#social 2020-11-07
2020-11-07 UTC
ilmu2, sl007, ben_thatmustbeme, Zakim and RRSAgent joined the channel
# 
RRSAgent logging to https://www.w3.org/2020/11/07-social-irc
ilmu joined the channel
manu joined the channel
# 
manu waves hello.
# manu Mumble is telling me that the server is using an old version of SSL and that the modern mumble client is refusing to connect to it.
# manu rhiaro -- anyone else having issues connecting via mumble w/ latest stable mumble client?
# sl007 we are on mumble
# manu it's the Ubuntu 20.04 mumble client -- server TLS negotiation failing because it's using old crypto :(
# nightpool[m] hm, I seem to be awake for this one, let me see if I can get on mumble
# manu could setup the W3C CCG channel for social if that would help...
# manu it uses open source jitsi -- browser based.
# manu trying to upgrade mumble client
# manu mumble is already the newest version (1.3.0+dfsg-1build1).
# manu looking at -- <W>2020-11-07 10:56:16.660 ServerHandler: TLS cipher preference is "TLS_AES_256_GCM_SHA384:TLS_CHACHA20_POLY1305_SHA256:TLS_AES_128_GCM_SHA256:ECDHE-ECDSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-GCM-SHA384:ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES128-GCM-SHA256:DHE-RSA-AES256-GCM-SHA384:DHE-RSA-AES128-GCM-SHA256:DHE-RSA-AES256-SHA:DHE-RSA-AES128-SHA:AES256-SHA:AES128-SHA"
# manu seems like new crypto
# manu rhiaro what version of mumble client?
vpzom[m] joined the channel
# vpzom[m] I'm on 1.3.3 and it works fine, but maybe ubuntu has it configured differently
# manu is scribin'
cwebber2 joined the channel
# manu Meeting: Social Web Community Group
# manu present: cwebber, nightpool, dmitriz, rhiaro, manu, vpzom
dmitriz joined the channel
# manu present+ sebastian
# manu nods.
# dmitriz present+
# manu how can we tell? :)
# manu cwebber2: Here's the agenda...
# manu pulls random levers, am i doing this correctly!?
# manu oooh, zakim
# kaetahbo don't
# manu rrsagent, make logs public
# kaetahbo do that
# manu sebastian: Would like to thank CCG for giving the call a home.
# manu cwebber2: We have a number of items -- let's start with alsoKnownAs
# nightpool[m] love a good p2p system :)
# manu cwebber2 if you're speaking, we can't hear you
# manu cwebber2 which browser are yo uusing?
# xkr47 rhiaro, I have no authority to do so, but this nickspamming is disturbing
# xkr47 rhiaro, it is fine :)
# manu cwebber2: Let's start with alsoKnownAs
# manu cwebber2: Basically, with this topic - I think Mastodon introduced alsoKnownAs term into Activity Streams namespace, has a well-followed pattern as to how it's used, not a part of AS.
# manu cwebber2: DID WG would like to make use of it
# manu cwebber2: So, we it would be great if we could, as CG, to ratify as an extension -- could do this in the meeting.
# 
rhiaro definition in DID Core: https://w3c.github.io/did-core/#alsoknownas
# rhiaro pull request to add it as an AS2 extension: https://github.com/w3c/activitystreams/pull/512
# manu cwebber2: Yes, that makes sense.
# manu cwebber2: Let's open it up for discussion -- feel like this could be an easy win, where we just vote and agree to do it.
# manu cwebber2: Just wanting to wait to see if there is any input?
# nightpool[m] q+ to talk about the current definition
# maymay +1 for alsoKnownAs
# dmitriz +1 for alsoKnownAs
# manu nightpool: Thanks Chris, I think we did some work in defining it in the wiki?
# manu nightpool: I'm trying to look it up -- thought some of those early extensions in AS namespace, thought we took a swing at defining them.
# sl007 q+
# manu nightpool: I know I made it for just one of them...
# manu nightpool: Let me look at my gist.
# manu sebastian: I think Amy mentioned the specs in social hub location.
# manu nightpool: I believe Amy just linked to Mastodon docs.
# manu cwebber2: We did it for sensitive at least.
# manu ack rhiaro
# 
cwebber2 I believe this is the discussed link https://www.w3.org/wiki/Activity_Streams_extensions
# manu rhiaro: I found that when I was looking this up originally, definition in Mastodon docs, write up a definition that made sense in DID WG perspective as well as make sense in Mastodon perspective. Put it on Social Hub forum to see if that made sense to folks.
# manu rhiaro: What I've done is make a PR on activity streams NS to add it as an extension... the concrete thing we should vote on is to determine if that PR should go in.
# manu rhiaro: Peopel from Activity Pub side or DID side change that before DID Core goes to CR.
# manu rhiaro: Might be worth talking about how to coordinate both groups.
# nightpool[m] Thanks Chris, i'm satisfied there isn't any unknown prior art here we should be looking at
# manu q+ to suggest we do this iteratively
# manu cwebber2: link in IRC
# nightpool[m] I was thinking of the #sensitive property primarily
# manu cwebber2: That has spec text definition, reading it out loud
# manu cwebber2: alsoKnownAs
# manu The value of alsoKnownAs MUST be a list where each item in the list is a URI conforming to [RFC3986].
# manu This relationship is a statement that the subject of this identifier is also identified by one or more other identifiers.
# manu cwebber2: That feels like a clean definition to me
# manu cwebber2: nightpool, does that address your understanding?
# manu nightpool: Yes, in actual AS namespace, just linking to DID Core.
# manu nightpool: Yes, that definition makes sense, although, I guess it's a little weird.
# manu q?
# manu cwebber2: Could I ask for clarity on what's weird?
# manu nightpool: Not entirely clear on DID spec, seems specific to DIDs...
# manu cwebber2: Let's get to the queue, Manu might comment on that.
# manu sl007: My question is about how we can express alsoKnownAs in HTML layer -- like rel attribute for links? If we should propose together with this proposal a spec for a rel attribute for a link.
# manu sl007: We have rel=me and nightpool pointed out that this is more like rel=alternate than rel=me.
# nightpool[m] Point of order: I said the opposite.
# manu sl007: The rel attributes can provide alternate protocols --
# nightpool[m] `url` is alternate, `alsoKownAs` is like me
# nightpool[m] q+ to the rel= question
# manu cwebber2: two things to respond to -- subject identifier, is it did specific, then rel stuff...
paul joined the channel
# cwebber2 manu: the first thing to let the pressure off of the decision... we might just want to agree to this iteratively. We don't have to agree to an exact definition today, we have time in the DID WG to modify it, if we can just get an agreement today that this is a good starting point we can refine, we have at least 9 months
# cwebber2 manu: second thing around nightpool's concern about talking about subject, also the "must be a list", the goal here is to share the semantics with the AS2 community. The DID spec just because of weirdness in that group has chosen to use introspect to use lists and sets and etc... so there's extra imposition on that
# cwebber2 manu: I think nightpool points out things that yes, you should be concerned about, but it's in no way an imposition on the AS2 spec... there's a general AS2 use, and the DID spec can help refine it without changing semantics... we hope it's more that the AS2 deinition is general and then the DID spec can put its own specific requirements for use on it
# manu rrsagent, draft minutes
# RRSAgent I have made the request to generate https://www.w3.org/2020/11/07-social-minutes.html manu
# manu nightpool: Yes, wanted to confirm that I think sebastian is not in IRC -- said something that didn't get proxied.
# manu nightpool: I said the same thing in IRC -- alsoKnownAs is more like rel=me than rel=alternate... there might be spec work to be done there, might be good to do unification there, agree with Manu, let's take this one thing at a time.
# manu nightpool: Agree with manu, let's make AS define something more general and then have DID Document refine it a bit... that's not what we have in namespace document, just a pointer to namespace definition -- we might wnat to do another draft somewhere w/ AS definition somewhere.
# manu rhiaro: The way it is written, definition in DID Core spec, first paragraph, DID specific stuff, actual definition - bold italics, that's the bit that's the definition of the term that should be generic for everyone, everything else should be DID specific -- distinct block that should be generic.
# manu rhiaro: Manu said there should be general definition in AS - no definition in AS for this, and I don't think we can change AS Vocab spec, which is a REC -- don't see it happening realistically, don't see how we get another term in there. The only definition that exists is in the DID Core spec. The way extension should work could be defined elsewhere.
# manu q+
# manu q+ to note Amy's right -- how do we want to proceed?
# manu cwebber2: Let me start out by saying -- we had some recent conversation at ActivityPub conference - we had a broad number of implementers at ActivityPub on what SocialWebCG should do moving forward, three things people wanted 1) This meeting about extensions and when meetings are important, 2) demonstrations of what people are building -- using screensharing about what they're building, 3) putting out CG reports especially around current state of
# manu extensions, both within and without AS vocabulary -- also about known patterns of deploying ActivityPub.
# manu cwebber2: First of two CG extensions - maybe we can at this time, not chartered to put out spec document, but can put out CG document... feels like good place to get consensus in the group -- excellent set of first things to add to that document. Even if it's not officially published thing, officially published in DID Spec, still useful to make record of this. I'd love to hear what Manu thinks of that.
# manu cwebber2: I have a suggestion of the path forward.
# manu q+ to defer to Amy :P
# manu cwebber2: Feels like we're going to have the followign things happen:
# manu cwebber2: 1) Everyone agrees that we would like to get alsoKnownAs in more general form as AS extension and have this group move forward with that.
# nightpool[m] rhiaro (IRC) I think Chris proposed doing it in a CG report
# manu cwebber2: 2) We don't need to have the exact phrasing right now, we just need to get agreement that this is the right definition in AS.
# manu cwebber2: 3) The rel question is interesting, but does not need to be solved immediately, given that there are two communities using alsoKnownAs, and that's not going to change given what we know.
# manu cwebber2: If we generally agree with that, we might be able to get consensus.
# sl007 q+
# 
csarven have W3C conneg the AS2 vocab to HTML(+RDFa) .. or create one eg. see https://www.w3.org/ns/ldp in text/html
# rhiaro csarven we have /ns/activitystreams, but for extensions it links to other specs for the text definition
# manu sl007: I'm just trying to generalize my question - as implementer of ActivityPub, about rel thing - if my software visits any website and wants to know what profiles are active, I'd check each rel=me link, basically. It's a "yes/no" question -- should we have an extra rel attribute where I can propose rel to IANA, for example, what AaronP did for micropub profile? Do we need something for ActivityPub?
# manu cwebber2: That sounds like a worthwhile thing to investigate, would you be willing to have us take that on, could you take leadership on that?
# manu cwebber2: Do you agree that it is not a blocker for this, but can be complimentary.
# manu sl007: Yes, exactly -- wanted to know your thinking, but would then take lead and talk w/ Aaron Parecki on Micropub.
# manu cwebber2: Let's make this a topic for next meeting? Sounds like you and nightpool have given it a lot of thought
# manu rhiaro: What you put in IRC, the 1-2-3 points sounds right.
# manu rhiaro: CG Report, good idea, wondering how we do it -- worrying about it being a bottleneck for the DID Work, can we do it in a way that things are decoupled? Will SocialCG stuff move quickly? One document for each extension that's ratified?
# manu rhiaro: Once CG Report is published, can it be updated?
# manu rhiaro: Thoughts on the suggestion?
# nightpool[m] (would like to note for posterity that existing implementations use Content-Type/Accept negotiation to do this, not HTML rels. and I think will probably want to continue to do this for the forseeable future)
# manu rhiaro: Starting w/ generic part of definition in DID Spec, AS points to it until something more appropriate per the definition points to something better.
# nightpool[m] (sorry, WRT sl007 (IRC)'s proposal, not the other discussion)
# manu cwebber2: That soudns really good to me.
# manu cwebber2: Wrt. CG report, don't want it to be a blocker. If it's a blocker, I'll walk back from it -- don't want it to block. CG work should be a complimentary publication, mostly about capturing current understanding and knowledge of the community... closer to indieweb on living spec/living document.
# manu cwebber2: General idea - create git repo with spec, encourage contributions, one about current practices, one about extensions, get leadership in Social Web communtiy to start working on that.
# manu cwebber2: does that sound acceptable?
# manu rhiaro: Yes, don't want to end up backed up into a corner - DID spec needs it's own defintion, term has to appear in AS namespace and JSON-LD context, I think we're good.
# manu manu confirms that he believes what rhiaro said to be correct.
# manu cwebber2: rhiaro, do you want to try to make a proposal?
# manu rhiaro: gimme two seconds.
# sl007 q+
# manu stops scribing dirty laundry. :P
# dmitriz hahahah wooot
# manu nods ins agreement.
# manu sl007: Just wanted to see if pukkamustard is here? Intro on fediverse enhancement proposals?
# manu dohs.
# nightpool[m] q+
# manu s/cwebber2: \/me/\/me/
# rhiaro PROPOSAL: we accept the alsoKnownAs definition in the DID Core spec ("This relationship is a statement that the subject of this identifier is also identified by one or more other identifiers. ") as a starting point that can be iterated on with the participation of both the DID and SocialCG communities
# manu s/PROPOSAL: we accept the alsoKnownAs definition in the DID Core spec ("This relationship is a statement that the subject of this identifier is also identified by one or more other identifiers. ") as a starting point that can be iterated on with the participation of both the DID and SocialCG communities//
# manu Amy makes two proposals.
# manu nightpool: I thought we wanted a more general definition that was unrelated to DID Core spec?
# manu nightpool: That doesn't seem like that's here in the proposals... but I guess DID Spec wants something normative.
# manu cwebber2: You're concerned because of language -- subject of this identifier and list thing is problematic?
# manu notes "as a starting point"
# manu cwebber2: This permits us to start iterating because of "as a starting point"
# manu cwebber2: I don't think subject and identifier are meant to be specific to DIDs... it's not meant to be DID specific, goal of starting point is to iterate.
# manu cwebber2: Amy does that match your thoughts?
# manu rhiaro: Yes, exactly -- identifier and subject are meant to be general...
# manu rhiaro: The subject could be an Activity Pub profile URl -- supposed to be general/generic language... relationship is alsoKnownAs -- ANY identifiers, HTTP URLs, DIDs, etc.
# manu nightpool: Yes, that makes sense to me... seems like there was some sort of disconnect -- define in AS and then nail it down in DID Core -- more conceptual sense to me.
# manu cwebber2: Ah, one more thing -- DID Core document contains same definition in CG Report, but it will also add some additional DID specific restrictions... will contain AS thing, and within DID Spec may put more requirements on top.
# manu cwebber2: Does that make sense?
# manu nightpool: Yes, that makes sense, and we don't want CG Report to be a blocker on DID WG.
# manu cwebber2: Is that aligned, rhiaro?
# manu rhiaro: Yes, DID Core one should be equal or more constrained of AS definition. They shouldn't ever be in conflict, and I don't think it's likely that AS would ever become incompatible w/ DID Core, it should be the more generic one, DID Core will be more specific.
sl0071 joined the channel
# manu rhiaro: About defining in AS -- where, what document? We can't change the spec, don't want there to be non-normative AS document that DID Core points to or depends on that we can't depend on from W3C Process perspective.
# manu cwebber2: It sounds like nightpool and rhiaro are in agreement, nightpool do you feel like that answered your questions?
# manu nightpool: Yes, I think we can move forward on this basis -- we can move forward asynchronously... like -- actors, do those make sense... can make it fairly general.
# manu cwebber2: Within AS usage -- could specifically mean this -- ok... let's get to the vote.
# rhiaro PROPOSAL: we accept the alsoKnownAs definition in the DID Core spec ("This relationship is a statement that the subject of this identifier is also identified by one or more other identifiers. ") as a starting point that can be iterated on with the participation of both the DID and SocialCG communities
# manu cwebber2: Let's get votes in..
# manu manu: +1
# dmitriz +1
# nightpool[m] +1
# paul +1
# paul not sure if I should have voting rights :9
# vpzom[m] +1
# paul I am
# sl0071 +1
# rhiaro RESOLVED: we accept the alsoKnownAs definition in the DID Core spec ("This relationship is a statement that the subject of this identifier is also identified by one or more other identifiers. ") as a starting point that can be iterated on with the participation of both the DID and SocialCG communities
# manu manu: +1
# paul +1
# dmitriz +1
# nightpool[m] i have one comment on the PR but it's not blocking
# sl0071 +1
# nightpool[m] +1
# manu yaaya!
# dmitriz wooooooooot!
# manu cwebber2: Yaay, the Social CG did a thing! This has been a productive meeting from my perspective. I think we're in a good starting place.
# manu cwebber2: Agreeign to merge that PR, that's a huge step there... we didn't get to all of our agenda items.
# manu cwebber2: Some of these other things might be easier now that we've had progress on alsoKnownAs --
# manu dmitriz: When is the next meeting?
# manu cwebber2: Good question -- happy to do one in as soon as two weeks ... but might be too close to the November holidays? Could do next week, non-holiday nearby next week?
# dmitriz +1 to either next week or the next nearby
# manu cwebber2: Next week or week of 21st?
# manu no strong preference.
# dmitriz yeyyyy
# manu cwebber2: Ok, let's do this next week then.
# manu cwebber2: I'll post on social hub.
# nightpool[m] manu++
# dmitriz lol
# manu lols.
# manu will try not to spend all that Karma in once place :P
# manu cwebber2: Great, congrats everyone on being productive during this meeting, take care!
# RRSAgent I have made the request to generate https://www.w3.org/2020/11/07-social-minutes.html Zakim
Zakim left the channel
# nightpool[m] rhiaro++ for bringing us the first useful #social meeting topic in months! :D
# manu HODLs on karma.
# manu +1 rhiaro++
# manu present+ paul
# manu rhiaro: I expect so :)
# manu rhiaro: I can report back to DID WG -- unless you want to do that?
# paul thx, I'm a total irc noob
# manu can do that now, might as well get it over with.
# manu paul: you're doing great :)
# manu Also, these IRC bots we're using are all sorts of arcane magic.
# manu paul -- https://www.w3.org/2002/03/RRSAgent
# manu oh no!
# manu what have I done!?
# manu paul++
# manu gasps.
# paul didn't hurt
# manu I've sullied paul's name... my apologies, paul.
# manu was just trying to share a link.
# manu paul ~~ https://www.w3.org/2001/12/zakim-irc-bot.html
# manu rhiaro: that's fine -- was trying to figure out how to signal more strongly to DID WG that progress has been made here... like, via email to mailing list.
# manu (helps demonstrate broad review/collaboration)
# dmitriz +1 emailing
# rhiaro https://www.w3.org/2020/11/07-social-minutes.html#ResolutionSummary yessss the arcane magic works so well
# sl0071 [pol, sorry] : you stare at a tv screen for days and then they announce it during the socialcg meeting …
# dmitriz hahahahah
# dmitriz yeah I just heard
# paul this work here was more historical IMO ;)
# dmitriz lol true
# nightpool[m] As another follow-up here, I wrote down what I would expect a "activitystreams-specific" definition to look like, based on the ways this is used on the fediverse: https://socialhub.activitypub.rocks/t/defining-alsoknownas/907/16?u=nightpool
# nightpool[m] posted a file: alsoKnownAs.md (1KB) <https://matrix.cybre.space/_matrix/media/v1/download/cybre.space/BPOvTYIFlvTFUErWivXjxcrB>
# nightpool[m] I’m thinking of this definition as kind of a “starting point” from the other side, so we can figure out what the generic definition might look like by unifying these two ideas and hopefully create something that doesn't feel too DID-focused. Please let me know if you have any feedback!
# nightpool[m] (It's also at the socialcg link I pasted above)
# nightpool[m] let me know if that's helpful!
# dmitriz thanks nightpool!
# dmitriz nightpool - great definition / starting point. My main question reading it is - from the AP side, does it require (or recommend) the two-way link? (as in, both Actor profiles have to have corresponding links?)
# dmitriz ahhh got it
# dmitriz one, that's really cool. (that it's progressive like that). and, worth mentioning in that definition
# nightpool[m] Yeah, i was a little torn on whether to emphasize that more or less—AIUI in mastodon there's no functionality until you get to a two-way link, and it might make sense to make it required, since we'd want to avoid the misleading/abusive case where someone could trick a implementation that only thinks only about one-way links for the purposes of harassment/etc
# nightpool[m] thoughts?
# nightpool[m] Ah, I may have been slightly imprecise—mastodon processes one way links for the purpose of providing the linked-to account the opportunity to confirm the links easily in the settings menu.
# dmitriz nightpool - yeah, I definitely think it's worth explicitly calling out both cases. the one-way properties, and the two-way security properties
# nightpool[m] yep, definitely! that was the needle I was trying to thread with my definition, will update to make it stronger + clearer
# rhiaro nightpool[m]: see the note following the alsoKnownAs dfn in DID Core - https://w3c.github.io/did-core/#dfn-alsoknownas - it's about that
# nightpool[m] thanks! looking now
# nightpool[m] I also really like the way that note explains it—the presence of the also known as assertion should not be confused for proof of the assertion
# nightpool[m] rhiaro: changes LGTM! I'm ambivalent on linking to the resolution in the namespace itself vs trusting Github's record of PRs or the git commit message to contain it, since parties interested in the procedures followed to produce the document will find the commit + PR anyway, but I'm not against including it
# rhiaro thanks nightpool[m]. Between github and the w3c servers it has to go through svn :D I expect the link to the resolution will be in the svn commit message too, but I also find it helpful to be very transparent about it, because if github goes down the commit history might not be easily available to many people
# nightpool[m] > Almost all cases on the web platform require an ordered set, instead of an unordered one, since interoperability requires that any developer-exposed enumeration of the set’s contents be consistent between browsers. In those cases where order is not important, we still use ordered sets; implementations can optimize based on the fact that the order is not observable.
tenma, includeals and someonewithpc joined the channel