Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines
Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.
REPEAT: Do NOT report bugs for outdated packages!
https://wiki.archlinux.org/title/Bug_reporting_guidelines
Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.
REPEAT: Do NOT report bugs for outdated packages!
FS#79902 - [fwupd] Build without passim dependency
Attached to Project:
Arch Linux
Opened by Jake Dane (jakedane) - Tuesday, 10 October 2023, 06:40 GMT
Last edited by freswa (frederik) - Tuesday, 10 October 2023, 21:42 GMT
Opened by Jake Dane (jakedane) - Tuesday, 10 October 2023, 06:40 GMT
Last edited by freswa (frederik) - Tuesday, 10 October 2023, 21:42 GMT
|
DetailsDescription:
Last month fwupd 1.9.5-1 added a dependency on passim, an unauthenticated, unsigned mDNS caching service on the local network. The use case for passim is "thousands of users" [1] on the current network needing the same firmware file, so they can fetch it locally instead of from a CDN. Is a significant percentage of Arch Linux systems being operated on a local network with thousands of users that have the same hardware? I don't think so. fwupd can be built without the passim dependency and I suggest it should. [1] https://github.com/hughsie/passim#introduction Additional info: * package version(s): 1.9.6-1 Steps to reproduce: - Install fwupd, passim gets installed as well |
This task depends upon
FS#79614. In general, Arch tries to enable most features. If no interest from PM, this request will be declined ASAP.> The use case for passim is "thousands of users"
Incorrect. Please don't latch on to a single phrase and twist it by omitting the most important word i.e. "perhaps"
> have the same hardware
Are you sure? It seems to be small metadata file that *everyone* downloads irrespective of hardware.
@freswa, it seems this issue is not going away :( Any chance you will reconsider splitting out the passim libs as it seems to be the "least friction" solution? I know we don't do that lightly, but there are some exceptions i.e. systemd-libs, libpipewire, boost-libs, etc. See also [1] for some choice ranting...
I personally don't care either way, so feel free to whack this on the head :)
[1] https://bbs.archlinux.org/viewtopic.php?id=289297