Re: [Space] GURPS Handbook of the Planets
 

"User" number is perhaps better called the "Universe" number. What it is is a seed for the random number generator the spreadsheet uses; on any (unmodified) copy of this spreadsheet, if someone uses the same User number and the same System number, they will get the same system, as it's using the same random numbers.

Re: [Space] GURPS Handbook of the Planets
 

Not a problem. I'd rather not-fix a not-bug than not fix a bug.

It doesn't.

Yes, that's right. The user number will remain "1" unless and until you choose another and insert it into cell C4 on the control panel.

To generate System 2, insert the number "2" into cell F4 on the System table, and so on for any other system number you care for.

What the spreadsheet does is hash the user number and the system number together to create a seed for the (rather crude) pseudo-random number generator. If you change either the user number or the system number you change all the die-rolls and get a different system. But if you ever put that combination of user number and system number in again you will get the same system as before, and if you tell them to me and I put them in I will re-create the same system that you get.

I coded it that way so that you can choose a user number to get a different set of systems than everyone else, and then record a system very compactly in your notes. For example, you might decide to use the user number 11032013 for a new setting, and record that in your notes. Then you can write the system numbers from 1 to as high as you like beside dots on you star-map and you can re-generate any of those systems exactly the same when-ever you want to refer to it.

I just generated the first 1,000 systems in "universe" 11032013, which took my machine about two minutes. I can look at the results, and see that systems 3, 22, 51, 62, 64, 72, 76, and 78 are highly habitable, and that 13, 18, and especially 73 are attractive for asteroid mining. I can record those facts, start writing up my setting notes, and generate the detailed results whenever I want to look at them. And I can give my players a copy of the workbook, and tell them "The setting is Universe 11032013. Persiphai is System 78, and that's the capital of the Alliance." Then they can put the universe number and system number into their copies of the spreadsheet and check out the details of Persiphai whenever they want. But it ia a lot more compact that printing out a system table and a planet sheet for each of the 67 highly-habitable systems in the first thousand systems of that universe. That's what makes this a handbook of planets (or rather, systems) and not a mere generator.

Not in the way I intended. Take a look at the tab called "Read me", and check out the instructions from Row 29 downward.


No, there are no back-connections or references to other documents. The workbook is entirely self-contained and will work fine on a computer that is not connected to any network. It's really just an insanely complicated spreadsheet.

No problem. I strive to please.

Re: [Space] GURPS Handbook of the Planets
 

The workbook I use to investigate a thousand systems at a time is called "GURPS HP beta21 explorer". It has a copy of the Handbook as its back end, and its front end is an Excel single-valued data table. The user can set any of the usual things in the Control Panel except the user number, but the usual System Table and Planet Display are disabled. You set the Universe Number on the front end, and tell it which chiliad of systems you want explored (chiliad number 1 is systems 1 – 1,000, chiliad 2 is systems 1,001 to 2,000, etc.). Then you select "calculate now" from Excel preferences, and over the next few minutes it fills in a table of the spectral class of the main star, the highest habitability and affinity of any object orbiting it, the number of habitable worlds, planets, and moons in the system (defined as having Habitability ≥4), the total carrying capacity of the system at your specified campaign TL, and the entire row from the System Table corresponding to the object that is closest to the main star among objects that have Habitability equal to the highest Habitability score of objects in the system and carrying capacity equal to the highest carrying capacity among objects with Habitability equal to that score. That is, in a system full of uninhabitable junk it sometimes produces the table line corresponding to the inner moonlets of the innermost gas giant that has moonlets.

It's quite finicky. My machine (which is ten years old and rather slow, admittedly) takes about two minutes to fill up the table. During that time I go and make a cup of coffee, because if I touch anything it will drop out and leave the table half-updated — and without producing an error message. I can't let it update in the background while I check e-mail or browse the web a bit. I ought to be able to, but I can't.

That being the case, leaving autocalculate switched on is totally impractical. You have to leave it switched to "automatic except for tables" and update manually only.

This isn't going to run at all on OpenOffice, nor (I suspect) only anything else but Excel. I wouldn't even have any confidence of it working properly on any version of Excel but Excel X for Mac, which I wrote and tested it on.

I haven't properly trapped all the errors. In the case of a system that has no planets or asteroid belts you get a row of #NA# results.

There are no specifications or instructions. There are definitely no user-serviceable parts inside. This is a cranky Rube Goldberg gadget, and it is not suitable to make available for general download and use.

Here's some sample output. I haven't tidied up all the formatting, but you ought to get the idea.

Re: [Space] GURPS Handbook of the Planets
 

I found a subtle but probably important error in beta 21. One column was referencing a 3d6 roll that had already been used for something else. That made a correlation between atmospheric mass and marginal atmosphere that ought not to have existed. I've mended the bug. You ought to download a new copy of the generator.

The latest version is "GURPS Handbook of the Planets beta22.xls".

Re: [Space] GURPS Handbook of the Planets
 

Thank you. It seems like it downloaded no problem, but it wants me to buy WinZip to open it because my evaluation period expired. I suppose I should just go ahead and buy it.

Re: [Space] GURPS Handbook of the Planets
 

I recently found this, and think it is a very well done system generator, and likely to be quite useful for anyone needing a randomly generated star system. I particularly like how it lays out the information.

Beyond the placement of moons, tidal calculations, tectonic and volcanic activity for young stars, and habitability modifications, were any other changes made to the generator as presented in GURPS Space?

Also, out of curiosity, what are you using to generate your random numbers? I took a look at how you did them, but wasn't able to figure out the method involved.

Re: [Space] GURPS Handbook of the Planets
 

Not only is it a random star-system generator, but also it has a planet-designing tool and will generate a random system around the the designed planet.

As far as I know there are no others aside from those documented on the "read me" tab. Unless you consider it a change that it calculates habitability on the basis of a world's final characteristics after tidal braking, not on the basis of what it would have been like if it hadn't tide-locked. As far as I know I implemented the GURPS Space 4th ed. star system generation sequence faithfully and in detail.

It's a linear conguential PRNG, adapted for producing reals in the range (0,1]. Or strictly its a small family of them, since I use a slightly different recurrence relationship generating the rows of seeds for the columns of PRNs that that which I use for the columns themselves.

The user provides a "personal user number" and a "system number", and these are hashed together to produce a seed for the pseudorandom number generator. Then there are two transcendental numbers, in cells D2 and E2 on the sheets that use random numbers. Each of those sheets has a row of seeds somewhere near the top. The first is generated from the PUN and system number, and each subsequent one is generated from the previous one by multiplying by one of the transcendentals, adding the other, and truncating. Then when a column of random numbers is needed the first is generated from the one in the row across the top, and each subsequent one from the one before it, by the same sort of method. Thus the numbers are uniformly distributed, but not truly random, and you can re-create the whole set by supplying the same seeds again.

As it happens I chose the two transcendentals rather poorly. The one in D2 is way too small, and as a result there are subtle periodicities down the columns. This doesn't affect the distribution of characteristics of worlds since their successive rolls are from different columns. But it would probably be ideal to use pi^4 or even pi^5 in place of pi in the D2 cells of the hidden sheets "star system" and "world details". That would eliminate subtle correlations that probably exist between successive orbital spacings.

Re: [Space] GURPS Handbook of the Planets
 

Sorry, I thought everyone had a free unzip facility these days. I'll put up a non-compressed version.

Re: [Space] GURPS Handbook of the Planets
 

Which I was aware. As far as I can tell, you've done a very good job of implementing the generator, though I haven't worked through all of the details to be sure of that.


Ah, I see how that works now. It's fairly simple, but despite the limitations you mentioned, I think it would take some careful statistical analysis of the results to turn up flaws in it. To the casual observer, it gives good results, which is good enough for this purpose.

The limitations of the GURPS Space generator are probably more significant in the long run. And while I'm quite interested in discussing them, would they take this thread off-topic too much? If so, I'll be happy to take that to PMs.

Re: [Space] GURPS Handbook of the Planets
 

While I have major problems in following these technical discussions, I am interested in reading them and maybe learning some small bits. So I would prefer it if you would continue here or open another thread which I can read.