macOS Monterey: "This app will not work with future versions of macOS"

I recently updated to macOS 12.0.1 (Monterey). Right after booting up the first time, I got a message informing me that Duplicati would not work with future versions of macOS. I found a similar issue for a different app on the Apple developer forum, and the image of the message there is similar to the one I saw.

To clarify, backups are working fine with the macOS 12.0.1 (not tested restore). However, I wanted to flag this in case it creates future issues.

1 Like

Here’s a screenshot I just captured after updating to macOS 12.1

Screen Shot 2021-12-13 at 8.04.09 PM

Hi,

I just upgraded my Duplicati machnine to Monterey too. And am getting this message. Is there any updates to this problem?

/Raymond

Could you click “Learn More…”? If it’s a public web page, you can post the link rather than screenshot it all

I don’t have a Mac, and when this initially came up there wasn’t any information search could find that fit…

This might need more volunteers with macOS who can help (especially if you have some developer skills).
Duplicati team is low on volunteers in general (any out there?), and might be especially low on Mac ones…

because a problem that no volunteer can even see is a problem that the volunteer can’t really investigate.

Duplicati is a community effort and is limited by what volunteer resources the community will provide to it.

Welcome to the forum @raymanoz

Could someone here please help on:

It sounds like it’s successful except some future version is in question, but I don’t want to speak for you…

Op here. I can confirm the Duplicati works on Monterey 15.1 and 15.2.

The issue that the dialog box is referring to stems from python 2 being deprecated.

I noticed a thread over a year old that dealt with upgrading references from python 2 to python 3

It look like it was solved, but maybe not for Mac OS

Learn More takes you to this page

https://www.python.org/doc/sunset-python-2/

Thanks for the tip! The more general Python script is utility-scripts/DuplicatiVerify.py. It looks at least largely Python 3 ready, as found by running 2to3-2.7 over it (per an article I found). Tool suggested one line deletion. Script header says “tested with python 2.7 & python 3.8”, but I don’t know what macOS’ complaint looks for.

In comparison, rumps.py which is macOS specific and launches the TrayIcon, seems old and needing lots.

https://github.com/jaredks/rumps/blob/master/CHANGES.rst

Add Python 3 support

might be newer than Duplicati’s version, but this needs more expertise. Any experts? Can you file an Issue?

If a daring enough macOS user wants to experiment with the conversion tool on rumps.py on their Mac, feel free to see if a conversion can avoid message while still keeping things going (save the old file just in case).

Rumps is useful, but https://macmule.com/2021/10/25/macos-monterey-prompt-needs-to-be-updated/
has a more detailed description of what causes the MacOS prompt

The prompt is triggered by a process calling: /usr/bin/python. It could be the pkg, the pkg’s preinstall, or postinstall script

Does this help?

Thanks for the tip. That helps, but now I need someone with a Mac to do " How do we start to uncover what is calling Python 2?" to see if it’s just rumps.py. My method was to 7-Zip extract a.dmg, then I search for import which Python scripts often use. It’s possible I missed one. As there’s a way to know for sure, let’s try that one.

Alternatively maybe filing an issue will get the development team working on getting someone in to work this…

I think an issue is the best path. There’s compiled code at https://github.com/duplicati/duplicati/blob/master/Duplicati/GUI/Duplicati.GUI.TrayIcon/RumpsRunner.cs that might be rumps start, meaning there’s a bit of a chain to trace through. Might need to see if the original author can help with this.

I think an issue is the best path. There’s code at https://github.com/duplicati/duplicati/blob/master/Duplicati/GUI/Duplicati.GUI.TrayIcon/RumpsRunner.cs that might be the rumps starter, meaning that there’s a somewhat long path there, and the debugging technique that article suggested might not quite fit well.

Could someone on this topic file an issue?

Issue is open. Thanks! Below link to it will post itself there, so we can connect this and issue both ways.