– DRAFT –
Social Web Incubator CG

<omz13> hello all

<KevinMarks> is jitsi waiting for the top of the hour to start?

<dshanske> KevinMarks: Apparently not

<bengo> hi folks

people who haven't officially joined the CG can do so here: https://www.w3.org/community/socialcg/

<dshanske> I am present for the first time. Haven't been available previously for this time.

<tantek> bumblefudge_++ for scribing

<Loqi_> bumblefudge_ has 2 karma over the last year

dmitriz: timeboxing would be good to get through agenda

decision-making policy proposal

capjamesg: let's get into the socialCG charter/process topics raised by bengo first (slight variation from the emailed agenda)

<dshanske> (Is the Jitsi supposed to have started?)

<capjamesg> dshanske Yes. Are you not in the call?

bengo: there's been a lot of text in various channels but the proposal itself is fairly simple, i'll summarize

<dshanske> capjamesg: It says Asking to Join meeting...but I just saw a Tantek message. I'll try restarting

<capjamesg> tantek https://lists.w3.org/Archives/Public/public-swicg/2023Sep/0118.html

bengo: the original email is here: https://lists.w3.org/Archives/Public/public-swicg/2023Sep/0118.html
… (proceeds to summarize rapidly)
… i also summarized the feedback on list in this digest: https://lists.w3.org/Archives/Public/public-swicg/2023Oct/0024.html

<dshanske> I have people...but no audio yet....but beeping noises. So assume it is working.

evan: how do you want to structure the discussion?

bengo: you can speak to any point you like, i don't have much agenda except wanting to move to consensus

evan: my concern is that this proposal frames itself as tabula rasa, while i think 5 years have passed on process defined at meetings and in minutes

<tantek> +1 evan

evan: i take the point that async decisions and inputs have been lacking, but this feels like a challenge to status quo

<tantek> does "every single minute" "every single resolution" mean of the CG or also of the WG?

bengo: but i read through all the minutes i could find and not finding much in the way of process-defining group decisions

evan: ok i can try to recap the history a bit
… : the CG was spun up when the WG spun down, probably without enough explicit/formal documentation of process
… maybe some of the rules have been unspecified or implicit/norm-based, and i think the general idea of this proposal is good, i just

<snarfed> this can't be the first time a CG needed a decision making process. why are we trying to invent one ourselves?

<dmitriz> @snarfed - definitely. and we're not reinventing; bengo's proposal reflects basically every other cg/wg policy

evan: want accepting this document NOT be taken as the first or only process of the group

bengo: actually in the minutes, i only see an explicit chair decision to NOT have a charter

<snarfed> @dmitriz good to know, thx

evan: but group decisions have been taken by +1/-1 for 5 years and that seems to be working, even without an explicit agreement

<bengo> james audio odd?

<bengo> refresh helps sometimes

evan: mailing list was disabled for 5 years, starting minutes after the CG was formed

tantek: the mailing list was turned off deliberately because we decided in that meeting it was counterproductive to the workmode of the CG
… namely, focusing on git issues and archived
… IRC channels for async and smaller side meetings

<capjamesg> Thanks bengo!

<capjamesg> I have disconnected my headphones which should help.

tantek: i don't think most of the messages since we turned the list back on are useful for moving this work forward
… I believe any kind of process needs to be centered around editor productivity
… and we should optimize for what the existing and past editors want

<snarfed> I assume "spec maintenance" doesn't include normative changes? is a CG allowed to make those...?

tantek: I think our choice of channels and media should protect mental health of editors
… e.g. monitoring lots of channels just distracts and burns out editors
… i think much of the list is currently distracting and harmful and hurtful and if i were an editor i'd ignore it

<dmitriz> -1 tantek :(

dmitriz: (queue update)

<capjamesg> Group GitHub: https://github.com/swicg

<capjamesg> (presently not active)

dmitriz: it is not up to any of us to make quality judgments about community discussion channels, CGs have an obligation to their communities
… our goal is not solely to editor productivity and specs, but to balance outputs against community inputs
… if I can turn to bengo's concrete proposal, I don't see how community feedback is a roadblock to productivity
… or onerous in general

<tantek> for the record, -1 on the proposal

<KevinMarks> I'm not on the queue

david somers: I think async is very crucial here, and lots of us are timezone distributed ffrom one another
… i like github issues for document-based and long-form/thoughtful contirbutions
… but mailing list can help people workshop ideas before they post them on github

<dmitriz> +1 github issues might be a great tool

<tantek> exactly what david somers said

<tantek> +1 david somers

lisa: not all community input is good! charters and codes of conduct exist to filter out non-constructive input and feedback

<tantek> +1 lisa

<dmitriz> +1 Lisa, I meant more about increasing community input slightly, from what it was before. I heartily agree there are careful limits, within charter

bengo: I would contribute my personal experience to say that in many CGs I have participated in, I use mailing lists are primary reporting and feedback mechanism
… and take decisions WITHOUT having to miss work and family obligations to attend sync meetings
… there is a section of "gaining consensus" in the w3c docs that i find really useful here, w3c has been based on these rules for 20 years
… a big part of open standards (and their usefulness to regulators and legislators) is gathering (filtered, of course, and weighted) the feedback of all affected parties, not just implementers or editors

tantek: huge +1 to async productivity. the w3c used to use mailing lists primarily but is more github based in the last 10 years for all issues of substantial or technical issues
… including issues of process, which can also be done in git-structured and issue-threaded ways

<bengo> I've never seen GitHub issues used for calls for consensus and notice of provisional resolution, which is all my proposal covers

tantek: i've seen mailing lists that are net-positive for many groups, but I see them less and less over the years
… i think there are some bad-faith or emotionally-charged "blowups", as Amy (rhiaro) wrote in their recent blog post, happening on the fediverse threads about these issues
… and i feel the collegiality is suffering from a general decline in buy-in and consensus
… we need to find a way to raise the bar and heal the tone and tenor of the discussion in the channels
… i would say that the dominant mode of decision-making in working groups in w3c is synchronous, in-meeting group decisions

<capjamesg> +1 re: new information.

tantek: getting input from others is important, and even synchronous meetings are susceptible to additional information reopening the issue

<dmitriz_> +1 that a lot of WGs make decisions on calls. But many/most of those groups also have a period for consensus/objections

tantek: but not simply by people who miss meetings saying they would have objected if they had been there

<bengo> My proposal includes a clause establishing that decisions are pending new information.

tantek: i agree with lots of the principles in the proposal but im still -1 on the proposal procedurally

evan: i'm really interested in how this has evolved

<plh> [I created a placeholder in w3c/strategy to register the conversation on a social weg wg: w3c/strategy#435 ]

<tantek> +1 dmitriz_ yes, period for async discussion BEFORE WGs make decisions on calls. thank you for noting that

<tantek> also, issues typically added to agendas of meetings in advance for discussion

evan: i think async is a clear value-add here, and i agree that making major decisions during sync meetings is a real handicap here to full participation by half the world's timezones

<Loqi_> [preview] [plehegar] #435 Restarting the Social Web Working Group

evan: i agree with tantek that we should optimize for the goals of the group, which have until now been very git-focused for deliverable work, particularly normative work

<dmitriz_> @tantek - that's essentially what Bengo's proposal is. adding a period for async discussion before decisions

evan: i think more administrative and organizational work benefits more from async decision making

<tantek> dmitriz_: not really, it's post sync which is not what w3c groups do

evan: so i support both. i think re-considered all prior decisions 1 by 1 would be a bad idea, but i'm glad to hear that bengo did not intend that

<dmitriz_> +1 evan that previous decisions are binding

<bengo> Evan, absolutely, I never meant to call into question past resolutions that have been posted.

evan: so i am more open to this given that was my only really blocking objection

capjamesg: i have to volunteer here that i am putting in a lot of time watching all the channels

<tantek> capjamesg++ chairs doing a lot of volunteer work

<Loqi_> capjamesg has 3 karma in this channel over the last year (122 in all channels)

<tantek> dmitriz++

<Loqi_> dmitriz has 1 karma over the last year

<omz13> one caveat with in-meeting decisions is that sometimes life will get in the way resulting in people not being able to attend (which is, for example, why I couldn't attend the last meeting)

<tantek> nightpool++

<Loqi_> nightpool has 1 karma over the last year

capjamesg: i take seriously my role as channeling and filtering inputs
… but chairs (current and future) have to ultimately mediate what gets discussed and decided by the group

<Zakim> tantek, you wanted to react to a previous speaker

capjamesg: curating finite discussion time and prioritizing and timelining
… it's also our duty to consider the process malleable and improve it over time
… so we (as chairs) are open to ideas like this to make the group work better

dmitriz: (queue note: kevin you're on the irc queue and we're using jitsi queue)

<bengo> Woah GitHub for process discussions is bold. Unlike when SocialWG editors started using it for certain work items, GitHub is now owned by a completely different company that has started datamining participation there to train AIs that ignore the licenses by people who use the platform. I have a big concern of starting SocialCG procedural chatter on GitHub.com via w3.org

dmitriz: evan, it sounds like maybe your main objections are met?
… tantek, how can we move towards consensus on this?

tantek: i think this freestanding document is a little odd, could we include this in a CG-wide charter?

<capjamesg> I am happy to lead an initiative on a CG charter.

<bengo> Charters are a kind of Operational Agreement. My proposal is to establish an operational agreement. A charter is not a meaningful distinction. We can pass operational agreements atomically, not a big document with pork/horse-trading.

tantek: one major concern is that it feels like a proposal in a vacuum. so maybe it's time to have a charter after all, since there is disagreement about how to proceed

<capjamesg> And work with chairs to help us get to a document ready for the edification of the community.

evan: maybe these one-off documents are good as an intermediate or stop-gap measure to build up a charter but editing it into a cohesive charter would be great

<dmitriz_> @bengo -- +1, we can work on Operational Agreements in parts. And "how do we make decisions" is an important part of the OA

<Loqi_> @bengo has 3 karma over the last year

capjamesg: i can definitely put some time in to make a full charter with the group, happy to take inputs via list

plh: i opened an issue to track the new possibility of a second working group, on w3c git repo...

<plh>

<Loqi_> [preview] [plehegar] #435 Restarting the Social Web Working Group

plh: it's a public issue, everyone is welcome to comment on it

evan: can we take a group decision now on the call?

<capjamesg> plh See https://www.w3.org/wiki/SocialCG/WG_Charter_Discussion#Deliverables for a page sent to the community for contributions.

kevinmarks: i think there's a tension between synchronous and async

<dmitriz_> PROPOSAL: adopt the proposal

<dmitriz_> https://lists.w3.org/Archives/Public/public-swicg/2023Oct/att-0024/swip-37f2-proposal.md

kevinmarks: i think keeping document-based discussion in a git-based platform like github is fine, and the prior art is heterogeneous here
… some groups find email and document-based to work well together, others don't, no one-size solution

<dmitriz_> +1 (we can modify the proposal to add Github to the process)

kevinmarks: i don't love the idea of going back in time 20 years and discussing documents over email

<tantek> -1 on the proposal, especially "provisional" point for sync decisions. async discussion should occur before an item is brought to a sync agenda

<bengo> The proposal text is

<KevinMarks> -1

<capjamesg> -1, but could be a +1 if the time period is reduced to 7 working days, and we use GitHub Issues.

<snarfed> -0

<bengo> irc... speaking of 30+ year old work modes

<bengo> +1

<pzingg> +1 but would like to have the email responses archived in a Git* issue

dmitriz: kevin would you be more ok with github being source of truth?

kevinmarks: no, i think email being source of truth is the problem

<eprodrom> +0 with the understanding that this proposal does not cancel out previous resolutions of the CG

<snarfed> GitHub _Issues_ vs files in git, very different!

bengo: could you explain more, kevin, about the CG groups that don't use email?

tantek: i disagree that email or async decision are the W3C norm
… and if anything email decisions are an anti-pattern
… i would support async like on GitHub issues discussion preceding official group decisions during sync calls

<tantek> -1 on requiring any sync meeting decisions as "provisional", that should be up to the discussions in the group and chairs facilitating decisions in the groups

capjamesg: young folks are easier to get on github than onto mailing lists

<lisarue> Sorry for joining IRC late; there's a little too much IRC to catch up on so quickly, but I did reply to the mailing list that a full-coverage call-for-consensus always-to-email process gets a "-1" from me.

capjamesg: even the notion of mailing list subscriptions can block participation
… also the multiple channels is an issue

<dshanske> I'm older than capjamesg and I find mailing lists ineffective for this. I only use mailing lists that offer static type notifications, not discussions.

<tantek> +1 dshanske

capjamesg: that said, imposing a discussion period is also good, to prevent "surprise decisions" during meetings

<eprodrom> PROPOSAL: form a CG charter task force

evan: i'd propose a charter task force?

bengo: other agenda items?

dmitriz: I think we've heard from everyone, let's continue this async

<bengo> I think my proposal strongly encourages sharing Calls for Conesnsus to other forums (e.g. SocialCG forum aka socialhub, and I think a GitHub repo is compatible with that too. One issue is that would be a brand new repo I assume, so I don't expect there is a lot of existing audience there unless there is a process repo for SocialCG I'm not aware of.

bumblefudge: i heard people on this call talking about documents needing a different workmode than administrative decisions, big +1, i think implicitly this proposal wasn't intended for normative decisions but for decisions about what documents go into git in the first place

Social WG Charter

<capjamesg> wiki page: https://www.w3.org/wiki/SocialCG/WG_Charter_Discussion

capjamesg: polling for editor interest is a top priority here

dmitriz: would git issues be easier than a wiki or list threads?

capjamesg: yeah I think swicg repo on github is good

tantek: +1 to issue-based next steps for this
… but wiki is maybe more open than github or mailing list

<snarfed> +1, issues and maybe wiki, not files in the repo

<bengo> point of info: anyone can email public-swicg@w3.org. Is the same true for who can edit/contribute to the w3c wiki?

^^^

tantek: plh is here, we should ask him about how to charter

bengo: strategy issue is there until the CG proposes a WG charter?
… is there a failure mode where CG doesn't have consensus on a WG charter?

plh: it's not a critical failure, a charter can still get accepted, but i strongly prefer the CG propose one charter
… multiple charters are a problem we prefer to zero charters

bengo: what does the w3c want to see?

<lisarue> I'm not a big fan of moving large amounts of discussions to github but I don't object strongly enough to object.

plh: like i said at TPAC, the top priority is a healthy feedback loop to the specs that maintain and make updates as needed to it

<dmitriz_> @lisarue - I think the idea is that discussions can be on list or wiki or github. And will be tracked/linked to/summarized on github

plh: if there aren't updates you're blocked on producing, you don't need a WG

bengo: can CG write its own updates to the specs?

<lisarue> @dimitriz_, I think your clarification makes my concerns worse.

plh: yes but we prefer a WG for IP purposes, even if it's just an "umbrella" documenting technically the CG consensus goals
… so the goals have to come first

<bengo> thanks plh

<dshanske> +1 to actually having clearly defined goals before building a political structure around it.

capjamesg: editors of other socialCG specs have proposed normative updates to those as well, as per my email last month

<dmitriz_> @lisarue - what would you like to see instead?

<capjamesg> Quick time check that we have 2 minutes left, but we can continue discussions async.

evan: is the WG requirement strict for normative changes?

<lisarue> @dmitriz_, let's not get into that here because I don't object.

<dmitriz_> @lisarue - thanks, noted.

plh: 7 or 8 years ago, it would have been, but we have shifted norms a bit to optimize for ongoing maintenance
… so that patent protection continues to protect ongoing updates to patent-protected documents

<bengo> It would be helpful to have a special topic call explaining the patent policy protection and risks of opening vs not opening a new WG that would have new patent exclusions on top of the ones that arleady happened in SocialWG

tantek: bengo asked about ambiguities
… and those sound normative to me

bengo: but errata and other amendments to docs CAN be accepted by w3c staff

<lisarue> (I'm happy to disagree and commit on many things but also happy to explore why I might have some disagreement but on the third hand it's a messy convo with lots of nuance and interdependencies)

plh: the w3c team CAN update specs directly, but it's more an exception than an option, we don't love the additional IP risk in those cases

evan: lots of anxiety in the broader community that a WG would change too much, embrace/extinguish, etc

<bengo> great question Evan

evan: are there ways we can get a WG that optimizes for CG input and allays those concerns?

<dmitriz_> +1 Evan

plh: actually, let me zoom in on a previous comment, w3c team has actually downscoped its leeway for those direct changes

<tantek> +1 plh

plh: to purely editorial or non-substantive ones

<bengo> I believe few ActivityPub list/forum/fediverse users have expressed an interest in substantive changes to the TR

<tantek> +1 bengo

plh: so i would mention that the WG can be scoped to a way that allay concerns about scope creep
… for example, a charter can say, "only clarifications and improvements to the existing functionality with no new functionality"

<lisarue> I'm going to leap ahead a bit -- I'm +1 forming a proper WG . Messy imperfect groups like the W3C exist to imperfectly protect public goods like these specs from capture.

<tantek> you can also do that per spec

<tantek> +1 forming a proper WG inclusive of specs with editors that want to work on them in the WG

plh: or you can even go so far as to say "WG can only change functionality or add features with CG consensus"
… that is also a valid option in a charter

<bengo> "Note: If there is no Working Group chartered to maintain a Recommendation the Team cannot make substantive changes and republish the Recommendation. It can, however, informatively highlight problems and desirable changes using errata and candidate corrections and republish as described in the previous section."

<bengo> I meant w3c team "informatively highlight problems and desirable changes using errata"

<bengo> https://www.w3.org/2023/Process-20230612/#revised-rec-substantive

bumblefudge: not seeing much wiki activity how do we get people writing to it? email?

<tantek> bumblefudge, lots of wiki edits, mostly from Evan :)

capjamesg: any CG member has access, email already went out, please go to wiki and provide features
… they want to see in-scope or out-of-scope of upcoming substantive changes to specs

<tantek> see https://www.w3.org/wiki/Special:RecentChanges for example, note all the contributions from eprodrom++

<Loqi_> eprodrom has 1 karma in this channel over the last year (2 in all channels)

<dshanske> Yes, I confirm I was here.