Please add your agenda items to our pad for the meeting minutes, so we know beforehand if there is anything to discuss (some people decide their attendance on that).
Side note: Oct 3rd is a public holiday in Germany. So it is very likely that many members of the community wonāt be joining. Letās see if we can gather enough contributors regardless.
Please add your agenda items to our pad for the meeting minutes, so we know beforehand if there is anything to discuss (some people decide their attendance on that).
Please add your agenda items to our pad for the meeting minutes, so we know beforehand if there is anything to discuss (some people decide their attendance on that).
Teufelchen: We could stick with the status quo. If we are fair, there should be nothing to hide
maribu: One of the more controversal maintainer proposals resulted into a rather unactive maintainer than soon after dropped out of the role. Maybe he did not feel welcome to actually engage.
Ann: Other point of view: People may censor themselves from raising issue if they know it is public
Martine: One example that would clearly be something that we should not make public is raising a CoC violation against proposed maintainer
Teufelchen: Letās revisit next week
Renaming ads101x driver
Teufelchen: (PR author) Started work, and later found something is in, but name didnāt match.
Teufelchen: Propose doing the change (breaking change but corrent and rewarding contribution)
Ann: I heard the driver is currently not working?
No itās working
crasbe: Letās be brave.
(Some nodding)
crasbe: Letās not pressure contributor to throw things overboard now.
maribu: Itās an API change thatās not hard to fix (sed).
Martine: Pseudomodule for name change?
maribu: Would also need set of headers for function calls.
chrysn: Would this even be well tested?
maribu: Lot of work for 5 users.
Koen: So the only remaining question is hard change vs. compatibility change?
Teufelchen: Just break, not worth the attention.
(Most nodding)
(Only remaining discuss from Josh from Matrix discussions: āno benefit from renamingā)
Upcoming Release
Ben under pressure, maribu will help.
Suggested release name āBurning Dumpster Fireā (also for API breaking change)
FOSDEM Booth (shared with Ariel OS)
Ariel folks want to collaborate. Is there RIOT interest?
maribu: related to devroom?
Koen: Not mutually exclusive if all OSes are in.
Teufelchen: Would be great
maribu: positive, but not sure if there (thus not sure if can contribute). Ben will be there. Not sure if Iām allowed to volunteer Ben, though.
Elena: Thinking about going there,
Teufelchen: Can we commit to it?
AOB
CI is happy with crasbeās PR 21800 that reduces code by 160k LoC.
Please add your agenda items to our pad for the meeting minutes, so we know beforehand if there is anything to discuss (some people decide their attendance on that).
Mikolai: Lot of high profile changes, it would be valuable to check for regressions without skipping release tests
Maintainer Proposal Threads on the Forum
Martine: Josh proposed to delete all threads
Martine: I donāt see why one would do that
Martine: regarding concrete case with maintainer proposal, that was in times of mailing list, so probably not the reason for them not staying active
maribu: but isnāt it enough that it may cause friction?
Teufelchen: might repeat myself: should leave it public, we shouldnāt trash-talk anyway (keeping it public is a way of ensuring that we behave nicely)
Martine: we wouldnāt make them maintainer anyways
AnnsAnn: Iād second Enochs opinion. probably very hypothetical, but I would not dare to openly criticize in that case
Martine: maybe, if there is valid criticism, double-check with other maintainers in private, then openly state it in the maintainer thread
maribu: donāt see advantage of keeping it at all
Martine: in favor of archiving
AnnsAnn: would be okay with unlisting
Martine: consensus: weāll unlist maintainer proposal threads right after sending out the invitation
Was discussed last week: A bit inconclusive but tending towards āletās rename itā but Teufelchen should checkin with Josh D. beforehand.
Teufelchen did checkin with Josh D.:
āthere is a reason I donāt want a renameā ā The reason is that if you check the git history of the driver, you will see that it always supported both families, and always under the current name. I presume that there are users of the driver and feel more obligated to support those users who initially created the driver than anybody else.
Then again, if there are really only 5 users (and maybe they care far less than I do), maybe the question reason is more ideological than practical.
Teufelchen: letās go for breaking change (rename), but only after hard-freeze
(some nodding in the room)
FOSDEM
HAW Hamburg gang (Teufelchen & AnnsAnn) will join
Benpicco has been on every FOSDEM so farā¦
Mikolai thinks heās going to be there
Koen likely and maybe more ARIEL OS devs.
Maribu: we should probably prepare talks for devroom
RIOT-Assembly @ 39c3
Teufelchen: I donāt think I would want an assembly, I am going to be bussy with other stuff
Teufelchen: But we should drop a lot of sticker!..
Teufelchen: ā¦and I would love to share a Mate in a tiny āMaintainer Assemblyā for an hour or two!
Teufelchen: I am going to be there, already got a ticket
Martine: want to be there, but depends on ticket or talk acceptance (also Coffeebots @ CfPunk)
Mikolai wants to be there, doesnāt have a ticket yet
Elena also wants to be there
Martine: given we are so many, we should register an assembly
Martine: assembly can be pretty low-key, no workshops etc needed, just a āhomebaseā for all RIOTers
Mikolai: when?
Martine: Call for assembly not out yet
AOB
Zephyr and RIOT are almost like Windows and Linux for the embedded
Just wanted to quickly mention before this spreads as misinformation as it might not be as clear in the notes, that this was an assumption from a very basic flyover of the code based on the included test using strings for all values, including integers/floats. Upon closer inspection the implementation does have CBOR support as far as I can tell! Just to not discredit any implementation based on this point, given that both are good PRs with a lot of care put into them.
2025-10-31T14:00:00Z we will have our next weekly coordination meeting.
Since it is a public holiday in some German states and I will be boarding a plane to Montreal around that time, @bergzand was kind anough to take over the moderation for that meeting.
Please add your agenda items to our pad for the meeting minutes, so we know beforehand if there is anything to discuss (some people decide their attendance on that).
Please add your agenda items to our pad for the meeting minutes, so we know beforehand if there is anything to discuss (some people decide their attendance on that).
for mguetschow: some tinydtls/crypto/lwip issues open, maybe fixed by recent changes?
crasbe: if you need procrastination topics, suggest some of the issues, some issues might not be relevant anymore
mikolai: will look into it when I have time
crasbe: we recently closed really old issue, thank you benpicco, RTC issue / alarm caused drift/variation, issue was over 7.5 years ago
RIOT at 39C3
mikolai: question who will be there, would it make sense to get a table or new shared table system that works more ad-hoc as elena suggested
(Agreement towards that idea)
elena: shared table might make more sense since we often are not at the table
mikolai: will more people from hamburg join?
ann: Teufelchen but busy with own table, Lasse potentially, unsure rest, (was asked to ask people)
How to proceed with Runtime Configuration PRs
Tom: Lasse rebased his PR to current master, we tried to contact Fabian to hear from him, didnāt succeed
Fabian: Sorry, but my Matrix account is not active
Ann: Question how you stand on your PR, see potential for working together, getting it merged, etc?
Fabian: Iām not insisting on my PR and see some advantages e.g. in doc and testing in Lasseās PR. I would be fine if Lasseās PR gets merged.
Mikolai: Could you take another look into Lasseās PR? Maybe integrate your perspective and use case?
Fabian: Currently no time for that planned. Need to get back to my boss to discuss priorities
Mikolai: Would be excellent if someone with a use case and understanding of the background would take a look
Tom: What is still outstanding is splitting the PR into 2-3 PRs. This is currently possible, as the PR is already relatively good structures. But all in one PR is overwhelming to review.
mikolai: found reason for weird bug, different modules of network stack called from different threads, consistency problems with prefixes for debug and other syntax, would it make sense to add QoL improvements to debug. There was one issue in the past but went stale.
karl: there is a problem where enabling debugging globally can cause the system to overload/cant catch up with the amount of debug logs
benpicco: it makes sense to have debug, lowers the barrier, log can be complicated/rarely used
mikolai agrees
karl: would it make sense to add prefixes to debug?
(General agreement that it makes sense to put some time into improving it)
Release
mikolai: release, anything we can help with, how is progress?
benpicco: release tess are running, looking good, everything parses
mikolai: any deadline?
benpicco: all chill, will work out
Tiny/Smallbuild
crasbe: @maribu wanna look at smallbuild / tinybuild docker PR, now works well, 20GB for normal, 1-2GB for tinybuild
maribu: will try this weekend
mikolai: so if I need both, I need even more storage?
crasbe: yes but download and build is far quicker
mikolai: what about version pinnings, are they the same?
crasbe/maribu: no, far newer, better diagnostics, would most likely detect bugs, positive but might be annoying
crasbe: migration of CI would be a different subject in future, RFC if we want this by default, finding out which build image you need can be annoying, not a docker wizard though
Please add your agenda items to our pad for the meeting minutes, so we know beforehand if there is anything to discuss (some people decide their attendance on that).
