Ship design spreadsheet

[Selentine]

If anyone else has wondered how much tech you need to put Reserve Tanks on a colony ship or an Ion Cannon on a small, I've put together a spreadsheet. Everything works under Gnumeric and should work with proprietary software (but I'd be happy for someone to find out for sure).

Ship design spreadsheet

Selentine

[Sirian]

It loads OK in "proprietary software". Looks like it could stand a little reformatting in there, but I think all the data is intact. No time for me tonight to actually examine the data, mind you. (That might be a while). But thanks for providing this resource.

- Sirian

[Selentine]

Looks like it could stand a little reformatting in there, but I think all the data is intact.

I gather it could be better. If the important columns don't fit on the screen or text overruns the fields or such, I'd like to know. Part of the reason I did this was to teach myself Gnumeric, and the layout might be n00bish and user-hostile. On reflection, displaying Space might be preferable even though Space is non-integer and behaves strangely. For the simplest example, Nuke Engines produce 20 power. Adding lasers (25 power) will alternately require 1.2 and 1.3 engines. Then not every demiEngine is the same size because fractions truncate. It's a mess and displaying integer quasi-Space would be misleading, but still I could have made better use of 800 pixelwidths.

I was mostly concerned that everything recalculated automatically despite some weird indirection, that everything fit, and that the space calculations came out right. It works here. If it also works for non-Gnome people I could reorganize it and perhaps include Cost.

Selentine

[Selentine]

Ok, Zed tells me some of the formulas were returning errors in his software. I'm guessing it was the weird indirection I was using to make it easier to reorganize the worksheet later. So I just reorganized it sooner and cleaned up the formulas to make them more portable. The layout is more like the D screen now and should be more user-friendly (instead of tester-friendly). I also added cost (seems decimal but rounds strangely) and a floatingpoint estimate for (unevenly trunced integer) space.

New and possibly improved spreadsheet

Selentine

[Zed-F]

Well now we have broken formulas all over the place instead of just in C8-F18. There's only one broken value when I load the sheet but as soon as I start filling in stuff (especially what warp engine I'm using) formulae start breaking all over the place. Could be more of the indirects are broken, I don't know enough about spreadsheets to say.

[Selentine]

Well now we have broken formulas all over the place instead of just in C8-F18. There's only one broken value when I load the sheet but as soon as I start filling in stuff (especially what warp engine I'm using) formulae start breaking all over the place. Could be more of the indirects are broken, I don't know enough about spreadsheets to say.

*inscribes regular 17-gon*
*rattles abacus overhead*
charis, CHARIS, CHARIS

I didn't know much about spreadsheet software ten days ago but I know a fair amount now. Enough to loathe kspread, anyway. The XLS export from Gnumeric may be weird, so I exported from oocalc and uploaded again. kspread, among other horrors, can't handle more than one argument to floor() and can't handle offset() at all requiring abuse of the indirect(address(row(),expr)) construct. Gnumeric and oocalc are both happy with the latest file. As I keep editing it I've added more formatting and validation which makes things happier here but may blow up in MSware.

For reasons I do not understand, a global replace of = with = may make broken formulas magically work; it worked here when importing gnumeric .xls into oocalc.

Kspread also had a problem with the construct: if (x>y; indirect(address(x-y;z)); 1), which I used for miniaturization since the miniaturization rates are table-lookup rather than algorithmic. It is possible bogusware will attempt to lookup (x-y;z) and barf when (x-y) is nonpositive. The workaround is indirect(address(if(x>y; x-y; 99);z)) with 100%s in O-P99. I find this inelegant, though, and if the formulas don't start broken I don't think this is the bug. Replacing = with = may shake the program into accepting the formulas. Or you could try

Third time lucky? spreadsheet

Thanks for bearing with me. I hope this spreadsheet will be useful even for non-Linux users.

Selentine

[Zed-F]

I think the global search and replace '=' with '=' trick is what did it. I get similar-looking errors on the spreadsheet as the last one when I first open it, but after I do the '=' replace on your current linked spreadsheet, it seems to work correctly.

[wycca]

A bit of necromancy I know (sorry), but the site with the spreadsheet no longer works, does anyone have a copy? Alternatively, I'd take Saragon posting the one he uses!

Thanks in advance.