Uhm, I don’t understand where the include error comes from then (ROOT might be trying to #include stuff on its own, but I don’t understand why that would fail).
I tried executing a ROOT macro from a directory with a space in its name, as well as reading ROOT data from a directory with a space in its name, and it worked fine (with latest ROOT master, not v6.14, admittedly).
Can you try to simplify gaussiansigma.C until it contains only the very few lines that cause the problem?
This definitely looks like a bug, but I can’t reproduce it locally. I put your macro inside file test.C, put that file in my cernbox directory, tried to run it, and it worked.
Maybe @Axel or @dpiparo have a better idea of what cling might be getting wrong.
Otherwise, feel free to open a ticket on https://sft.its.cern.ch/jira/projects/ROOT, possibly specifying a clear sequence of steps to reproduce the problem (including your cernbox setup, because things work fine with mine…?)
I’m sorry I can’t provide better suggestions.
and this dir (/Volumes/DATI/Dropbox) is a symbolic link since now I have TWO Dropbox directories, one for my “personal” (“Personale”) account and one for my team account (“AMS Perugia”). This is what is explained here and creates this stupid dirs with the spaces:
Both if I reach the dir following the path with the symbolic link (i.e. WITHOUT the space), or following the path with the space, the macro has that error.
I didn’t say but should be clear from the “simplified” macro: If, instead, I move the macro in a “normal” dir (for example my Desktop), it works fine.
As in: cannot use (. I’m debugging why - it’s likely because the parsing misinterprets this as the argument .x foo.C(12). (But I have a huge backlog after vacation, so please ping me next week should you not have heard back from me!) That’s now https://sft.its.cern.ch/jira/browse/ROOT-10097