enforce createrepo_c gzip compression (f39+ switched to zstd)
self-identify the resalloc resource in logs
dropping the documentary playbooks from copr-backend payload
nicer unknown-resalloc-tickets output
worker to not call keygen for source builds at all
don't sign products of srpm-build
longer timeout for fallback generating GPG keys after build
recreate missing repodata so that prunerepo doesn't traceback
use the rename HashWorkerLimit instead of GroupWorkerLimit
provide per-arch & per-owner worker limit implemented
collect and compress fedora-review logs after run
react on staled SSH connections in some cases

dump the /update/ payload to worker.log
don't run external command(s) to collect built packages
don't eat the "build detail collecting" traceback
fixes in the unknown-resalloc-tickets.py helper
more careful format_evr() method
fix tests for zst compression on F39+
log task dict in case of error returned from redis
skip builds for ExcludeArch and "not" ExclusiveArch
offload NEVRA (s)rpm parsing to copr-rpmbuild
redis authentication support added

copr_prune_results.py: work-around the arg_max problem

copr_prune_results.py: don't enforce appstream-builder, ask FE
copr_prune_results.py: just one API call to FE per one repo

Forking: better handle FileExistsError
Run the copr-rpmbuild command with task URL, not build ID

Bump version for release mess

Run createrepo without --database
Make copr_messaging optional

Don't include package name into srpm result dir name
Remove libmodulemd1 dependency
Hardlink RPMs while doing rawhide_to_release
Make sign key domain name configurable

Skip the test_run_prunerepo test because of Koji

Fix chroot version parsing with new python-packaging
Fix traceback for non-existing tasks
Python: drop the unneeded marshmallow dep
Log general exceptions to find issues more easily
Proper log argument formatting instead of .format
Use spdx license

start systemd services after the redis.service
build worker - list the built RPMs with rpm --nosignature

use OpenPGP v4 signatures
migrate from pipes to shlex
require redis.service to be started
move to GitHub home page
add resultdir cleaner
move dispatcher and background workers to copr-common
de-prio IO for the analyze-results script
don't traceback when there are no files in the S3 storage
allow devel instance to remove access files
send non-CDN hitcounter hits in chunks
copr-backend-unknown-resalloc-tickets script
work with multiple CDN hostnames per instance
move setup_script_logger to copr-common

sign everything EPEL-5+ with sha256 hashalgo

aws-hitcounter: remove temporary files as soon as possible
aws-hitcounter: ignore downloaded SRPM files
aws-hitcounter: decode special characters from URLs

count only hits from an appropriate CDN hostname
add option for infinite number of attempts to the hitcounter script
print more reasonable output from AWS hitcounter script

log every request that is sent to frontend