Re: [Rpm-maint] [rpm-software-management/rpm] SPECPARTS dir in %_builddir/%buildsubdir is leaking to setuptools package discovery (Issue #2532)
Could the default be `SPECPARTS-%{name}` instead of `%{name}-SPECPARTS`, so
that `$RPM_BUILD_DIR/%{name}*` doesn't change meaning?
--
Reply to this email directly or view it on GitHub:
https://github.com/rpm-software-management/rpm/issues/2532#issuecomment-1705676298
You are receiving this because you are subscribed to this thread.
Message ID: ___
Rpm-maint mailing list
Rpm-maint@lists.rpm.org
http://lists.rpm.org/mailman/listinfo/rpm-maint
[Rpm-maint] [rpm-software-management/rpm] RPM 4.19.0 RC1 released! (Discussion #2645)
This is a release candidate with minor enhancements and bug fixes since BETA. With the key features in place since BETA, most of these changes are centered around the build process and the test-suite, and thus are only relevant to the packagers and developers of RPM. Among those little features that made it at the last minute is a `find_package()` config for CMake which may come in handy for projects linking to our libraries. Of particular note also is the new default of C.UTF-8 when building RPM, this can be changed back to the original value of C at configure time by setting the `ENABLE_CUTF8` CMake option to `OFF`. Details and download info at https://rpm.org/wiki/Releases/4.19.0 -- Reply to this email directly or view it on GitHub: https://github.com/rpm-software-management/rpm/discussions/2645 You are receiving this because you are subscribed to this thread. Message ID: ___ Rpm-maint mailing list Rpm-maint@lists.rpm.org http://lists.rpm.org/mailman/listinfo/rpm-maint
[Rpm-maint] RPM 4.19.0 RC1 released!
This is a release candidate with minor enhancements and bug fixes since BETA. With the key features in place since BETA, most of these changes are centered around the build process and the test-suite, and thus are only relevant to the packagers and developers of RPM. Among those little features that made it at the last minute is a find_package() config for CMake which may come in handy for projects linking to our libraries. Of particular note also is the new default of C.UTF-8 when building RPM, this can be changed back to the original value of C at configure time by setting the ENABLE_CUTF8 CMake option to OFF. Details and download info at https://rpm.org/wiki/Releases/4.19.0 ___ Rpm-maint mailing list Rpm-maint@lists.rpm.org http://lists.rpm.org/mailman/listinfo/rpm-maint
Re: [Rpm-maint] [rpm-software-management/rpm] RPM 4.19.0 RC1 (PR #2644)
Merged #2644 into rpm-4.19.x. -- Reply to this email directly or view it on GitHub: https://github.com/rpm-software-management/rpm/pull/2644#event-10272225146 You are receiving this because you are subscribed to this thread. Message ID: ___ Rpm-maint mailing list Rpm-maint@lists.rpm.org http://lists.rpm.org/mailman/listinfo/rpm-maint
[Rpm-maint] [rpm-software-management/rpm] RPM 4.19.0 RC1 (PR #2644)
Brings the `rpm-4.19.x` branch up to date with `master`.
You can view, comment on, or merge this pull request online at:
https://github.com/rpm-software-management/rpm/pull/2644
-- Commit Summary --
* Bump CI to Fedora 38
* Use CMake Lua module
* Fix leftover lua.h import from days long gone
* Make the higher level file info selection flags available to Python
* Fix file signatures getting loaded when not asked for
* Fix thinko in archive size tag(s) description
* Dont mask the user verify fail on invalid symlink test
* Verify user/group in tests when possible
* Silence an apparent false positive warning from gcc in -Og mode
* Fix a copy-paste --help description of --whatconflicts (RhBug:2208661)
* Macro documentation improvements
* Add %{rpmversion} builtin macro for getting the running rpm version
* Fix Lua docs rpmio mode table rendering (missing newline)
* Export rpmGlobPath() to Lua as rpm.glob()
* Remove obscure check for package build time from --rebuilddb (#2527)
* Add RPM_LD_FLAGS to build environment
* rpmrc: Fix how x86 models are derived
* Fix error handling of rpm.execute() and rpm.redirect2null()
* Add %specpartsdir to macros.in
* Drop unnecessary internal includes from the IMA plugin
* Add doc note about disabled option processing
* Add doc note that getopt processing only supports short options
* Show the tag number for unknown tags in --xml output
* Fix possible null pointer reference in ndb
* Only add arch, os and platform, self-provides once
* Fix rpmDigestBundleFinal() and Update() return code on invalid arguments
* Document spec parse recursing at BuildArch
* Fix per-file plugin hook regression introduced in 4.18
* Use __builtin_cpu_supports in x86-64 CPU feature level detection
* For completeness, make sure CPUID leaf 7 is supported before using it
* Drop unnecessary internal includes from the fsverity plugin
* Dont muck with per-process global sqlite configuration from the db
backend
* Actually return an error in parseScript if parsing fails
* Add some Python binding examples
* Really handle hardlinks in the rpmunpack python example, doh
* Require real names for contributions
* Fix reproducability test --define usage
* Ensure reproducability test runs with hard requires for user/group
* Unset RPM_CONFIG_DIR for package building scripts
* Mention systemd dynamic user feature in user/group manual
* Make user/group info reliable operation across chroot
* Simplify rpmug caching
* Test user-creation + verify in a --root operation
* Fix library version test to only look for ourn own libraries
* Drop bogus %{?dist} from test-suite specs
* Silence mknod failure in the test-suite setup
* Replace AT_CHECK and AT_CLEANUP with RPMTEST_*
* Fix systemd-sysusers helper path discovery
* Remove redundant runroot_other
* Capitalize our cmake vars in tests/
* Cluster add_custom_*() calls together
* Move COMMANDs onto separate lines
* Fix WITH_CAP check in tests
* Dont depend on %__cc from build system in tests
* Let rpmtests target depend on local.at
* Dont init plugins in Python tests
* Find required test programs in tests/ cmake file
* Drop test-related *dir vars from cmake
* Drop redundant atconfig
* Drop redundant TOPDIR use from tests
* De-duplicate the long command in rundebug()
* Wrap fakechroot in new sandbox function
* Break the long run*() commands into lines
* Sanitize the helper tool search path in cmake
* Gah, another last-minute edit brought back a debug warning...
* Add sandbox exec command
* Add RPMTREE variable to tests
* Rename testing/ to tree/ in tests
* Turn populate into mktree
* Create the SOURCES symlink in RPMTREE
* Shuffle configure_file() around a little
* Add sandbox mount command
* Remove toplevel _dbpath in tests, not just files
* Set %_sysusersdir independent from our prefix
* Make RUNDIR lower case (rundir)
* Fix broken link to downloads page
* Actually ensure no rpmdb in no rpmdb test
* Silence awk if no RPM_CONFIGDIR in atlocal
* Refactor mktree
* Rename sandbox() to snapshot()
* Use $RPMTEST in dir check
* Bootstrap a self-contained OS image with mktree
* Automatically remove test image on make clean
* Add support for make env
* Add RPM_CONFIGDIR_PATH to atlocal
* Replace fakechroot with proper container tech
* Start with clean environment in bwrap containers
* Add support for make shell
* Mention new test features in README
* Containerize the whole test-suite
* Base per-test snapshots on the root filesystem
* Add convenience DNF wrapper to make env
* Place TESTPROGS binaries into PATH
* Rename mktree.in
* Add support for mktree backends
* Drop ci/ directory, move Dockerfile to tests/
* Turn Dockerfile into a standalone RPM image
* Add mktree.common
* Refactor cmake
* Add mktree.rootfs backend
* Add mktree.podman backend
* Mount the common tree snapshot as
Re: [Rpm-maint] [rpm-software-management/rpm] Always run tests without --nouser and --nogroup, now that we can (PR #2641)
I think I tested that it now works without it, whether we should is, I don't know. It's a somewhat problematic feature to begin with because it's kinda experimental yet enabled by default (when running as non-root). But then, it shouldn't affect the test-suite one way or the other as-is. -- Reply to this email directly or view it on GitHub: https://github.com/rpm-software-management/rpm/pull/2641#issuecomment-1704860682 You are receiving this because you are subscribed to this thread. Message ID: ___ Rpm-maint mailing list Rpm-maint@lists.rpm.org http://lists.rpm.org/mailman/listinfo/rpm-maint
Re: [Rpm-maint] [rpm-software-management/rpm] Eliminate RPMDBI_SIGMD5 and RPMDBI_SHA1HEADER rpmdb indexes (Issue #2633)
Yes various places show them because ... it's there. And course, for the longest time that was the only checksum kind of thing that we had on the payload. -- Reply to this email directly or view it on GitHub: https://github.com/rpm-software-management/rpm/issues/2633#issuecomment-1704822481 You are receiving this because you are subscribed to this thread. Message ID: ___ Rpm-maint mailing list Rpm-maint@lists.rpm.org http://lists.rpm.org/mailman/listinfo/rpm-maint
Re: [Rpm-maint] [rpm-software-management/rpm] Add a new perl.prov script to generate normalized module versions (PR #2586)
We can do that, or just create the repo in this organization to begin with. Either way, just file the request at https://github.com/rpm-software-management/org-admin -- Reply to this email directly or view it on GitHub: https://github.com/rpm-software-management/rpm/pull/2586#issuecomment-1704813446 You are receiving this because you are subscribed to this thread. Message ID: ___ Rpm-maint mailing list Rpm-maint@lists.rpm.org http://lists.rpm.org/mailman/listinfo/rpm-maint
Re: [Rpm-maint] [rpm-software-management/rpm] Always run tests without --nouser and --nogroup, now that we can (PR #2641)
I'm wondering if we can't or shouldn't also remove `--nouserns` from `runroot()`? -- Reply to this email directly or view it on GitHub: https://github.com/rpm-software-management/rpm/pull/2641#issuecomment-1704786137 You are receiving this because you are subscribed to this thread. Message ID: ___ Rpm-maint mailing list Rpm-maint@lists.rpm.org http://lists.rpm.org/mailman/listinfo/rpm-maint
