RV Sonnenkreis - Decoding Universal Cartographics

[Esvandiary]

I don't want to mess up the online version with all that stuff especially when it's speculative.

That's what different worksheets are for!
Seriously though, I know what you mean... I'm a little nervous having all my working on Bitbucket, considering I know at least some of it is wrong. I had to throw an extra five prefixes down into the "low confidence" list this evening because I realised they were off and I wasn't sure about them... Perhaps more will follow.

Cheers for the updated list, very helpful!

Imagine the galaxy is a giant slice of Battenberg which for reasons beyond our ken has had small chunks of carrot cake pushed into it all over the place...

Absolutely beautiful analogy. Class 1 and 2 are now forever going to be carrot cake and Battenberg in my mind.

 

[Ziggy Stardust]

Imagine the galaxy is a giant slice of Battenberg

Can't do it. It just reverts to chocolate cake in my head straight away. And penguins. And now I've got the penguin eating the cake, and I know that's not right.

On a different note: Meerkat! \o/

 

[Macros Black]

As an (innocent) bystander, it looks to me like Jackie and Esvandiary is trying to reproduce the Stellar Forge algorithm, well, parts of it, based on data observations.

- - - - - Additional Content Posted / Auto Merge - - - - -

Trying to get to grips what you fellers are on about. Trying and failing spectacularly.

But it all sounds quite marvelous! Keep up the good work chaps!

And yeah, I do miss the meerkat.

PS: Also like the meerkat.

 

[Jackie Silver]

No great progress to report. I stopped off at a large (and largely untouched) nebula in PHEIA AUSCS to carry out another survey, and I found a few Water Giants (a couple purely by accident, while I was looking for a moon to park the ship.) Yesterday was Christmas chez nous (don't ask!) so I've been the worse for wear today.

 

[Esvandiary]

Today's first discovery: it appears prefixes are not the only things that use the run length limits...

I noticed that in row 95 of the Google Docs spreadsheet, we have "Graeh" that is a little way further along followed by "Groec". Seemed a bit weird since it's the same prefix, and yet "h" is nowhere near the end of the suffix list.
I took a look in the galaxy map, and "Graeh" is immediately followed by "Groeb" - i.e. it's jumped to the next infix.

On a hunch, I checked how long the run is for the "Ae" prefix, since I didn't think it was full-length. The answer? 12. How far into the suffix list is "h"? 12th position. Looks like infixes also use those run counts!
There aren't many infixes that also appear in the prefix list with non-full-length run counts, so it was quite a lucky spot really!

This does also explain how 35-length prefixes get "out of sync" (runs starting with suffixes halfway through the list) - they've previously gone through the non-full-length infixes at least once.

Still no closer to working out the discrepancy between the full-Z-slice length and how long it "should" be though. Clearly additional inspiration required.


Edit: OK, so I have a new theory about the continuation behaviour...
I think the behaviour we see (the later phonemes happening to continue when the prefix changes) is basically due to chance.
I believe each prefix keeps its own "counter" for how far through the list it's got. Since the majority of the prefixes are 35 long, this counter is in exactly the same place for all of them - thus, it appears to repeat.
That's also why it goes "odd" for non-full-length prefixes - its counter is further back because it hasn't had as many sectors, so it produces different names.

 

[Jackie Silver]

I think what we've got is a setup similar to the one for system names, but with a very peculiar base system - I don't know what the mathematical name is for a system like this (edit - a mashed radish?). I think what we get is a unique positional number for each sector (starting at bottom-left-down, right through to upper-right-top) but it's reached by adding up a whole set of different length digits, the phonemes. If we can get the lengths of all of them... or perhaps if we lump some of the harder-to-find ones into groups and just treat them as a single thing without needing the full details... I dunno. I'll keep plugging away at getting the lengths - looks like we've got a large chunk of it sorted. If I could be more certain of the SPL one we'd have even further.

I'm about 19 Kylies from the Bubble now - when I reach home, of if I lose my ship, I'll concentrate more on the mapping. I've marked a copy of the sector map with places I need to check. There's a lot of them!

(edit) And of course there's two different sets depending on whether the first phoneme begins with a vowel or consonant; I'm guessing that e.g. the EA of SYSTEASI is numerically identical to the PH of AIPHAIRSCH a little further along. Or possibly one lower, depending on whether the sequence ticked up in the gap.

(edit) It's a complete mess and somewhat behind the Google Docs version, but here's the map I've been working on for the sector boundaries, just for illustration.

 

[Esvandiary]

I think what we've got is a setup similar to the one for system names, but with a very peculiar base system - I don't know what the mathematical name is for a system like this. I think what we get is a unique positional number for each sector (starting at bottom-left-down, right through to upper-right-top) but it's reached by adding up a whole set of different length digits, the phonemes. If we can get the lengths of all of them... or perhaps if we lump some of the harder-to-find ones into groups and just treat them as a single thing without needing the full details... I dunno. I'll keep plugging away at getting the lengths - looks like we've got a large chunk of it sorted. If I could be more certain of the SPL one we'd have even further.

Yeah, I agree - lots of what we've seen has made me think "this is clearly just a highly obfuscated incrementing number". The Battenberg effect from class 2 especially, that made me think very strongly of flipping bits as you increment an integer.

Fly safe!

Edit: Quick query... How sure are you about Sh being full-length?
I have a sneaking suspicion that both Sc and Sh may be the same length and slightly short (somewhere around 33), but Syr, Cy and My may actually be full-length.
This is assuming my earlier theory about a per-prefix "counter" is correct.

 

[Jackie Silver]

Edit: Quick query... How sure are you about Sh being full-length?
I have a sneaking suspicion that both Sc and Sh may be the same length and slightly short (somewhere around 33), but Syr, Cy and My may actually be full-length.
This is assuming my earlier theory about a per-prefix "counter" is correct.

Urk. I was not certain, but fairly sure. I will check.

No, it looks almost definite - we know the exact boundary of SPLEEDEA - AGNAIHS, then following it along with 35s leads to DUMBAAE - SKOOHOU. Unless I've misplaced something, I'll cross-check that the map positions are right when I come to the end of this series of jumps.

 

[Esvandiary]

Urk. I was not certain, but fairly sure. I will check.

No, it looks almost definite - we know the exact boundary of SPLEEDEA - AGNAIHS, then following it along with 35s leads to DUMBAAE - SKOOHOU. Unless I've misplaced something, I'll cross-check that the map positions are right when I come to the end of this series of jumps.

I just found the same set and came to the same conclusion, so don't worry about it
I'll have a closer look at other examples (and other prefixes with similar behaviour to what I was seeing with Sh/Sc) and see what I find.

Edit: Yep, false alarm, just something worked a bit differently to what I was expecting... So in that case I think Cy, My, Sh and Sc are all 35, and Wr is 31.

 

[Macros Black]

This is probably a boring and perhaps pointless suggestion; have you considered asking FD for the algorithm?

 

[Jackie Silver]

This is probably a boring and perhaps pointless suggestion; have you considered asking FD for the algorithm?

I thought about it, yes. They've been forthcoming with e.g. the shield calculations before so there's every chance they would. Maybe if we get really stuck? It's fun working it out though a lot of it is coming down to leg-work.

(I'm kinda hoping I'll stumble onto something while trawling the galmap - every now and then I've bumped into a permit sector, though no new ones!)

 

[Esvandiary]

I thought about it, yes. They've been forthcoming with e.g. the shield calculations before so there's every chance they would. Maybe if we get really stuck? It's fun working it out though a lot of it is coming down to leg-work.

I agree... With that said, I am a little worried our current progress is akin to getting a better and better definition picture of the wall we haven't explained yet

Then again it wouldn't surprise me if, once we've got all the run lengths figured out, one of the total lengths involved turns out to be obviously significant. We'll just have to see!

Edit: Which reminds me... I think I can say I'm relatively confident about the numbers currently in my code up until somewhere around Myc - Gyr. The reason for this is that above that point, the run lengths also affect class 2 names, so if any of them are out, my test modes start shouting quite loudly. There's still a possibility of two of the "invisible" ones being off by the same amount in the opposite direction and cancelling each other out, but the overall run length up to that point should definitely be correct.
I think I have a few in that list in low confidence still... That's because I know they add up to a certain amount, but am not confident about the exact values for each within that total.

 

[Jackie Silver]

I made it home safely after some heavy jonking. Covered around 230 Kylies over 11,000 jumps. Phew! Much, much, much clicking later the data is safely sold and I can relax.

 

[Esvandiary]

I made it home safely after some heavy jonking. Covered around 230 Kylies over 11,000 jumps. Phew! Much, much, much clicking later the data is safely sold and I can relax.

Crumbs, that's over 40% of my total jumps so far, which I thought was a positively crazy number

Only about 30% of the total distance though... I hit 750 kylies travelled on the final jump to Beagle Point
That'll go up a bit tonight, as I head out to the most remote currently-reachable system, at 65627Ly from Sol... Assuming I can find someone to hold my limpets and hitchhikers so I'm light enough, that is!

 

[Jackie Silver]

No great progress over the last few days; I spent some time random-walking around the Bubble carrying out odd-job missions, smuggling Onionhead (to both sides, for giggles) and firing the odd shot here and there.
I have managed to fill out some more spaces on the map and there's a large amount of the sequence complete - mostly it's now a matter of handling the smaller sectors, which is not to say it's easy or nearly done.
I've uploaded the latest sequence list to the same link as before.

I'd replaced the Whitespace II with a freshly-minted Anaconda, the Paranoid Style, and after messing about with the loadout for ages I left for a shakedown trip out to Alnitak and Mintaka to check the data I had on them (they were just about the first places I went, and I didn't have an ADS fitted so I'd missed some planets at Alnitak.) I had never been able to get to the third of the belt stars, Alnilam (EPSILON ORIONIS in game) as it's much further out and in a sparsely star-populated area, so I headed down through Barnard's Loop and again found my way blocked by jumps longer than my range (about 32 ly, as the ship is still heavily fitted) so I've expended some FSD boosts to reach my destination.

On arrival I found that one of the worlds there is a 7+ g landable world - I know there's a 9 g world out there somewhere but this is the highest I've seen and certainly the highest I've landed on. A steady glide down at ~10 degrees was enough to make it without scratches though. The SRV feels like it's glued to the ground, unless I try to park it on a slope.

(I wasn't sure about the green camo paint, but it kinda works in this light.)
I'm thinking about heading off to the Formidine Rift some time in the near future, but it looks like I'm going to need to strip more fittings out of the 'Conda. Not sure about the main reactor but the sofa in the forward viewing lounge stays.

Incidental music: the Jam.

 

[Esvandiary]

Been having a bit more of a look at the one-word names this morning... I think there's a bit more to the continuation that we haven't quite got yet (I think I've inadvertently mentioned one example before).
I've tried some tests and just using the naive approach (continue using each infix/suffix while you can, then roll over) that takes way too long to get to the names we can see.

It looks like at some points, it'll jump infix/suffix out of nowhere; I think this is probably where the discrepancy is coming from. These "jumps" happen now and then, which is throwing everything off.

Examples:
Prieleau --> Prielee --> Priemoe --> ??? --> ??? --> Prungeou
Cleeque --> Cleeqai --> ??? --> Cloomoi

Further examples would be appreciated, I don't think I've spotted any others; basically any cases where the prefix continues as normal but everything else changes out of nowhere in the middle of a run.
This might be odd behaviour of limited-length infixes or somesuch... But basically, I'm not sure we have enough visible data to really find out other than trial and error.

 

[Jackie Silver]

Further examples would be appreciated, I don't think I've spotted any others; basically any cases where the prefix continues as normal but everything else changes out of nowhere in the middle of a run.

I will search for them - at the moment I've been concentrating on the run lengths, I hope to get all the runs from levels -2, -1, 0, 1 found so I've mostly skipped to the starts and ends of sequences. I'll start checking for infix changes now as well.

I've been a bit preoccupied the last few days, mostly wandering round the Bubble scanning a few hundred systems and being shot at - repeatedly and with gusto* - at the Community Goals. After much experimentation I found an Asp build with good jump range and that's decent to fly, so I'll probably head down to the Pleiades later to take a gander at the Barnacles and then get on with some more sector hunting.

*today's discerning psycho flies either a 'Lance or DBS fitted for stealth and bristling with railguns. The FAS is so last season, darlings.

(edited to add)

I arrived at Pleione earlier and took a look at them there Barnacles. Kinda smaller than I expected. Looks like it's growing a horribly be-tentacled starship to me.

(edited to add)

I got some more lengths - Sly is confirmed as 4 and from that we know that Dry and Sch are both 35, which gives a good chunk of sequence there. Can probably get more from that when I look at it further. It's nice to get the little ones!
Uploaded to same links as previous.

 

[Bravada Cadelanne]

There are a few sectors that are shifted on the spreadsheet map to where they are in the ED galaxy:

At row 82 stack -1 column 79, the spreadsheet map has:

FLYUE BYOE	AUPHAD	AUPHAVSKY	PRYUSKS	FLYA BYOEA	PRYUFF	PREWOE
Throsoa	PROO BLI	THROSEE	THROTOE	THROTIO	THROTOEA
	STRIECHIE	FLYA BLI	FLYOO BLAO

while the order in the galaxy map is:

FLYUE BYOE	AUPHAD	AUPHAVSKY	PRYUSKS	FLYA BYOEA	PRYUFF	PREWOE
Throsoa	THROSAAE	PROO BLI	THROSEE	THROTOE	THROTIO	THROTOEA
	STRIECHIE	FLYA BLI	FLYOO BLAO

At row 74 stack -1 column 83, the spreadsheet map has:

FLYOO PRAU	FLYA PRAU	DRYIGNAE	FLYAI PRAO	FLYUE PRAU	FLYAI PRAU
GOOROA	PRUE FLYAO	PRAI FLYAO	PRUE PLOE	PRAI PLOE
FLYA PLOE	BLEEROI	BLEERAA	BLEERUA

while the order in the galaxy map is:

FLYOO PRAU	FLYA PRAU	DRYIGNAE	FLYAI PRAO	FLYUE PRAU	FLYAI PRAU
GOOREA	GOOROA	PRUE FLYAO	PRAI FLYAO	PRUE PLOE	PRAI PLOE
FLYA PLOE	BLEEROI	BLEERAA	BLEERUA

I've also been looking at the hand-placed sectors using data from EDSM. The sector origin is placed at an arbitrary location where each of the X, Y and Z coordinates are an odd multiple of 5, and each level grid is shifted as necessary to line it up with the same level procgen grid while still having the larger AA-A x0 grid cube contain the smaller AA-A x0 grid cube. This seems to suggest that the hand-placed sector grid cubes alias and override procgen grid cubes (boxels?).

Now with that information, checking for out-of-place and invalid system names, and verifying in the galactic map what was at the location of those out-of-place and invalid systems, I have a list of 133 systems that need to be moved in EDSM. There are also at least 4 systems in EDSM where there seems to be nothing at the trilaterated location.

 

[Jackie Silver]

Thanks Bravada, I'll sort that out later.


I think I've made some major progress - it looks like the total stack is 128 sectors high, for a total of 128 * 128 = 16384 sectors in a slice, before it moves to the next line up the map.

I think the total length of the known sequences is 3037 sectors.

We can see six sets of sectors (6 * 128 = 768) which means that the unseen area is the total length 16384 minus 768, that is 15616 sectors.
15616 % 3037 is 431 sectors - which is the observed gap in the sequence between the end of one line (+2) and the start of the next (-3, one line up).

I'm still checking it over (it's a case of finding runs where we know for certain what it is at the start of -3 and end of +2) but I'm confident that we've got a 128 * 128 * 128 sector cube here.

 

[Esvandiary]

I'm still checking it over (it's a case of finding runs where we know for certain what it is at the start of -3 and end of +2) but I'm confident that we've got a 128 * 128 * 128 sector cube here.

Holy hell! That's bigger than I'd anticipated... That's even bigger than the internal coordinate system goes! (If the system coordinates MB provided back in beta are anything to go by, anyway)
I look forward to seeing the results once you've checked them, very interesting stuff!

I'm happy that you actually managed to find some numbers that even remotely make sense.