How can I verify the format of my local mix? | mixer versions update

[mvincerx]

How can I verify the format of my local mix?

I want to run a mixer versions update to increment my mix (and incorporate new/edited bundles). Suppose I'm using CL 29740 (format 27).

Recreate the problem:

Set up the mix:
mkdir mixer4
cd mixer4
mixer init --upstream-version=29740
[... swupd bundle add ...]
[... swupd bundle remove...]
sudo mixer build update --native --clean

When I run
sudo mixer versions update

Results show

• "Old upstream 29740 (format 29)"
• "New upstream 29740 (format 29)

Question: Why doesn't the "Old upstream", above, show "(format 27) which is the format to which 29740 belongs? Or do I not understand "mixer versions update" command?**

Then I run:
mixer versions

Results show

• "Current mix: 20" as expected
• "Current upstream: 29740 (format 27) as expected

I run
cat mixer.state
Results show
[Mix]
FORMAT = "29"
PREVIOUS_MIX_VERSION = "0"

This appears to suggest that my mix, starting with 29740 (format 27), pulls from mixer.state (or my OS) instead of from upstream format associated with 29740.

[mvincerx]

[rchiossi]

I had to agree that this information is confusing and, but the behavior is correct.

When you create a mix based on an upstream version (in this case, 29740), you are using the content of that version, not the update metadata of that version.
What this means is that you will download the rpm packages that version 29740 uses and create your own bundles and manifests using those rpms. Since the format is a property of the manifests, not the content, you can create a manifest in any format you wish for that given content. The default is to use the format of the system you are running mixer on. That is why in the first command it shows Old upstream 29740 (format 29) - mix content copied from upstream 29740 with manifests in format 29.

Now, when you run mixer versions, instead of the format of your mix, what is printed is the format of that upstream release, so 27 is printed instead.

I'll mark this issues as enhancement since our current output is very misleading.

[mvincerx]

Thank you very much! @bktan8 FYI.