I'm not sure - it looks like they do have a signed release file here: https://mirrors.sonic.net/pureos/repo/pureos/dists/byzantium/

That works. It must be a problem with mirrors.sonic.net.

You can try the main repos;

Thank you @mak ! I will retry in a few days then. Do you know how many days?

What do you mean with "breaks"? Does it crash? Is it not installable? In the former case, a backzrace or at least console output would be nice.

This should be fixed now, but the update will need a few days to migrate out of landing.

Exciting - we should test this in Byzantium and hopefully we can close this bug.

Debian bug report: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=928893 indicates that this has been fixed.

@jeremiah.foster Perhaps this might help:

I am sorry that did not help.

Thank you @Wayne, but I don't find a solution there or something that could help me.

You may want to check and track this post. It may be related to your issue.

they're already on it!

If you follow the upstream discussion you can see that they're already on it! :-)

ok, so we should contact the Litrokey app devs? To see if they merge that request?

Debian maintainer states "I'm concerned that the addition of vendor_id might cause an ABI break. There are multiple packages that use libnitrokey in Debian (not just nitrokey-app) and so we can't silently change the ABI."

I've pulled down the Debian packaging from Salsa and reused that. I've added in the patches from upstream and built package for PureOS Amber and Byzantium.

Please note that this issue is about whether or not a package in PureOS is violating policies defined by Debian and/or FSF.

I wanted to add to this issue that Geogebra is no longer updated in Debian due to the license, cf. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=692728. The problem is not only the documentation is licensed for non-commercial, but also the language files. As a result, it seems impossible to me to have a package with the free parts only, because such a version would not include any strings and hence be unusable?

I tried octave in Byzantium and the bug is not present (at least the proposed example above works).

Looks like this is the diff;

Matthias is looking upstream to see if the fix has been already made in effort to de-duplicate work.

Yes, it has. The PoC is basically replacing the nitrokey pro device and vendor ID with the Librem Key device and vendor ID

Has the PoC been tested with the app?

No, that repo is a poc.

So, if I understand correctly, this repo holds the patched libnitrokey: https://source.puri.sm/joao.azevedo/libnitrokey/

Thanks @selea.

This does not break an install, that is correct.

I'm happy to work on this TODAY but this is not "Unbreak Now!" because it is not breaking anything for our users and, we have a perfectly viable alternative that is a tiny bit better, namely the Librem Key.

I confirm that it is working properly in debian-live with Gnome 3.

I forgot to say that the version of debian 10 I tested was using XFCE. I will try with Gnome so maybe the error is related to that,

So, what can I do?

@hansolo Yes, thank you! What I hope to do is to determine that any action we take affects only the Librems with a US keyboard. We're going to have to determine the best fix for the US keyboard and then deploy.

This issue is caused because g-i-s follows RHEL/Fedora useradd rules which are different from debian. Fedora/RHEL allows usernames with upper case letters.
And g-i-s provides no error message when a PureOS user tries to create a username that does not comply with the useradd rules followed in PureOS.

I have a German keyboard layout, can I help?

It's unfortunately one with a US keyboard :-/

@mak Do you have a German keyboard now? Or is your L13 v4 US keyboard?

I'm going turn this task into a milestone so we can track the release period.

The OEM installer converts a user Full Name into a Username in lower case letters, but it is possible for a user to manually edit the username and add a upper case latter again. Like in the picture bellow.

Both those mentions in the logs at boot time are "normal".

Apparently an upstream bug: https://github.com/systemd/systemd/issues/12401

Thanks for reporting!

After several days and upgrades, it seems, that unattended-upgrades got removed automatically, so I don't bother anymore :)

The versions in Amber and Byzantium are not far apart; https://software.pureos.net/search_pkg?term=libccid
In Debian's Salsa it appears they're about 88 commits apart: https://salsa.debian.org/rousseau/CCID/-/compare/ccid-1.4.30...ccid-1.4.32