Debian 11 “bullseye” is released: Debian -- News -- Debian 11 "bullseye" released
When is a good time to upgrade my Vero 4K + to bullseye?
The official repo http://apt.osmc.tv still has no bullseye in dists.
Hi,
There will be an announcement on Blog - OSMC and in the news section of the Discourse.
Thanks Tom.
Thanks, but that doesn’t answer all questions.
-
Is there an ETA? Days, weeks, months?
-
Is it possible/good to upgrade Debian now and upgrade the apt.osmc.tv packages later?
No such ETA exists but the last range is the most likely
Possible yes, but will lead to issues and problems
You may get an operating system but you will break the packages we build against Buster, including Kodi.
1 Like
Out of curiosity - what is your use case / need to update to Bullseye?
When we officially support it.
I upgraded my NAS to bullseye and I now am having issues with Samba due to to the different versions between OSMC and my NAS.
Also, I’d like to finally use DNS over TLS in systemd-resolved, but that needs Version 247.
There are many other new features like native exFAT in Kernel.
You yourself have mentioned issues being solved with bullseye, f.e. [TESTING] Kodi v19 builds for Raspberry Pi 2/3/4 - #6 by sam_nazarko
But really, software is a process, not a state. Features and fixes are ever-evolving. Debian stable releases are solid and… stable. Would be really great to enjoy this new release, too.
Can you provide a rough estimate for an ETA?
Thanks!
There shouldn’t be any issues that could not be overcome by respective SMB config on the client side
I have taken this opportunity to retire Samba and move to NFS.
But I feel the issue should not be my reasons for wanting to upgrade, but to get some information about the roadmap, that’s public and findable. I am probably not the last person asking this question.
I understand it’s probably a lot of work, else you could provide an ETA of days or weeks. But out of curiosity: What is all that work? The repo and builds are probably relatively straight-forward. Obviously there’s a level of testing (which I dont know). Am I missing anything else? Do you need to write new code? Is there anything we can do to help?
I don’t want to be pushy, apologies if it comes across like that. I am just interested and excited. Afterall, one of the reasons I bought my Vero 4K + is the underlying Debian. 
No you won’t be. No offence, but it’s quite tiresome to be asked this question again and again with every new kernel/kodi/RPi/Debian version.
OSMC relies on volunteers who are free to dedicate as much of their free time as they can spare. That’s the nature of Open Source, and why it can be futile to set deadlines. Meanwhile, as a business selling hardware, (without which OSMC would not exist for any platform) the focus has to be on stability rather than the latest and greatest.
Bullseye will come, don’t worry.
Just my opinion. If I’ve got that wrong, Sam can correct me.
6 Likes
We ship our own kernel - not upstream. So no benefit there.
Stretch to Buster took me over a year. No ETA for Bullseye. If anyone wants to lend a hand, I’d be much obliged.
I’ll start looking at Bullseye when the dust settles on v19.
3 Likes
Thanks for the reply. Is there any way we can help? What are the tasks that need to be done?
Thanks for all your work on this project!
Also, I’d like to finally use DNS over TLS in systemd-resolved, but that needs Version 247.
I am using DoT on my OSMC raspberry via kresd:
cat /etc/knot-resolver/kresd.conf
-- Switch to unprivileged user --
user('knot-resolver','knot-resolver')
-- Unprivileged
cache.size = 1*GB
-- Load modules
modules = { 'policy' }
-- forward to mdosch.de
policy.add(policy.all(policy.TLS_FORWARD({
{ '5.181.50.75', hostname='mdosch.de' },
{ '2a03:4000:3f:1c8:8839:33ff:feba:fe4a', hostname='mdosch.de' },
})))
I just disabled whatever OSMC was using by default as resolver, installed kresd, enabled it and put nameserver 127.0.0.1 in /etc/resolv.conf. Works fine.
Thanks, I am doing this, too.
But this post is not about my reasons for wanting to upgrade.
It is to find out about the timeline and steps when the upgrade can happen.
Indeed, but I didn’t know whether you knew about alternatives to get the DoT working.
I’m also interested in the upgrade to bullseye as 3/4 of my debian stable machines are upgraded to bullseye already (my main laptop is always using testing) and only the osmc raspberry is missing.
What are you specifically looking for from Bullseye?
Is it just because it’s newer?
If I can get a compelling reason, I’ll bring it up on my list of priorities.
Our focus is always on stability first.
As mentioned in Debian bullseye - #9 by ilf, for me a key new feature is systemd v247 with verifying DNS-over-TLS in systemd-resolved.
As mentioned in Debian bullseye - #11 by ilf, for me the Debian on Vero 4K + is not just nice and familiar, but makes the Vero 4K + a true general purpose computer that also happens to ship with a pre-installed Kodi. I run many other services on my Vero 4K + too, Debian enables that.
I would love to upgrade this other software I use for new features or stipped legacy (GCC 8.3 → 10.2, OpenSSH 7.9 → 8.4 with new default crypto, Python 3.7 → 3.9, completely removed Python 2, etc.)
Debian bullseye is stable. I was not yet able to find the differences between Debian upstream and Debian OSMC to understand the differences between them and thus the work required to adjust that delta to the new release.
I take it OSMC ships patches for kernel and driver(s). Are they complicated to rebase against the new versions?
What other work needs to be done, besides testing?
Re-compiling the whole OS against the new GCC, libc, verifying kernels are modern enough, patching Debian for any changes introduced by systemd changes, downstream package changes.
We ship our own kernel so no rebase is necessary here fortunately.
But userland changes will be extensive and invasive.
Feel free to build and upgrade yourself, but we will move when we are happy and ready with a new Debian.
You could always build your own systemd on Buster and use that if this is a priority feature.