On 01/10/2016 03:44 AM, Michael Gilbert wrote: > On Sat, Jan 9, 2016 at 9:27 PM, Jens Reyer wrote: >> Is there any reason not to use exec in the wrapper script's last line? >> Otherwise I'd suggest to do that. > > Is there a particular reason to? The overhead of the shell process > continuing isn't so large.
I'm not aware of any issues that actually do harm. But it's generally recommended for wrapper scripts (based on the first three results of a quick web search for exec+wrapper, including the FreeBSD porters handbook [1]). Reasons stated are: - PID stays the same, allowing e.g. easier process communication, however I'm not aware of this being relevant for wine. Not sure, especially for the wineserver, where I don't use exec either. Tests so far didn't show any issues however. - Saves system resources (you already commented on that). - Saves one process fork. ... and imo the output of ps is easier to read. >> AFAIK manpages.patch was needed because the "-" were incorrectly >> encoded, and didn't work for copy&pasting. This seems to be fixed now, >> so unless there are other issues we can drop that. > > This fixed a lintian error at the time. If lintian no longer flags it > anymore, then ok, otherwise no. Yes, lintian doesn't complain about that. Committed. Greets jre [1] http://www.freebsd.org/doc/en/books/porters-handbook/dads-sh-exec.html _______________________________________________ pkg-wine-party mailing list [email protected] http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-wine-party
