Igor

From Armbian perspective, application image is minimal + official way of installing things. This is how we do it: https://github.com/armbian/os/blob/main/userpatches/extensions/omv.sh This is what they do: https://github.com/OpenMediaVault-Plugin-Developers/installScript/blob/master/preinstall https://github.com/OpenMediaVault-Plugin-Developers/installScript/blob/master/install Some things in this long config are doubled, some deprecated, some makes no sense. I don't think we need to deal with all that. If you can access OMV via WEB interface and if you can upgrade it, that's about it.

That was quick fix to secure compiling. Checking all functions on all devices for each kernel upgrade is impossible, so I don't even try. Automated testing is primitive and only catches fatal problems, while now not even that as its currently broken. And this device is not a part of this. I only have one of such device, probably from pre-production series. Try to use a device-tree from a kernel that works - on 6.6 and 6.11. Then lets diff the code.

This is highly expected. We don't even try to support EDGE kernels at end user level. This kernel is EOL and will be replaced with a v6.12.y in next year Q1 release. But we can't secure you that this problem will be fixed as support budget is negative (we have some support for other boards from Radxa) and we, nobody in the team, can't afford to spent days or weeks resolving this. However, you are welcome to spent at least time and join debugging this - problems is as yours as it is ours. The reasons why we keep build framework in a good state is that you can help resolving problems https://github.com/armbian/build We are and will always be too small to face this mountain of problems that are generated by everyone, compensated by no-one. Not by us, but all community that is poking kernels, and hardware dealers that are flooding this market.

When you cloned our repo, you got 10 years of specific work on common work and that was contributed by 500+ people. Most of people are resolving their specific problems, yes, but we all try to do it the way that is reusable, that person behind has less work. We also got a lot of things prepared. Doing things universal should always be considered. Also because you might get help from someone. Which has absolutely same problems, but is having hardware from another vendor. In our world, SoC defines how things work, not the one that put SoC on PCB and name / sell it. This means that most of the drivers are the same, especially this area is shared - low level communication protocols such as GPIO, I2C, SPI. This is always shared among boards with the same SoC. This goes further. It is shared among same families and we also already have some common generic top level API / libraries that can be used. https://docs.kernel.org/driver-api/gpio/consumer.html (but here my understanding ends, I am not updated with the state of this) Yes, it has no point poking into. WiringPi library is a dead project for many years and its no point dealing with this. What vendors do here is, trying to be "as Rpi as possible", so they are doing their quick and dirty assemblies of that, always with partial functionality and absolute absence of maintenance. Their goal is to sell, regardless what is the quality of software. Is that your goal? I doubt. I am not an expert in this field, so those are general tips, but there are people on and around this forum that knows this stuff well, as they developed those libraries. I am just giving you some ideas / tips. Decision what you will do with your time is yours. Another one that came into my mind, while replying: https://github.com/eclipse/mraa

@going Probably @wwortel means to develop a minimal boot image which would download and dd image to eMMC. I agree that would be nice to have, but develop that requires time / resources we don't have. If you look into the script that installs OS to eMMC (and other media) its pretty complex and hard to maintain https://github.com/armbian/build/blob/main/packages/bsp/common/usr/bin/armbian-install Not to mention that people doesn't pile up to help us around that. There is / always was just a few people that sacrifices their precious family time to provide you current experience. Where we can always find something to fix and improve.

New version (v24.11) with even more targets has been released. Home assistant takes few minutes and might require reboot at first install or there will be some errors related to privileges. This gets away after waiting few minutes and reboot. Testing was done on Odroid M1. Latest: https://github.com/armbian/distribution/releases/tag/24.11.1 Previous: https://github.com/armbian/distribution/releases/tag/24.8.4

We are cleaning those builds as they have very little value. But you were lucky: https://github.com/armbian/community/releases/tag/24.11.0-trunk.351

Take a look at universal projects that are made far better, and are universal by design. https://forum.armbian.com/forum/40-reviews-tutorials-hardware-hacks/ (pinned topics)

They are not needed. We use flash media friendly and more advanced ZSWAP solution by default. https://www.google.com/search?q=zswap If you don't want to have this, you can of course disable this feature in /etc/defaults/ (will be soon added to armbian-config) and add swap file. But that is not recommended.

Try here: https://rsync.armbian.com/oldarchive/ We can't possibly maintain this further, sorry.

Running Debian bullseye? This is EOL. Edit: Aha, yes, that's it. https://unix.stackexchange.com/questions/669004/zst-compression-not-supported-by-apt-dpkg

Yes, we should probably add REPLACE: neofetch to the control file. Well, there are several reasons for this: - we are patching neofetch on our own to add Armbian branding. As Neofetch is not maintained anymore, we can't add Armbian normal / official ways by opening a PR at their GH - we would like to switch to fastfetch but this has to be resolved first https://armbian.atlassian.net/browse/AR-2336 (we have extreme lack of hands since ever as we focus into kernel space hard problems, so there is no way to proceed on this) - during the switch to new generation of armbian-config there are many things to tackle and this (known problem) simply live somewhere in backlog This is how this transition is made and we could probably remove this double packaging to make it easier. We already pack it at Armbian config repo https://github.com/armbian/configng https://github.com/armbian/build/blob/main/lib/functions/compilation/packages/armbian-config-deb.sh

Test most recent builds: (still in testing) https://rsync.armbian.com/incoming/schwar3kat/rock-5c/archive/

I think it is better to expand script to support those scenarios. https://github.com/armbian/build/blob/main/packages/bsp/common/usr/sbin/armbian-install Worse case, write things here: https://docs.armbian.com/User-Guide_Getting-Started/#how-to-install

Headers are released via repository and that is done when images are tested. In about a week.

Great job! Interested for integration into main config tool? https://github.com/armbian/configng Now this tools is properly made. https://docs.armbian.com/User-Guide_Armbian-Software/#adding-example

Can be found in https://github.com/armbian/community

REQUIRES Armbian 24.11 (currently availability is limited to rolling release or self build image) Armbian config is a configuration utility for customising and automating tasks within Armbian environments. These scripts help streamline setup processes for various configurations and use cases. It helps end-user to setup their network, ssh, configure hardware features and provides simple install and uninstall methods for your software packages. armbian-config Documentation: Adding a new feature or software install title User guides: system features software installer Please open bug reports or feature requests directly on GitHub: https://github.com/armbian/configng * valid for new version of tool that comes from Armbian v24.11 onwards

The problem you are experiencing is probably related to some incompatibility when assembling 32bit OMV. We can't test all assemblies manually (we have no other option) on all targets. Perhaps moving installer to run-time is better? https://docs.armbian.com/User-Guide_Armbian-Software/#adding-example We will eventually add this, but help is appreciated.

I am slowly working on a script, but currently doesn't provide expected results: https://github.com/armbian/configng/blob/main/tools/modules/system/release_upgrade.sh Also it will remain as experimental as this goes into upstream bugs which are beyond our control or resources to fix after thousands of people that are behind all those packages. Where single one can cause this upgrade to break ...

Also documentation (related to this topic) is now in right state of mind: https://docs.armbian.com/User-Guide_Armbian-Software/#adding-example It is hard to add software title we never tried, installed and / or know nothing about, then expert user of this software adding it to (armbian-config) installer. We focused to make this task as simple as possible, to encourage you doing that and not opening us tasks (we already have 1000 x too many and won't ever complete). We added popular titles to the system, which have some value for us, while everything else is mocked by more important tasks in other sections. But mechanism is there - use it, its user / developer friendly. Also its a nice and easy way to contribute to the open source.

Perfect example to explain how anyone can fix such trivial problems without involving me. Which is essential for the project to survive. There are and always will be way too many problems for too little maintainers ... Bug is here, yes, and can be easily fixed, especially if you already have a solution. I am happier to provide guidance then doing this on my own (which takes a lot less efforts). Solution is adding a package here: https://github.com/armbian/build/blob/main/config/desktop/jammy/environments/gnome/config_base/packages (jammy has same packages as noble and is symlinked). Every contribution counts, even this banal. We have different Chromium versions, patched for this and that, then with and without snap ... It is a surprise that we can provide a desktop where accelerated video playback works on Rockchip RK35xx platform, which we support better then Rpi. Where all this probably just works if you download any Chromium?

If this was not mentioned in a release log https://docs.armbian.com/Release_Changelog/ and since you already asked, there is no need and you should not ask again. https://forum.armbian.com/store/category/2-support/ and I will try to hire someone for you or fix on your own. This is your problem, not ours. Buy some other hardware. it is cheaper then fixing this. I hate being an ass-hole, but If we don't stand for ourselves, nobody will, certainly not constantly unhappy and desperate free-software-riders users. Remember this rule once and forever: applying a pressure to open source developers is worse thing you can do here. There are thousands of other problems when we can only address one. And you cover 0.5% for that one: https://github.com/sponsors/armbian Do the math. There is build framework https://github.com/armbian/build download it, dig into the code and fix it. Then open a pull requests. As you can see, we are physically, emotionally and financially unable to support your (endless) greed for software perfection. Don't take this personally as its not. Most of people on cheap hardware, that fully relies on exploiting open source developers, acts similar. We develop and maintain open source software for fun and we chose what that is! With little time that is available. Average resolution time for bugs in open source software is 5-7 years and there is nothing we can do. If you scream, complain or emotionally blackmail us. Closing. @Werner Feel free to re-open, once this kernel will be supported, probably Q1 - Q2 / 2025. Only confirmation that "it works" is expected to be added here, we don't need yet another "still doesn't work" ... ever.

If you would like to use 2 years old kernel with not a single fix applied after initial release, use version of Armbian from Orangepi site. This question was answered on Orangepi download pages. For everyone. To spare answering same questions over and over again. Which is devastating for mental health and expensive when we pay everything and you nothing. Only when you like - donation.