AMP ‘Decadeus’ 2.4.2 Release Notes
This release is a major update, with significant new features and enhancements - with changes in just about every single component in AMP.
The easiest way to update (For both Windows and Linux) is to open a shell/command prompt and run:
You can find the full guide on upgrading AMP here: How to update AMP to the latest version · CubeCoders/AMP Wiki · GitHub
10 Years of CubeCoders!
This update commemorates 10 years of CubeCoders! In honour of this occasion, we’ve slapped together a swanky new theme for you all to enjoy. Select it from the Theme setting and do a force refresh to get that purply goodness!
The descriptions of settings can now include links to other settings that when clicked, take you directly to that setting. Very useful when two settings are closely related or depend on each other.
As you can see in the preview screenshot above, you can now toggle on/off schedule triggers! This lets you temporarily pause and resume specific triggers without having to remove and re-do everything.
This new feature accommodates situations where AMP isn’t reasonably able to correctly guess what hostname should be presented to users. This setting lets you set this for a given target and all instances within it. This affects things like the Connect to Server button and the SFTP server links.
Dismiss all notifications
If you’ve had a pile of notifications pile up while you were away, dismissing them all could be annoying or required you to restart that instance entirely. Now there’s a convenient ‘Dismiss All’ button that does it all in go.
The height of the notifications area has also been limited and is scrollable to avoid large numbers of notifications dominating the display.
Resource monitoring display in Docker
Previously, even when resource limits such as a memory limit were imposed on a Docker container, the AMP instance within it was still aware of the amount installed on the host and assumed that to be the limit when displaying metrics. This now behaves in a more intuitive manner and shows the container limits that the application will be limited to.
ARM64/aarch64 resource monitoring
In previous releases, ARM systems could not show some resource metrics due to differences in how things like the number of CPU cores were reported. AMP now understands this difference in behaviour and accommodates it.
Permissions system updates
A few changes have been made to the permissions system. Namely that users created inside an instance can no longer see any users that were created outside that instance when they look at the list of users.
- Search results for settings no longer show markdown links as their markdown, and instead just shows the text, ignoring the URL.
- Search results no longer try and show hidden settings that aren’t accessible to the user.
- Fixed the container notice being incorrectly shown for applications that support containers, but that don’t outright require/recommend them.
- Fixed update stages stalling on certain types of long-running tasks.
- Browsing a directory in SFTP that contains a file whose name contains a backslash will no longer prevent you from seeing files in that directory (but the file with the slash in its name will not be visible)
- Fixed an issue where update stages might run out-of-order
- Fixed a handful of issues with BattleEye RCON (although some sporadic issues may remain)
- Fixed an issue where failing to sleep could cause AMP to stall entirely because the status wasn’t checked.
- Fixed the CreateFile update stage not honouring the OverwriteExistingFiles flag correctly.
- Fixed PreStartStages not being filtered based on OS.
- API change - Parameter names are no longer case sensitive.
- Permissions system change - Users can no longer add users to a role if they do not already have every single permission within that role.
- Fixed an issue where AMP startup could stall if the external IP address could not be resolved promptly.
- Fixed an issue where it wasn’t possible to remove a target from a controller if the target was offline.
- Fixed an issue where if using BypassController - plugins were not loaded on target ADS instances which made instances unmanageable with the default configuration.
- BattlEye and SourceRCON clients now tolerate bad packets and just try to carry on rather than dropping the connection on a bad packet.
- Added an additional Bedrock Version fallback for Linux systems.
- Updated Rust module to include new Query port.
- Updated URLs for FiveM module updates.
- Bad certificates for web server will actually prompt AMP to shut down and trigger diagnostics rather than sitting there doing nothing.
- Fixed not being able to trash files/directories that don’t have a matching directory structure within Trash already.
- Fixed a crash on shutdown when AMP tries to end all existing sessions where some of those sessions don’t have associated Websocket sessions.
- Miscellaneous UI tweaks and improvements.
- Fixed a divide by 0 error if number of cores/threads is reported as 0. This will never report below 1 now even if something goes wrong.
- ADS waits for a full init when logging in rather than a fast-init - normally fast is fine but because of the delay in getting instance information, this change avoids a jarring pop when the instances load.
- ADS instances lazy-load their background images and fade them in nicely as you scroll down the list. Much nicer to look at rather than images jumping in randomly.