From 91a4d1370457a2c2cd20566052adf15793fef1aa Mon Sep 17 00:00:00 2001 From: Clement Ho Date: Thu, 24 May 2018 14:05:12 +0000 Subject: Add example of when a regression fix will not be picked into subsequent releases --- PROCESS.md | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'PROCESS.md') diff --git a/PROCESS.md b/PROCESS.md index f206506f7c5..f19f9efed67 100644 --- a/PROCESS.md +++ b/PROCESS.md @@ -230,6 +230,11 @@ For instance, if 10.5.0 adds a feature, and that feature doesn't work correctly, then this is a regression in 10.5. If 10.5.1 then fixes that, but 10.5.3 somehow reintroduces the bug, then this bug is still a regression in 10.5. +In addition, if a regression appears in 10.5.0 for example and we are no longer +doing regular patch releases for 10.5 versions, the fix will not be patched for +the subsequent releases unless it is critical. In the event that the fix needs +to be patched for the subsequent releases, an [exception request](#asking-for-an-exception) can be made. + Because GitLab.com runs release candidates of new releases, a regression can be reported in a release before its 'official' release date on the 22nd of the month. When we say 'the most recent monthly release', this can refer to either -- cgit v1.2.1