WIth rootfs, you can use Method 1 or Method 3. Method 2 is for the ddimg.
Pudge
I used Method three to prepare a btrfs SSD which then booted an RPi 4b and ran calamares. Smooth and error-free.
If you happen to use that method please give feedback on readablilty, if it’s understandable, and if it doesn’t work, etc.
The sentence at the end (“After installation, Post-Install after Method 2”) threw me completely.
I thought it was telling me something extra I had to do, so I read and re-read Method two but couldn’t find anything. Eventually concluded that this sentence is a sort-of pre-sub-header for the following header (“After Image is installed”).
So (unless I’ve misunderstood) I’d suggest dropping that sentence to avoid confusion.
Also, for newbies like me, I’d suggest that the page should make clear that this is always a two-step process (1:Prepare bootable image on uSD/SSD, 2:Insert that into other device, boot and run calamares), but for step 1 you can choose one of the three methods. And newbies like me think that ‘installing’ is what you do with the ISO (ie step 2), so using the word ‘install’ to describe step 1 is potentially confusing.
Replying to myself:
One thing I noticed is that this setup seems to idle at 2-3°C cooler than an almost identical setup on another RPi4, the main difference being that this setup uses ext4 and the other (from the old image) uses btrfs.
So I look forward to testing the rootfs image, to see whether the new image also runs cooler under btrfs.
I have now tested installations using ext4 (via Method two) and using btrfs (via Method three).
Both on same hardware, both using Cinnamon DE, but with the btrfs install I added snapper, snap-pac and btrfs-assistant.
I ran a simple script which prints current time + cpu temperature every 12 seconds, left the machine doing nothing else for a few hours, and then looked at the readings for the last 10 minutes.
- With ext4, the temperature readings varied between 30°C and 35°C.
- With btrfs, the temperature readings varied between 35°C and 40°C.
Nothing alarming, but I’m surprised there’s such a noticeable difference.
Assuming that temperature correlates to cpu activity, it looks like I’m paying a (small but permanent) price in cpu cycles if I choose btrfs, which is a bit concerning for an always-on NAS server.
A few months ago I tried with method 1, and btrfs didn’t work, only ext4.
I do not own a PineBook Pro laptop, so I have no way to test the new pbp images.
There is only one sure way to find out, try it and see what happens.
Pudge