This second RC of 1.5.0 fixes some regressions observed with the first one and also improves some things in newly added and changed functionality:

Improvements

  • #3804 - Improve handling of corrupt users.yaml file and refuse server start if broken and default file based UserManager is configured.
  • #3811 - Only start tracking resend ratio after a set amount of lines has been transmitted.
  • #3822 - Prevent circumvention of access control through admin subgroups/permissions

Bug fixes

  • #3805 (regression) - Fix an internal server error on GET /api/files/local/<filename>
  • #3812 (regression) - Backup: Fix an AttributeError on CLI usage
  • #3813 (regression) - Discovery: Fix zeroconf discovery under Python 2
  • #3814 - Fix an error on language pack upload under Python 3; not a regression, but still included due to being minimally invasive (see also #3815)

Please also take a look at the following heads-ups:

Heads-up: Access Control is now mandatory and no longer can be disabled

If you so far had Access Control disabled, upon upgrading to 1.5.0, OctoPrint will prompt you to create a user name and password for the (first) admin user. This step was sadly necessary as too many people still will happily expose their completely unsecured OctoPrint instance on the public internet, causing additional support overhead from both attacked users and security researchers. See this guide for a way to have OctoPrint log you in automatically when connecting from an internal IP.

You can find the full changelog and release notes as usual on GitHub.

Special thanks to everyone who contributed to this release candidate and provided full, analyzable bug reports, you help making the next release as stable as possible!

A special Thank You! to this fine person for their PRs!

As the past RCs have shown me that a lot of people appear to be unaware of this: Please do not install this RC if you expect a fully stable version. It is not a stable release, it is a release candidate: severe bugs may occur, and they might be bad enough that they make a manual downgrade to an earlier version necessary - maybe even from the command line.

You should feel comfortable with and capable of possibly having to do this before installing an RC.

If you want to and can help test this release candidate, you can find information on how to switch to the "Maintenance RCs" release channel in this guide if not already done (also linked below).

Please provide feedback on this RC. For general feedback you can use this ticket on the tracker. The information that everything works fine for you is also valuable feedback 😄. For bug reports please follow How to file a bug report - I need logs and reproduction steps to fix issues, not just the information that something doesn't work so make sure to fill out all fields of the issue template.

Thanks!

Depending on the feedback regarding this version I'll look into fixing any observed regressions and bugs and pushing out a follow-up version as soon as possible and necessary.

Discuss!