Closed GoogleCodeExporter closed 8 years ago
Yes the Detour format changed in R252
http://code.google.com/p/recastnavigation/source/detail?r=252
I try to change the format as little as possible, but sometimes it needs to be
done.
If it is too much trouble for you regen the data, you can undo R252 quite
easily and revert the version number. Check the diffs for DetourNavMesh.h/cpp
and DetourNavMeshBuilder.h/cpp
Original comment by memono...@gmail.com
on 23 Nov 2010 at 6:29
[deleted comment]
however i would to update the code to fix the bug 144, so what should i do....
Original comment by zhibin.l...@gmail.com
on 23 Nov 2010 at 9:25
Bug 144 touches only Recast, so it is safe to keep R250 and only apply the
changes in R253 over it.
I'm not very good using SVN, so I would probably just sync to R253, copy
RecastArea.cpp and RecastRegion.cpp aside, sync back to R250, and copy
aforementioned files over the existing ones.
Sorry for the trouble. Do you have some ideas how to handle file format changes
in the future? The fact is that it will change. I hope to change it as little
as possible, though.
Original comment by memono...@gmail.com
on 23 Nov 2010 at 9:35
if it is only change of format, the data itself is still valid, you can keep a
set of loaders of different version. for example, Load250(), Load253(), or
LoadVersion1(),LoadVersion2()...and so on. In the Load() function, use a
switch(version){} way to load differnt version.
However, if the data itself is not valid anymore....OK, we just rebuild it.
AnyWay, thanks again for your hardwork
Original comment by zhibin.l...@gmail.com
on 23 Nov 2010 at 11:00
To make things efficient, the data stored for Detour is used as is, that is,
there is no layer to translate the data read from disc to the one that is used
in memory. The data that is loaded to memory is used directly. It makes it
really fast to load, but really hard to load old versions of data.
I could try to write conversion code between different versions, but that would
mean huge amounts of extra work for me.
Original comment by memono...@gmail.com
on 23 Nov 2010 at 11:23
Original comment by memono...@gmail.com
on 26 Nov 2010 at 12:50
Original issue reported on code.google.com by
zhibin.l...@gmail.com
on 23 Nov 2010 at 3:44