Sender: tech-pkg-owner@NetBSD.org

Rather than killing manually, perhaps we should see about putting in
some sort of watchdog timeout mechanism.  I know it could be done in
shells, and is relatively simple to do.  But the best method would be
smart enough to set that the file is being updated and reschedule, or
to teach the ftp client to have a timeout.  I know I have one web bot
which has such a timeout, which is started before every read or other
blocking operation and canceled after it returns.

Thoughts?

- Doug

Quoting Thomas Klausner (wiz@NetBSD.org):
> On Mon, Oct 18, 2004 at 04:35:49PM +0200, Roman Kennke wrote:
> > I also read the recent discussion about the 'make update
> > hell'. I also encounter this kind of problem: sometimes
> > while fetching distfiles, I get stalled connections. This
> > forces me to interrupt with CTRL+C, which then leaves me
> > with deinstalled packages here and there.
> 
> A better way to handle this particular problem is to kill
> the ftp process from a second shell -- that way just the
> next site will be tried, and the build will continue.
>  Thomas

-- 
Douglas Wade Needham - KA8ZRT        UN*X Consultant & UW/BSD kernel programmer
Email:  cinnion @ ka8zrt . com       http://cinnion.ka8zrt.com
Disclaimer: My opinions are my own.  Since I don't want them, why
            should my employer, or anybody else for that matter!