passing a URL to fetch it from.

I'd probably just stick them on AWS S3 as publicly readable, with a "SECT 3240" filename scheme, just like I do for world images/maps today.
Thanks! I think passing it as a URL to fetch it from would work best; URLs cap out at 2k officially, and the JSON data typically goes 140k. And it's going to get larger as I add in more features. I'll email you when things are further along.

Grav_Moped - the point about users being able to generate a JSON blob is an excellent one. My idea is to make TravellerWorlds user-friendly enough that anyone can edit all of the map data for every world and every single statistic of each world, and then save it as a JSON. At present, TravellerWorlds has deeper detail for some things than many canon systems, but obviously not others such as the interesting Boughane situation that's been discussed here. This is the nature of literally decades of science fiction world building by fans to create interesting adventure settings. It's a good problem to have.

My future to-do list is:
  • Edit world maps using a drop-and-drag interface
  • Asteroid / Planetoid belt (editable!) details including the Carbonaceous / Stone / Iron-Nickel zones in the style of the MegaTraveller WBH (Marc Miller has assured me I can legally build the software)
  • Gas Giant (editable!) details and maps
  • Generate & edit Beast Encounter tables - both for the whole world, but also generated for every local Terrain hex when it is generated.
  • Local weather and seismic events for every Terrain Hex when generated
  • Generate & edit sophonts
  • Export the lot to a Google Doc with a reasonably pleasing layout that users can then go and edit to their hearts' content. This allows a better way for users to consolidate data and free text (such as the interesting Boughane situation!).
  • Export system and world details to a Google Sheet with built-in formulas so users can analyse changes in an alternative format and muck around with the data.

This is all fantasy at the moment, but I've discovered that I'm knocking off my goals one by one. Map editing is on the way, a lot of necessary background changes have taken place. It's just I'm an amateur and it all takes time.
 
One thing that I would dearly love to see on TravellerWorlds is a better image layout scheme for the world map.

Take Regina/Regina/Spinward Marches for example.
I look at the image of the world map and ... it's just confusing to look at.
There is Stuff on top of Stuff on top of Stuff even when zoomed all the way out to see the entire world map.

Or take a look at Vilis/Vilis/Spinward Marches.
The symbols used for stuff just pile up on top of each other in confusing ways, making things hard to read.

To be specific here, I'm talking about the symbology used by the map legend. The presentation just makes the symbols all jumble up on top of each other when there are multiple symbols per hex. I'm thinking that's a problem. So what I'm talking about is more a matter of visual presentation/layout issues than anything else.

Ideally, it would be nice to have checkboxes on the map legend items such that details could be turned on or off. By being able to selectively view information (such as, where are the Resources) it becomes possible to de-clutter the map in order to have to map selectively tell you what you are wanting to know.

Alternatively, various map symbology elements could be better "arranged" within the confines of each hex (so, layout issue) which reduces the opportunity for those elements to overlap each other.



I mean, if we're doing wish list items, updating the presentation of the world maps would seem to be a Nice To Have™ feature upgrade.
Could even be something as simple as checkboxes and an Update Map button to impose the changes onto a new render of the world map (along with an automatic refresh of the web page). That way you could use the checkboxes to modify the page URL and use that to control the variations in map presentation upon page refresh.

Also, you might as well just make the Ice Cap terrain a white hex with no symbol for it. The symbol just clutters up the map hex presentation for little in the way of added value (especially at the north pole hex).



And something I just noticed ... how does Equus/Lanth/Spinward Marches manage to have a liquid water ocean when the surface temperature at the equator is -60º C during the day at the equator? Considering that Equus is supposed to have a significant Dolphin sophont population (10%, which is 20 million Dolphins🐬), if Equus is an ice ball due to freezing temperatures at the equator that tell Antarctica "hold my beer" ... I'm thinking all of those Dolphins would undoubtedly DROWN underwater due to a lack of access to the atmosphere through the "hard candy shell" of ice covering the entire surface of the planet preventing Dolphins from breaching the surface in order to breathe.

Simplest solution would be to update the orbit of Equus to be somewhere that results in a more temperate climate suitable for liquid water ocean on much of the world's surface.



Oh and I think I found a bug for you.
This way of looking at symbols taken from a post by Thomas Jones-Low found here.
The link provided does not work.
Probably an artifact of the move to new forum software.
Simplest fix would be to find the post in the new forum software and update the link URL.



Hmmm ... :unsure:
Upon further poking around, I do not see any links on page at TravellerWorlds to either TravellerMap nor to the TravellerWiki (let alone links that take you to a presentation of the world under examination on TravellerWorlds). Might be worth adding such a feature (although TravellerWiki can sometimes be tricky to link to when there are multiple worlds with identical names, such as Kinorb or Aramis, for example).



Paya/Aramis/Spinward Marches may be a Garden World ... but according to TravellerWorlds, only at night when you won't expire from heat exhaustion.

Incidentally, Paya is a Disaster World according to TravellerWiki.
Paya was settled by colonists from Rethe and Inthe in the Regina subsector, Paya grew quickly, ultimately reaching a population of some 12 million people before disaster struck in 1075.
Considering that the disaster killed more than 99% of the population, there probably ought to be an impact crater somewhere on the map (a big 'un). 💥
 
Last edited:
Turning symbols off and on is a great idea. Consider it added to the to-do list.

On temperature and the map. A limitation at present is that the map is generated using the T5 mapping rules but has no link at all to the other details that are generated. The way around this at present is for the user to edit other details (e.g. orbital distance, albedo, greenhouse effect) to bring temperature in line with the generated map. This isn't terribly satisfactory.

But I need to explain the assumption behind TravellerWorlds. It's targetted at Referees who need materials for an upcoming session. Users can take whatever they like from the details. I suspect people often go to it for the maps, an explanation of the UWP digits to get a picture, and ignore everything else. That's completely valid. TravellerWorlds is not an attempt to document the entire OTU, but rather give Referees a world-building tool to make their lives a little easier. Also, I just find delving into the rules and developing the code fun.

This is why it makes no reference to the wiku or other sources - if you want a wiki to look up canon about the OTU, we already have one. I'm still keen to develop TravellerWorlds for editable JSONs that are stored somewhere for canon worlds such as the Terran system. But TravellerWorlds is a tool based on randomness not documentation.

Always remember, a good idea for an adventure setting trumps a bunch of randomly generated stats every time. UWPs are there to spur the imagination, not constrain it.
 
There's a manually edited JSON file for Byret on page 6 of this thread: Imaah Chehesti link

There are some new features like borders, which could be used on balkanized worlds or, in Byret's case, for a Larianz Autonomous Region on the planet.

In addition, you can place features where you want them, such as towns, cities, noble estates, etc. It doesn't take much JSON knowledge to understand how to just add an existing terrain feature type.
 
"Upon further poking around, I do not see any links on page at TravellerWorlds to either TravellerMap".

Just on TravellerMap links, on TravellerWorlds go to Generate - New Star System and you'll see the link there. But it's easier just to navigate using TravellerMap directly.
 
Back
Top