From mboxrd@z Thu Jan 1 00:00:00 1970 From: Adam Porter Subject: Org should follow SemVer conventions [Was: Automatically save the archive file after org-archive-subtree] Date: Tue, 09 Oct 2018 01:05:00 -0500 Message-ID: <87tvlv4pv7.fsf@alphapapa.net> References: Mime-Version: 1.0 Content-Type: text/plain Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:56631) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1g9l95-0001Mh-Vv for emacs-orgmode@gnu.org; Tue, 09 Oct 2018 02:05:28 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1g9l8z-000188-8w for emacs-orgmode@gnu.org; Tue, 09 Oct 2018 02:05:26 -0400 Received: from [195.159.176.226] (port=54512 helo=blaine.gmane.org) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1g9l8q-00016X-R0 for emacs-orgmode@gnu.org; Tue, 09 Oct 2018 02:05:15 -0400 Received: from list by blaine.gmane.org with local (Exim 4.84_2) (envelope-from ) id 1g9l6g-0001Zr-LD for emacs-orgmode@gnu.org; Tue, 09 Oct 2018 08:02:58 +0200 List-Id: "General discussions about Org-mode." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org Sender: "Emacs-orgmode" To: emacs-orgmode@gnu.org Kodi Arfer writes: > As of > > https://code.orgmode.org/bzg/org-mode/commit/b186d1d7236c0dc397eadeb004c9a17eaffd3aab > > archiving a subtree no longer automatically saves the archive > file. How can I get this behavior back? > > A Debian bug was also opened for this issue: > > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=887332 According to that bug report, the behavior changed when upgrading from 9.1.2 to 9.1.5. I think this shouldn't happen. "Patch"-level releases (incrementing the third number) should only contain bug fixes. Changes which change default behavior belong in, at minimum, "minor" releases (incrementing the second number). Users should be able to freely upgrade from 9.1.x to 9.1.x+1 without having to worry about reading a changelog and adjusting their config to avoid changes in behavior which could lead to data loss. In my own (much smaller) projects, I follow this practice, which is basically SemVer. When I fix bugs, I apply the fixes to the latest stable branch, make a new stable patch-level release, and then merge the fixes into the master branch (in some projects, by rebasing the master branch on top of the stable branch, and in others, by a non-fast-forward merge of the stable branch into the master branch). The master branch becomes the next stable branch when its new features and other changes have been merged for long enough and have no known bugs. New features and refactorings are generally developed in a feature branch and then merged into the master branch when ready, eventually being released in the next stable branch. The master branch is intended to be always usable, but only recommended to users who are willing to potentially encounter bugs in new features or as a result of other significant changes. (In practice, this ends up being the branch used by 99.9% of MELPA users, but that's a MELPA-specific issue.) Most users should use the latest stable branch. This practice also makes it much easier for downstream packagers, like Debian, because changes to the stable branch are ONLY bug fixes. Could Org start doing this and call it Org 10.0?