We do not seem to have a coherent approach to how to handle
debian/changelog in trees (eg, vcs branches) which are not yet ready
for upload.
See below two examples of things done differently.
The questions are:
Q1. Should the suite in the changelog entry be UNRELEASED,
or the suite at which the vcs branch is targeted ?
Q2. Should the changelog entry be finalised ? That is, should it
have an uploader name and date ?
Q3. Should the version number be the intended next version number,
or should it be decorated with ~something ? If it should
be decorated, what should it be decorated with ?
Discussion:
Q1: Replacing the suite with "UNRELEASED" means that it is not
possible to distinguish a branch intended for experimental from one
intended for sid, other than by branch name (and of course branch
names are often used for other purposes). I think this is very
undesirable. AFAICT only reason to use UNRELEASED is to prevent
accidental upload, but maybe we can do that some other way.
Q2: The information in the trailer line of the changelog entry in a
new branch is not meaningful. Depending on the vcs etc. in use, it
can cause pointless merge conflicts. Often the information is quite
wrong: for example, a date which is earlier than the most recent
commit. The Emacs changelog mode doesn't like editing finalised
changelog entries. IME several of our tools deal badly with
unfinalised changelog entries. For example, mergechangelogs
mishandles them, and some parts of the package building toolchain
sometimes complain. However, such bugs can be fixed and the fixes
deployed fairly easily (via backports if necessary).
Q3. If the version number is not decorated, then binaries (or even
source packages!) built for testing (or for other reasons other than
upload) are not always distinguishable from intended-for-release
builds. IMO this is very undesirable. Also, it can make the "version
X contains changelog entry for version Y" relation confusing: ideally
that would mean that X contains all the changes in Y, but if Y had a
vcs branch with finalised changelog entry that was not uploaded, and
the person who uploaded made further changes but forgot to push, then
X might be missing those changes from Y, accidentally. Therefore IMO
"working" changelog entries should have a decorated version number. I
suggest "~UNRELEASED" because (i) it resembles the UNRELEASED suite
which is sometimes used; (ii) U sorts before all the uppercase
letters, so it's quite early.
Proposal:
Step 1: dak should be changed to reject all uploads whose version
number contains UNRERLEASED.
Step 2: All tools should be changed to produce/expect that
working versions have:
- suite = the intended suite
- unfinalised changelog entry
- version ending in ~UNRELEASED
In particular:
* Tools which currently have a "suite=UNRELEASED" check should check
the version number too, the same way as dak.
* Tools which add/edit changelog change items should insist that the
changelog is unfinalised and contains ~UNRELEASED in its version.
* Tools which generate new suite=UNRELEASED changelog entries should
be changed to generate unfinalised version~UNRELEASED ones, with a
real suite, instead.
* Tools which prepare changelogs for upload (eg by finalising the
changelog) should now remove ~UNRELEASED from the version, and
finalise the changelog. Tools which currently replace UNRELEASED
with a suite can continue to do so, but should print a warning when
they find themselves doing so.
* Anything which barfs on or mishandles unfinalised changelogs should
be fixed.
What do people think ?
Ian.
dgit (3.10~) unstable; urgency=medium
Bugfixes:
* dgit: Copy several user.* settings from main tree git local config
to dgit private workarea. Closes:#853085.
* dgit: Strip initial newline from Changes line from dpkg-parsechangelog
so as to avoid blank line in commit messages. Closes:#853093.
* dgit: Do not fail when run with detached HEAD. Closes:#853022.
* dgit: Be much better about commas in maintainer changelog names.
Closes:#852661.
Test suite:
* quilt-useremail: New test for user config copying (#853085).
* lib-import-chk: Test that commits have smae authorship as appears in
the changelog. (Or, at least, the same authorship set.)
* import-maintmangle: New test for changelog Maintainer mangling.
Documentation:
* Fix typos. Closes:#853125. [Nicholas D Steeves]
--
sysvinit (2.88dsf-59.9) UNRELEASED; urgency=medium
[ Martin Pitt ]
* Mark sysvinit-utils as Multi-Arch: foreign, like sysv-rc and initscripts.
[ Michael Biebl ]
* Demote priority of sysv-rc and initscripts to optional.
[ Petter Reinholdtsen ]
* Avoid remounting tmpfs, linprocfs and linsysfs on kFreeBSD
(Closes: #833687). These file systems are not remountable. The
change avoid a warning from mount. Based on Patch from Jon Boden.
[ Ian Jackson ]
* Add myself to Uploaders, as part of adopting the package.
Closes:#811377 (RFA bug).
* Add Benda Xu to Uploaders, as requested in #811377.
[ Ben Hutchings ]
* Keep /usr mounted read-only on shutdown (Closes: #757083)
-- Ian Jackson <ijack...@chiark.greenend.org.uk> Wed, 18 Jan 2017 23:23:35
+0000
--
Ian Jackson <ijack...@chiark.greenend.org.uk> These opinions are my own.
If I emailed you from an address @fyvzl.net or @evade.org.uk, that is
a private address which bypasses my fierce spamfilter.
