[Q] Why use `-c` option for loading config in launch scripts?

[Apaczer]

I assumed it was to replace default path&name of standard gmu.conf, so when running through one of launch scripts e.g.: gmu-pandora.sh we would end up using always gmu.pandora.conf for read/write.
However instead the script only performs read on that custom conf, and all the settings are reseted on next run - nothing never changes as such when running through script. My small remedy:

if ! test -e firstrun; then
    ./gmu.bin -c gmu.pandora.conf
    touch firstrun
else
    ./gmu.bin
fi

Maybe all scripts should be updated this way or perhaps I should use an -d option in a script such way:

mv gmu.pandora.conf gmu.conf
./gmu.bin -d $(pwd)

@jhe2 How do you aproach this for less possible end-user confusion when shiping distribution pkg?

[jhe2]

So what the -cflag was supposed to do is let you specify an initial config file for reading the configuration. It was never supposed to write to that config file, but instead write a new config file (typically in the user's home directory in ~/.config/gmu/gmu.conf). If such a config file was then found on the next run, Gmu would use that instead of the specified initial config file.

If it doesn't do that, then this is a regression, that I should look into. The way it was meant to work, your script changes shouldn't be necessary. So I would rather like to fix the issue, since it really wasn't meant to be that broken. ;)

I can see how this can be confusing for those explicitly using the -c flag, so maybe it should be documented better.