Log message:
py-kombu: updated to 4.5.0

4.5.0:
- The Redis transport now supports a custom separator for keys.
- When the SQS server returns a timeout we ignore it and keep trying
  instead of raising an error.
- Properly declare async support for the Qpid transport.

Log message:
py-kombu: updated to 4.4.0

4.4.0:
- Restore bz2 import checks in compression module.
  The checks were removed in celery/kombu-938 \ 
<https://github.com/celery/kombu/pull/938>_ due to assumption that it only \ 
affected Jython.
  However, bz2 support can be missing in Pythons built without bz2 support.
- Fix regression that occurred in 4.3.0
  when parsing  Redis Sentinel master URI containing password.
- Handle the case when only one Redis Sentinel node is provided.
- Support SSL URL parameters correctly for rediss:// URIs.
- Revert celery/kombu-954 <https://github.com/celery/kombu/pull/954>_.
  Instead bump the required redis-py dependency to 3.2.0
  to include this fix andymccurdy/redis-py@4e1e748 \ 
<https://github.com/andymccurdy/redis-py/commit/4e1e74809235edc19e03edb79c97c80a3e4e9eca>_.
- Added support for broadcasting using a regular expression pattern
  or a glob pattern to multiple Pidboxes.

Log message:
py-kombu: updated to 4.3.0

4.3.0:
- Added Python 3.7 support.
- Avoid caching queues which are declared with a TTL.
  Queues that are declared with a TTL are now also be excluded from the
  in-memory cache in case they expire between publishes on the same channel.
- Added an index to the Message table for the SQLAlchemy transport.
  The index allows to effectively sorting the table by the message's timestamp.
- Added a timeout that limits the amount of time we retry
  to reconnect to a transport.
- :class:celery.asynchronous.hub.Hub is now reentrant.
  This allows calling :func:celery.bin.celery.main to revive a worker in
  the same process after rescuing from shutdown (:class:SystemExit).
- Queues now accept string exchange names as arguments as documented.
  Tests were added to avoid further regressions.
- Specifying names for broadcast queues now work as expected.
  Previously, named broadcast queues did not create multiple queues per worker.
  They incorrectly declared the named queue which resulted in one queue per
  fanout exchange, thus missing the entire point of a fanout exchange.
  The behavior is now matched to unnamed broadcast queues.
- When initializing the Redis transport in conjunction with gevent
  restore all unacknowledged messages to queue.
- Allow :class:kombu.simple.SimpleQueue to pass queue_arguments to Queue object.
  This allows :class:kombu.simple.SimpleQueue to connect to RabbitMQ queues with
  custom arguments like 'x-queue-mode'='lazy'.
- Add support for 'rediss' scheme for secure Redis connections.
  The rediss scheme defaults to the least secure form, as
  there is no suitable default location for ca_certs. The recommendation
  would still be to follow the documentation and specify broker_use_ssl if
  coming from celery.
- Added the Azure Storage Queues transport.
  The transport is implemented on top of Azure Storage
  Queues. This offers a simple but scalable and low-cost PaaS
  transport for Celery users in Azure. The transport is intended to be
  used in conjunction with the Azure Block Blob Storage backend.
- Added the Azure Service Bus transport.
  The transport is implemented on top of Azure Service Bus and
  offers PaaS support for more demanding Celery workloads in Azure.
  The transport is intended to be used in conjunction with the Azure
  CosmosDB backend.
- Drop remaining mentions of Jython support completely.
- When publishing messages to the Pidbox, retry if an error occurs.
- Fix infinite loop in :method:kombu.asynchronous.hub.Hub.create_loop.
- Worker shutdown no longer duplicates messages when using the SQS broker.
- When using the SQS broker, prefer boto's default region before our hardcoded \ 
default.
- Fixed closing of shared redis sockets which previously caused Celery to hang.
- the Pyro_ transport (:mod:kombu.transport.pyro) now works with
  recent Pyro versions. Also added a Pyro Kombu Broker that this transport
  needs for its queues.
- Handle non-base64-encoded SQS messages.
- Move the handling of Sentinel failures to the redis library itself.
  Previously, Redis Sentinel worked only if the first node's sentinel
  service in the URI was up. A server outage would have caused downtime.
- When using Celery and the pickle serializer with binary data as part of the
  payload, UnicodeDecodeError would be raised as the content was not utf-8.
  We now replace on errors.
- Allow setting :method:boto3.sqs.create_queue Attributes via transport_options.
- Fixed infinite loop when entity.channel is replaced by revive() on connection
  drop.
- Added optional support for Brotli compression.
- When using the SQS broker, FIFO queues with names that ended with the 'f' letter
  were incorrectly parsed. This is now fixed.
- Added optional support for LZMA compression.
- Added optional support for ZStandard compression.
- Require py-amqp 2.4.0 as the minimum version.
- The value of DISABLE_TRACEBACKS environment variable is now respected on \ 
debug, info
  and warning logger level.

Log message:
py-kombu: updated to 4.2.2

4.2.2
- Support both Redis client version 2.x and version 3.x.

Log message:
py-kombu: updated to 4.2.1

4.2.1:
The 4.2.0 release contained remains of the async module by accident.
This is now fixed.
- Handle librabbitmq fileno raising a ValueError when socket is not connected.

4.2.0:
- Now passing max_retries, interval_start, interval_step,
  interval_max parameters from broker transport_options to
  :meth:~kombu.Connection.ensure_connection when returning
  :meth:~kombu.Connection.default_connection
- Qpid: messages are now durable by default
- Kombu now requires version 2.10.4 or greater of the redis library,
  in line with Celery
- Fixed ImportError in some environments with outdated simplejson
- MongoDB: fixed failure on MongoDB versions with an "-rc" tag
- Ensure periodic polling frequency does not exceed timeout in
  virtual transport
- Fixed string handling when using python-future module
- Replaced "async" with "asynchronous" in preparation for \ 
Python 3.7
- Allow removing pool size limit when in use

Log message:
Reset maintainer

Log message:
4.1.0

- SQS: Added support for long-polling on all supported queries. Fixed bug
  causing error on parsing responses with no retrieved messages from SQS.

- Async hub: Fixed potential infinite loop while performing todo tasks

- Qpid: Fixed bug where messages could have duplicate ``delivery_tag``

- MongoDB: Fixed problem with using ``readPreference`` option at pymongo 3.x.

- Re-added support for :pypi:``SQLAlchemy``

- SQS: Fixed bug where hostname would default to ``localhost`` if not specified
  in settings.

- Redis: Added support for reading password from transport URL

- RabbitMQ: Ensured safer encoding of queue arguments.

- Added fallback to :func:``uuid.uuid5`` in :func:``generate_oid`` if
  :func:``uuid.uuid3`` fails.

- Fixed race condition and innacurrate timeouts for
  :class:``kombu.simple.SimpleBase``

- Zookeeper: Fixed last chroot character trimming

- RabbitMQ: Fixed bug causing an exception when attempting to close an
  already-closed connection

- Removed deprecated use of StopIteration in generators and invalid regex
  escape sequence.

- Added Python 3.6 to CI testing.

- SQS: Allowed endpoint URL to be specified in the boto3 connection.

- SQS: Added support for Python 3.4.

- SQS: ``kombu[sqs]`` now depends on :pypi:`boto3` (no longer using
  :pypi:`boto)`.

    - Adds support for Python 3.4+
    - Adds support for FIFO queues
    - Avoids issues around a broken endpoints file

- Zookeeper: Added support for delaying task with Python 3.

- SQS: Fixed bug where :meth:`kombu.transport.SQS.drain_events` did not support
  callback argument

- Fixed bug around modifying dictionary size while iterating over it

- etcd: Added handling for :exc:`EtcdException` exception rather than
  :exc:`EtcdError`.

Log message:
Changes 4.0.2:
- Now depends on :mod:`amqp` 2.1.4

    This new version takes advantage of TCP Keepalive settings on Linux,
    making it better at detecting closed connections, also in failover
    conditions.

- Redis: Priority was reversed so, e.g. priority 0 became priority 9.

Log message:
Switch to MASTER_SITES_PYPI.

Log message:
Add SHA512 digests for distfiles for net category

Problems found with existing digests:
	Package haproxy distfile haproxy-1.5.14.tar.gz
	159f5beb8fdc6b8059ae51b53dc935d91c0fb51f [recorded]
	da39a3ee5e6b4b0d3255bfef95601890afd80709 [calculated]

Problems found locating distfiles:
	Package bsddip: missing distfile bsddip-1.02.tar.Z
	Package citrix_ica: missing distfile citrix_ica-10.6.115659/en.linuxx86.tar.gz
	Package djbdns: missing distfile djbdns-1.05-test25.diff.bz2
	Package djbdns: missing distfile djbdns-cachestats.patch
	Package djbdns: missing distfile 0002-dnscache-cache-soa-records.patch
	Package gated: missing distfile gated-3-5-11.tar.gz
	Package owncloudclient: missing distfile owncloudclient-2.0.2.tar.xz
	Package poink: missing distfile poink-1.6.tar.gz
	Package ra-rtsp-proxy: missing distfile rtspd-src-1.0.0.0.tar.gz
	Package ucspi-ssl: missing distfile ucspi-ssl-0.70-ucspitls-0.1.patch
	Package waste: missing distfile waste-source.tar.gz

Otherwise, existing SHA1 digests verified and found to be the same on
the machine holding the existing distfiles (morden).  All existing
SHA1 digests retained for now as an audit trail.