EventStore 3.5.0

3.5.0 is up with lots of new projection fixes and UI goodness.

Event Store Server

(#782) - (All Platforms) - Show scavenges and their statuses in the UI.
(#792) - (All Platforms) - Show TCP connections and their statistics in the UI.
(#793) - (All Platforms) - Default the db to the current directory if
an Access Denied exception is encountered.
(#794) - (All Platforms) - Adds logging for http requests and
responses. Enable with --log-http-requests
(#799) - (All Platforms) - Shows statistics about replicas in a cluster
(#808) - (All Platforms) - External TCP service will now use the
correct heartbeat interval.
(#809) - (All Platforms) - Projections routes will now return 404 if
projections aren't enabled. Was previously timing out.
(#811) - (All Platforms) - Add Trusted Writes to TCP. This enables
nodes to send writes over the internal TCP connection if the user is a
trusted user. In this case, system.
(#813) - (All Platforms) - Fix stuck projections and include common logging
(#816) - (All Platforms) - Fix incorrect arguments in verbose logging
in the client api.
(#818) - (All Platforms) - Persistent Subscription correctly reflect
last known event number
(#819) - (All Platforms) - Introduce a System Ready message

Event Store UI

(EventStore/EventStore.UI#108) - Show scavenges in the UI.
(EventStore/EventStore.UI#111) - Show TCP stats in the UI
(EventStore/EventStore.UI#113) - Shows stats about replicas in a cluster
(EventStore/EventStore.UI#114) - Adding seconds on the Created Date
and Timestamp columns
(EventStore/EventStore.UI#116) - Redirect the user if they try to
access projections when they're disabled

.NET Embedded Client

(#807) - (All Platforms) - Embedded Client will now correctly set
stats on public. Was previously setting admin on public.
(#798) - (All Platforms) - Fixing Occasional Race Condition With Authentication

Great! Thanks for this, rather significant, upgrade :slight_smile:

is there a binary release too? Specifically looking for windows.

Congrats! Any tips for upgrading a live cluster?

Take a slave down. Upgrade. Take a slave down upgrade. Take last node
down upgrade.

This is valid for all semantic versioning without a major version change

Awesome, thanks!