3 * In #841294, the Technical Committee was asked to overrule the
4 maintainer of the 'global' package to get a new upstream version
6 * As a matter of fact, at the time #841294 was filed, the 'global'
7 package's latest upload to unstable had happened in October 2010,
8 despite several requests for newer 'global' upstream releases and
10 * The discussion, involving various people ranging from bugreporters,
11 Debian contributors, the 'global' maintainer, and some TC members, has
12 clarified two lines of argumentation around the maintenance of the
14 - global is fine as it is, version numbers are no silver-bullet, and
15 there are severe problems in the new upstream versions, that are
16 being discussed with upstream. New features could always be
17 backported to the Debian version if worthwhile bugs were reported.
18 - there's a rightful expectation to get new upstream versions, even if
19 they introduce regressions or functionality losses. No amount of
20 upstream problems justify holding new versions back over multiple
25 * Our Social Contract's "We don't hide problems" implies that
26 maintainers go through reasonable effort to make their packages'
27 problems visible; and the usual way is to use the Debian bug tracker.
28 It also implies reporting upstream flaws to upstream, ideally in public.
29 Adding references to the BTS would avoid the impression that nothing had
31 * Integrating recent versions of upstream software is a maintainers'
32 duty, as Debian is a primarily a software distribution; distributions
33 exist to facilitate users' access to upstream software. Uploading recent
34 versions and making them available to Debian users on a somewhat regular
35 basis is our way to find, address and correct problems brought in by new
36 upstream releases. The 'experimental' suite exists explicitly for the
37 purpose of testing software not immediately suitable for release towards
38 future stable releases.
39 * If the maintainer decides that our users will be best served by not
40 upgrading, this should be explicitly stated. The README.Debian file
41 of the package would be a good place to do this, as well as in response
42 to bugs requesting upgrades.
43 * The argument that features could easily be backported would carry
44 significantly more weight if there was evidence of patches for past
45 bugs being acted upon in a timely manner.
49 - Option A - Reaffirm Ron Lee as 'global' maintainer (§6.1.2)
51 - Option B - Declare Wookey as 'global' maintainer (§6.1.2)
53 - Option C - Decline to rule, consider case closed
55 - Option FD - Further discussion
59 We invite all interested parties to contribute in good faith for the
60 best possible 'global' package. Filing bugs with appropriate severities
61 is every user's duty, and it is important that those who understand the
62 package best continue to provide their best inputs.