Demi Marie Obenour <demiobenour@gmail.com> writes:
On 11/13/25 11:49, Alyssa Ross wrote:
Demi Marie Obenour <demiobenour@gmail.com> writes:
There is now a way to update the OS, so the previous documentation is now stale!
Signed-off-by: Demi Marie Obenour <demiobenour@gmail.com> --- Documentation/installation/index.adoc | 3 ++- Documentation/using-spectrum/index.adoc | 2 ++ Documentation/using-spectrum/updates.adoc | 29 +++++++++++++++++++++++++++++ 3 files changed, 33 insertions(+), 1 deletion(-)
diff --git a/Documentation/installation/index.adoc b/Documentation/installation/index.adoc index d67c88dda062066c19c3b21e699f074cc18a6dbc..536c3dd9f78faa2ecad4127dc9ccc2058a230b1a 100644 --- a/Documentation/installation/index.adoc +++ b/Documentation/installation/index.adoc @@ -18,6 +18,7 @@ development.
== Uninstalling and Updating
-Currently, there is no implementation for a software update. +See xref:../using-spectrum/updates.adoc[Updating the OS] for how to enable +updates.
Let's phrase this so it says that there's work going on to enable updates but it's not all set up yet. User-focused documentation shouldn't really be suggesting that people will have to build their own images and run their own update servers.
Would it be okay to mention that it is WIP, and also add a link to the build configuration options for those who *have* built their own images? That will continue to be relevant even after official binary releases are available. Developers are users too, and they might be a bit confused when their image either doesn't update at all or updates to an official build without any of their changes.
I think it would make more sense to cover that in the section about building your own images, because ideally you find this out just before you build your image, whereas here you might end up only finding it after and have to rebuild.
+Updates are atomic and take effect after the system reboots. +If the system is rebooted, crashes, or loses power during an +update, the update will automatically be rolled back. Updates
Is this currently true?
If not, that's a systemd-sysupdate bug.
I thought you took out the boot counting stuff?