I made a missile for bombing runs...

Seafood.

i actually prefer organic shaped ships than the tec shapes. 

Helio,

You need to finish this, so that I can start using it, or so you can make models for me!

DANMAN

I just added a feature that changes it up a bit, letting you define limits in the shape of the mesh, at the time you press the button, sort of like the STORE button on a calculator... ... and it's the first step in making meshes with multiple stacked shapes, allowing you to create separate engine pods and the like...

  Here's a better pic of the whale base from above, compared to a similar design...

In contrast, here's a couple less organic ships showing how the mesh can be truncated...

Finally, in contrast to the original mesh I used to decypher the SINS mesh format...

Helioforge,

   I would like to playtest your mesh generator.  The ships your program tends to make are perfect for my new race.  I am making a variation on the 8472 bio-ships.  This is why I wanted the Borg in, but instead of 8472's, I will make it my own race that I named myself.

With the power of your generator, I can open "The FleetAlex Fleetyards", because without a fleet, I'm just Alex.

Helioforge,

if you put your focus on making the generator and I focused on making ships, releasing a ton of mesh-packs, then it gets tested and Soase gets it's ships more quickly.

   This seems to be most efficient way to go.

(resistance is futile)

Alex:  Since I've put considerable effort into programming it from scratch, I'll be claiming at least the first race's worth of ships and bases for the race I'm designing...

All the local militia work (and a hundred minor experiments in modding SINS) have given me the experience to assemble a race with original designs, powerful features, and careful balance, with strengths and weaknesses that change the nature of the game, without breaking it...

  Since my developer's tool was only started a month ago, I'll wait for more features to be added, to clean up the interface (currently it looks like a control panel from the space-shuttle), and to track down anomalies created by the interaction of those features...

  Oh, and a friend showed me the SPORE interface (it had a TREMENDOUS interface), but all you can make with it is SPORE creatures, so as perfect as the engine is that drives it, ... ...but I could go on for a week about the amazing user-friendly interface...

  If IronClad adapted the SPORE interface/engine to do what I'm attempting, SINS would never run out of ships!

Thanks Mansh00ter...

(I know a lot of people's Anti-Troll filters have been set to HIGH because we are all waiting for 1.1 proper, and waiting causes frustration, so the disclaimer is understood)...

Actually, I was considering rewriting the program from scratch, because one never thinks of the best way to do something the first time...

Right now, the program simply uses polar coordinates to wrap a sheet of points 41 across and 21 high.  In a rewrite, the sheet needs to be 3X2 all the way up to 60X30, with everything in-between (sort of how the backbone in SPORE works, but in both directrions)...  ...and all the features need to be rewritten to handle variable mesh sizes...

As for the UV stretching, I could simply write an algorithm that slides points around to fill in all those gaps that cause the stretching.   The other thing that causes the stretching is the way the sheet of points is wrapped into a sphere (I'm trying to figure out a reliable way to keep the points from concentrating at the poles of the object, as caused by the algorithm)

   I already have spare mesh-data-sets in place, which is the backbone of making the saved mesh contain several shapes at once, but I would still need to make the display show them at once, and to save them in one mesh file...

   Right now the program is at the point that the best thing to make it better is to redo it (a lot of the algorithms can simply be transferred, or rewritten quickly)... ...However, my interface and display are simply terrible (compared to anything 3D out there), and to do it right, it would need a display engine as powerful as SINS

   With only one mesh to display, the SINS engine would handle it with no problem, and the slow part would be all those iterative calculations needed to merge all the primitive shapes (something SPORE skipped over by making a ton of preset parts (with cutesy names))

   With just a simple addition to my code, my program could save meshes in algorithmic mode, instead of as a mesh, allowing for the algorithm to be imported back in (you could save and load your own parts, and make engines that could be swapped in on the fly)...  ...But if all the parts cost a thousand triangles, it gets big very fast, this is why I would need to make the algorithm handle variable sized meshes.  Little things like antenna could cost 12ish points instead of doubling the info for a minor add-on...  All the pieces are roughly in place, but for it all to work, it needs to be rewritten.

What I just wrote may look contrary to your suggestion, because of UV stretching and a declaration of variable-sized data-sets for meshes...

  The main reason my UV maps stretch is that the proportions of the triangles don't match the proportions of the texture map.  I would move the points around to fix those proportions rather than to move the reference points on the UV map (of course, I am in a unique position to do it this way, given the algorithmic nature of the program)...

  I'm not sure how I would move the points around to compensate, but I do remember math classes that talked about calculating the "curviness" of functions.  I would have to "unhook" the points from hard-coded positions in polar coordinates, and let them float on the surface. (it may take 60 years, and Teal'c may get a grey streak in his hair, but I think I could figure it out)...

Stop the techno-babble!

I'm a geek, not a nerd.

I went ahead and coded the program for variable data sets for each mesh component, but doing so breaks most of the data for the functions... The spheres still work (in all sizes), and the cones will be easy to fix, along with the cyllinders, after that, fixing each function will be progressively more difficult (so far I didn't have to rewrite the program...yet...)

Actually, just the cubes will be difficult, as most of the other functions rely on data relative to the size of the data-set, regardless of what numbers are entered (so it just has rough edges to smooth out, but until then, I'm getting some really wildly twisted meshes, maybe I should save one and show it)...