https://sourceware.org/bugzilla/show_bug.cgi?id=25502
Aaron Merey <amerey at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |ASSIGNED
--- Comment #5 from Aaron Merey <amerey at redhat dot com> ---
(In reply to Mark Wielaard from comment #4)
> (In reply to Tom Tromey from comment #3)
> > (In reply to Mark Wielaard from comment #2)
> > If the file can be deleted and then re-created without difficulty,
> > then it should go in the cache directory as well.
>
> It could be, if the database is gone then it can be recreated from scratch.
> But it might take a lot of work depending on how many packages/dirs you want
> to index.
>
> On the other hand if the database is important it might be explicitly be
> given on the command line.
>
> Any other opinions on where the default .debuginfod.sqlite should go (and
> how it should be named)?
According to the XDG spec, XDG_CACHE_HOME is meant for "non-essential data".
I'm not 100% sure what constitutes essential data here but if a server is able
to rebuild its database after a wipe and continue functioning as it previously
did, then I suppose the database isn't essential in this sense (even if it
might take a while to rebuild).
I think it's fair to default to XDG_CACHE_HOME since if the database is
particularly important to a user they'll probably just specify its location
like Mark said.
--
You are receiving this mail because:
You are on the CC list for the bug.
