Re: Mailcap, tmpfiles and master processes such as browsers
Previously Mark Brown wrote:
> I would assume the tmpfile is removed by whatever started the second
> process rather than by the second process. How's it supposed to know
> that the file is still being used when the program it invokes exited?
I'ld argue that only a process that created a tmpfile should remove
it since that is the only process that knows when it is done with
the file. If you can't be sure and just blindly remove files things
break as this galeon bug demonstrates.
/firstname.lastname@example.org This space intentionally left occupied \
| email@example.com http://www.liacs.nl/~wichert/ |
| 1024D/2FA3BC2D 576E 100B 518D 2F16 36B0 2805 3CB8 9250 2FA3 BC2D |