Hello Alastair,
On Mon, May 2, 2016 at 4:04 PM, Alastair McKinstry <alastair.mckins...@sceal.ie> wrote: > Hi, > > Can you point me to somewhere I can get such files (I will work with > EUMETSAT to do so, if necessary), > > and some hints as to what processing you did? Thanks for your interest in the matter. You can find a sample data file as well as a subset of the C code that we use to read it at http://labrador.oblomov.eu/grib/ —the grib file itself is compressed to reduce load on my home server. Simply gunzip the grib and make && ./test *grb will exhibit the issue. The segfaulting instruction is a printf in my grib2load.h which simply tries to access some fields that should have been initialized correctly. If you trap with gdb before it, and inspect the fields data, you'll see that it's pretty much bogus. Compared with what you get with the older version of the library to find the sane (and correct) values. Let me know if you need anything else to reproduce. Best regards, -- Giuseppe "Oblomov" Bilotta
