Pathing w/Recast + Detour - Potential issue - thoughts?

[Tanaris4]

All-

So I thought I had this working properly but I'm seeing some strange results when I try to load my meshes into RecastDemo.

This is tile 31_47 loaded in Azeroth (3x3 around 31_47), everything looks pretty ok:


But then if I load 31_47 within a 3x3 around 32_48, it looks quite strange (maybe my generation of 32_48 + the 3x3 tiles around it didn't extend far enough, so it stretches far):


And even more strange, here is if I load a 3x3 around 32_48 and try to load the nav mesh I created for each tile:


This doesn't look normal - does anyone have any ideas on why this is happening? It seems like the bottom left and top right tile (of the last image) don't even have nav meshes.

Here is what I'm generating them with:

#define m_cellSize (1.0f / 3.0f)
#define m_cellHeight 0.3f

#define m_agentHeight 1.652778f
#define m_agentRadius 0.35f

#define m_agentMaxClimb 1.0f
#define m_agentMaxSlope 50.0f

#define m_regionMinSize 8       // 20
#define m_regionMergeSize 20    // 40

#define m_edgeMaxLen 12.0f
#define m_edgeMaxError 1.0f

#define m_vertsPerPoly 6.0f
#define m_detailSampleDist 3.0f
#define m_detailSampleMaxError 1.0f

#define m_tileSize 1600.0f

#define m_keepInterResults false

#define m_monotonePartitioning false

And I do extend beyond the tile a bit by doing:

    // we want to go over the tile's edge a bit!
    m_cfg.bmin[0] -= m_cfg.borderSize*m_cfg.cs;
    m_cfg.bmin[2] -= m_cfg.borderSize*m_cfg.cs;
    m_cfg.bmax[0] += m_cfg.borderSize*m_cfg.cs;
    m_cfg.bmax[2] += m_cfg.borderSize*m_cfg.cs;

Note: I haven't tried to generate a path, I just assumed I did it wrong given recast is displaying it this way.

Thoughts?

Thanks in advance!
~ Tanaris