This is the mail archive of the
mailing list for the GDB project.
Re: New deprecation procedure
- From: Joel Brobecker <brobecker at adacore dot com>
- To: Pedro Alves <palves at redhat dot com>
- Cc: Ulrich Weigand <uweigand at de dot ibm dot com>, GDB Patches <gdb-patches at sourceware dot org>
- Date: Thu, 11 Sep 2014 12:03:48 -0700
- Subject: Re: New deprecation procedure
- Authentication-results: sourceware.org; auth=none
- References: <201409101635 dot s8AGZrEC010511 at d06av02 dot portsmouth dot uk dot ibm dot com> <54109F88 dot 9090405 at redhat dot com>
> >> So let's discuss the new obsoleting procedure, so we can document it:
> >> . I think that the first 4 steps (post email on gdb@, wait a week,
> >> then on gdb-announce, wait another week) are fine. Anyone thinks
> >> we should go straight to gdb-announce?
> >> My thinking is that people interested in maintaining a port
> >> with enough skills to do so are likely to already be on gdb@,
> >> so we can avoid sending an extra mail to gdb-announce. But
> >> the traffic on gdb-announce being very low, and the frequency
> >> at which we deprecate targets being fairly small as well,
> >> I wouldn't object to a simpler procedure where we email
> >> gdb-announce directly.
> >> . Remove steps 5 & 6 that mark the code as obsolete, only keeping
> >> the last test, which removes the code. I'd add a note to add
> >> a NEWS entry.
> > Sounds all good to me.
> To me too.
OK, thanks all! Wiki page updated accordingly.
Sharing a thought that crossed my mind: I thought about increasing
the amount of time we wait between steps, from 1 week to say, 2 weeks,
giving anyone about a month to step up. I eventually dropped the idea
because someone stepping up late should easily be able to revert
the removal, particularly now that we've switched to git. In the
meantime, since we suspect no-one is usually going to step up,
waiting longer just defers the corresponding cleanups we want to do.
If you guys agree with that, I'll add something to the wiki page
to explain the thought process.