Bug 2350301
Summary: | Upstream or namespace PLATFORM_ID | ||
---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | Lennart Poettering <mzeuom> |
Component: | fedora-release | Assignee: | Zbigniew Jędrzejewski-Szmek <zbyszek> |
Status: | ASSIGNED --- | QA Contact: | Fedora Extras Quality Assurance <extras-qa> |
Severity: | unspecified | Docs Contact: | |
Priority: | unspecified | ||
Version: | rawhide | CC: | alciregi, fedoraproject, kevin, mboddu, pbrobinson, samyak.jn11, sgallagh, thrcka, zbyszek |
Target Milestone: | --- | ||
Target Release: | --- | ||
Hardware: | Unspecified | ||
OS: | Unspecified | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | --- | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | Type: | Bug | |
Regression: | --- | Mount Type: | --- |
Documentation: | --- | CRM: | |
Verified Versions: | Category: | --- | |
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
Cloudforms Team: | --- | Target Upstream Version: | |
Embargoed: |
Description
Lennart Poettering
2025-03-06 11:27:12 UTC
The PLATFORM_ID field is pretty much vestigial at this point in Fedora. It was introduced during the Modularity days and it provided the identifier for the virtual "platform module" that other modules could depend on. I believe it's still in use in RHEL, but it's probably safe at this point for us to remove it from Fedora. oh, ok, removing sounds great too! I did a little further digging here and we're in kind of an awkward position. Technically, Fedora has retired Modularity and its module build service (MBS), meaning we don't ship any modules and cannot build or update any that we used to ship. This happened in Fedora 39, which SHOULD mean that anything that used to be a module is long since migrated. HOWEVER, it turns out that DNF (both 4 and 5) retain support for Modularity (largely for RHEL reasons) and it's possible that there are third-party modules out there that rely on PLATFORM_ID which is, as you note, a non-standard extension to os-release. Since PLATFORM_ID was effectively a published API for Modularity AND we have no reference implementation for module-building, we don't have an easy way to rename the option. I also don't see any value in trying to get PLATFORM_ID standardized upstream at this point, given that it's a dead technology. Would it be sufficient for us to just stick a comment in os-release indicating that the option is a non-standard extension that is preserved for the moment for historical reasons, but it should not be relied upon for any purpose? Could we also do a change for F-43 to remove it so it's an advertised removal, and maybe suggest a replacement identifier if people need something, and it just goes away? (In reply to Peter Robinson from comment #4) > Could we also do a change for F-43 to remove it so it's an advertised > removal, and maybe suggest a replacement identifier if people need > something, and it just goes away? Changing or removing it has knock-on effects with DNF that I don't have the cycles to address in the immediate future. If someone wants to drive that effort, I absolutely support it. I just can't commit to it. I think it'd make sense to file a Change proposal for this. https://sourcegraph.com/search?q=context:global+%27%22PLATFORM_ID%22%27&patternType=regexp&case=yes&sm=0 suggests that it's used for el8 and el9 by bootc, possibly by container-inspector. (It's hard to say if it's used, or just shown in samples…) But that's probably enough usage to not just drop it immediately. |