[GRASS-dev] Re: About the vector changes

[Glynn Clements]

Markus Metz wrote:

> Maybe it is less effort and less complicated to make everything LFS-safe 
> instead of creating workarounds for modules that are not LFS-safe when 
> LFS is enabled for lib/gis? One way or the other, I guess these modules 
> need to be identified and possibly modified.

It certainly looks that way.

It's basically a case of examining anything which uses lseek(),
fseek() or ftell(), converting fseek/ftell to G_fseek/G_ftell
respectively, and ensuring that offsets are calculated and stored
using off_t.

I *think* that I fixed all of the lseek() cases some time ago, but I
didn't touch the fseek/ftell cases (we didn't have G_fseek/G_ftell at
that time).

-- 
Glynn Clements <glynn at gclements.plus.com>