Re: [PATCH] tail: add comments noting potential inotify-related problems

[Giuseppe Scrivano]

Hi Jim,

Jim Meyering <address@hidden> writes:

> I'm not convinced that adding a lot of new code just to make tail -f
> handle a far-fetched case like that is worthwhile.  But that's just
> my opinion, and if someone can present a use-case that makes it seem
> the additional code would be put to good use, I'll keep an open mind ;)

So you don't want tail -F to handle the case that the parent directory is
removed and after re-created?  tail will not open again the watched file
in this case, or in the case the parent directory was created after tail
initialization.
As any parent directory up to '/' can be removed, if we decide to handle
the case the parent can be removed, it is not a bad idea to do in a more
generic way.
I think the difference, in lines of code, between handle just the parent
or the full hierarchy is not much, it is just a generalization.

On the other hand, as user I rarely use -F and -f is enough, I really
don't remember any case where the parent directory could be removed or
renamed and don't handle the parent directory will not be a real
problem.

Cheers,
Giuseppe