I am trying to read the "isobaric main" files at the OSO server directory ncepe/pggrib/..., which I understand are in GRIB format, into plain text for further processing but am unable to find a decoder which works for all such files and hope someone may steer me towards a solution.

I first tried the WGRIB program I found at wesley.wwb.noaa.gov. That program built sucessfully on my SGI and worked for several of the files - but for many others I got the error message "ran out of data or bad file".

I then tried following the link on the MAPS page to www.unidata.ucar.edu. There seems to be a program GRIBDUMP which would do what I want but I was unable to build that package. Its documentation is not very helpful, but my guess was that the reason for the failure was that I needed to have the NETCDF package already installed. However, I was unable to build that package due to a failure when running its "configure". (I should note that I have built many, many packages successfully on my SGI.)

I then found some decoders at ncardata.ucar.edu, but they
seem rather old and I see references to grib1 & grib4 & grib4a, which I do not understand.

Other references to GRIB decoders seem to be part of packages such as WXP or GRADS, to which I do not have accesss.

So at this point I seem to be unable to penetrate the GRIB wall surrounding the OSO files and am hoping that someone can provide a roadmap towards a solution.