Is it by design that the SNAP gpt command requires full file path names for everything(graph file name, input file names) and does not first check the current working directory? Additionally, the default directory for output products ends up being the snap home directory, which is confusing. This was not the behaviour of BEAM gpt.
Will this be fixed in a future release? I am wondering whether I must rewrite my scripts.
Thanks!
(P.S. It is also a little odd that the SNAP installer does not create a symlink for gpt as it does for the snap and snap-conf-optimiser commands.)