This is the mail archive of the
mailing list for the glibc project.
Re: [PATCH] manual: Correct type in struct timeval/timespec.
- From: Joseph Myers <joseph at codesourcery dot com>
- To: Roland McGrath <roland at hack dot frob dot com>
- Cc: J William Piggott <elseifthen at gmx dot com>, Carlos O'Donell <carlos at redhat dot com>, <libc-alpha at sourceware dot org>
- Date: Fri, 13 Feb 2015 12:25:47 +0000
- Subject: Re: [PATCH] manual: Correct type in struct timeval/timespec.
- Authentication-results: sourceware.org; auth=none
- References: <6948004 dot VhtuX0CvNp at descartes> <54DD0944 dot 7070003 at redhat dot com> <54DD4AFF dot 5030505 at gmx dot com> <20150213025214 dot 2A2142C3C18 at topped-with-meat dot com>
On Thu, 12 Feb 2015, Roland McGrath wrote:
> I don't think we should require bugs filed for changes to the manual.
Agreed. I think the expectation to file bugs should only be for issues
that were bugs in installed code, not documentation, that were
user-visible in releases. Not for new features, not for documentation
issues, not for non-user-visible issues such as cleanups or testsuite
defects (as opposed to user-visible bugs that happened to be shown up by a
testsuite failure). (It can still be useful to file bugs in some such
cases - for example, to record that a well-defined cleanup is desired but
you're not working on it - but not required for cleanups, manual issues
etc. if you're sending the patch and taking charge of pinging it until it
Joseph S. Myers