15:01:29 RRSAgent has joined #social 15:01:29 logging to https://www.w3.org/2019/10/09-social-irc 15:02:11 I can scribe 15:02:20 scribenick: bengo 15:02:31 TOPIC: Evergreen Recs (followup if any after cwebber reached out to w3c staff) (DELAYED) 15:02:54 cwebber2: Let's start. First topic is I'd said I'd reach out about evergreen spec stuff. I've been busy and haven't been able to, but hopefully will be next meeting. 15:02:56 TOPIC: socialhub.activitypub.rocks 15:03:38 hellekin: Status update: there are a number of tasks before launch, e.g. creating categories, but we can make them later. 15:03:50 hellekin: nightpool is now an admin btw 15:04:59 hellekin: We still have a few topics to fill on datashards, OCAP, relationsihp between ActivityPub SIG and SocialCG 15:05:23 hellekin: We are ready to launch any time, but we can also wait a bit longer for cwebber2 and serge to start nice topics 15:05:31 https://socialhub.activitypub.rocks/t/last-tasks-before-launch/78/11 15:05:35 hellekin: There is a poll about whether to launch or wait ^ 15:05:41 q+ to address ActivityPub special interest group 15:06:06 q? 15:06:07 hellekin: We need more people to get involved and fill more roles, so we have a team and no one person burns out. Other than that, it's pretty much ready. 15:06:33 ack cwebber 15:06:33 cwebber, you wanted to address ActivityPub special interest group 15:06:37 cwebber2: Explaining if people want attention to type 'q+' 15:06:57 cwebber2: hellekin are you talking about the ActivityPub W3C CG, chaired by Evan? 15:07:20 Prague session https://redaktor.me/pad/p/community 15:07:21 hellekin: No I'm referring to the discussion we had in Prague about what the role of this socialhub.activitypub.rocks forum would be. 15:07:57 cwebber2: Got it. The reason I was unsure is because all before ActivityPub was worked on by w3c SocialWG, Evan had made a w3c cg called ActivityPub. Nevermind then. 15:08:35 cwebber2: The general conversation we had in Prague was that this SocialCG maintains more than just ActivityPub 15:09:15 cwebber2: But we discussed it could be useful to have a more detailed collaboration space for ActivityPub since there is no SocialCG mailing list, and in there ideas related to ActivityPub could be discussed and percolate, and then could be brought to SocialCG. 15:09:31 cwebber2: I am A+ in on that. We could do a vote to give socialhub an official purpose, but I'm not sure it's needed. 15:09:41 q+ emacsen 15:09:47 ack emacsen 15:09:50 rigelk has joined #social 15:10:31 emacsen: I was at this discussion, but didn't fully have the background. I request a quick clarification because datashards came up. 15:10:41 q+ to answer emacsen about datashards and socialhub 15:10:59 (that was Pleroma, but not me directly) 15:11:06 q+ 15:11:08 emacsen: Datashards isn't an AP protocol, but there are some AP implementors eager to integrate datashards into their projects. But that's a separate activity from AP. So is datashards discussion allowed/encouraged on socialhub.activitypub.rocks 15:11:10 ack cwebber 15:11:10 cwebber, you wanted to answer emacsen about datashards and socialhub 15:12:04 cwebber2: 1) socialhub.activitypub.rocks is going to be "the" AP forum and sicuss issues, then raise important topics to SocialCG as needed. 15:12:35 cwebber2: 2) The people running socialhub, including hellekin and nightpool[m] want to bootstrap the forum with topics of interest to AP community before opening it to general public 15:12:43 +1 15:12:53 cwebber2: One topic might be OCAP w/ AP, another might be Datashards w/ AP. 15:13:03 q? 15:13:07 ack sl007 15:13:09 ack sl 15:13:56 sl007: As hellikin already wrote on the forum, discourse will eventually implement AP, so the forum will be federated via AP. Also wrt datashards, there are categories in the forum, so to discuss datashards you could use #fediverse tag. 15:13:58 q+ to suggest how emacsen could help with the datashards stuff 15:14:12 q? 15:14:14 ack cwebber 15:14:14 cwebber, you wanted to suggest how emacsen could help with the datashards stuff 15:14:14 sl007: In general the forum is interesting for all kinds of 'fediverse' people, so it would be useful to have a summary of datashards there. 15:14:46 cwebber2: I think emacsen and I should have a conversation offline about getting this stuff going on the forum. I could use help since I'm timestrapped, emacsen is too. So emacsen and I should coordinate offline ASAP after call. 15:15:15 cwebber2: hellekin there has been a question about when forum is open to public. Do you want to respond to that? 15:16:10 hellekin: There is a poll in the forum. nightpool[m] and I think we're ready to go. Other, we're not sure. We have options: e.g. 15th of October fixed date. 15:16:19 hellekin: I'm okay with going as we are and launching right now. 15:16:37 proposal... launch socialhub.activitypub.rocks right now 15:16:45 cwebber2: My personal attitude is 'release early, release often' 15:16:49 q? 15:16:53 cwebber2: But I leave it up to forum admins 15:17:05 q+ 15:17:12 ack hellekin 15:17:13 (i can't see what is on it, as i haven't asked for an account yet) 15:17:26 hellekin: Melody earlier didn't ask for invitation, but was wondering why people here couldn't participate 15:17:30 q+ 15:17:38 ack ben 15:17:40 ack bengo 15:17:49 bengo: I typed a proposal suggestion earlier 15:17:59 i couldn't hear any of that 15:18:25 PROPOSED: Launch socialhub.activitypub.rocks as soon as possible 15:18:29 +1 15:18:33 +1 15:18:35 +1 15:18:39 +1 15:18:40 +1 15:18:43 +1 15:18:43 cwebber2: You can type +1 if you are in favor. -1 if you are against. or +0, 0, -0 if youa re wishy washy 15:19:16 +1 15:19:20 emacsen: +1 15:19:30 RESOLVED: Launch socialhub.activitypub.rocks as soon as possible 15:19:47 it's online :) 15:19:48 q? 15:19:58 https://socialhub.activitypub.rocks/ 15:20:03 cwebber2: Let us enjoy having the liveness of it 15:20:08 cwebber2: 15:20:11 q? 15:20:48 TOPICS: Introductions 15:20:51 TOPIC: Introductions 15:21:07 cwebber2: I realize we have some people who are new to the call. It's a tradition to have everyone introduce themselves. 15:21:53 bengo: hey I'm bengo / benjamin goering (sp?) and I used to be part of the socialwg and did two small implementations and happy to be here 15:22:35 emacsen: Hi I'm serge (emacsen). This is my second call. Im very interested in AP, no current implementations, but I hope to soon. And I'm interested in tangentially related technologies like datashards 15:23:04 hellekin: Hi I'm Hellikin. I run a nonprofit in Brussels caled petit singularity (?). I've been catherding, and work for a nonprofit that helps fund free software like ActivityPub 15:23:17 I'm jess. this is my second meeting. communication protocol nerd and writing a AP server implementation in an IRC that almost federates! 15:23:31 got the name bang on cwebber2 15:23:34 je so po 15:24:04 Petites Singularités participates in the NGI0 consortium (see https://nlnet.nl/discovery) 15:24:19 kaniini: Hi. I'm me, kaniini. I work on Pleroma. I've also worked on Mastodon in the past. And also GNU Social very very very long time ago. 15:24:31 kaniini: I also started the litepub group, which is kind of like the opposite of the SocialCG. 15:25:35 melody: I'm melody. I'm working with a small team at my cooperative on potential AP implementation. I'm mostly concerned with anti-harassment and anti-abuse online. 15:25:57 melody: you're super welcome to the well-being.team :) 15:26:05 rigelk: I'm rigelk. I'm working at a University in the East of France on Olki, a social platform for scientists that federates 15:26:51 sebi: Hi this is Sebastian. You might know me from the Prague conference. Aside from that I'm doing redaktor, a multimedia CMS that is based on ActivityPub. I'm trying to push self-hosting. 15:27:18 sebi: You can read about this in the forum's software category. Before, I worked as a photojournalist for 20 years, but unfortunately my camera sank to the Pacific ocean. 15:27:42 sl007: I posted some grants on the forum. If you want to help with redaktor it uses TypeScript, node.js 15:27:48 (hope sl007 is sedi) 15:28:09 present+ 15:28:10 present+ 15:28:10 present+ 15:28:10 present+ 15:28:11 present+ 15:28:12 cwebber2: I forgot another thing to start the meeting with the bot. we need to type 'present+' if you are present 15:28:13 present+ 15:28:15 present+ emacsen 15:28:16 present+ 15:28:24 chair: cwebber2 15:28:26 cwebber2: That way the logs will pick you up as having attended the meeting 15:28:30 present+ 15:29:05 cwebber2: Let's move forward to next topic: "Issue Triage from GitHub". First I want to talk about how we deal with it. 15:29:11 Agenda: https://www.w3.org/wiki/SocialCG/2019-10-09#Topics 15:29:39 cwebber2: We've handled github issues in other groups. We usually ask people to raise issues ahead of time, and if they don't, we just go through the tortuous process of going one by one 15:30:18 cwebber2: Im interested in the metaconversation about how to handle these. The W3C Wiki is hard to edit for everyone and add topics to the meeting page 15:30:43 cwebber2: So I wonder if there is another way to raise issues to the group. Maybe we could have a thread before the meetings, including issues. 15:30:58 q+ 15:31:06 ack bengo 15:31:12 scribe: cwebber2 15:31:19 q+ 15:31:46 bengo: it seems like the github is where most of the conversations would happen for a few years, I found it disillusioning in the past that I would raise things and then they would never get addressed 15:32:12 q+ 15:32:15 bengo: I think a real question is whether or not the ideal point of a forum to have "issue zero" level of issues, have them all resolved 15:32:16 q+ 15:32:19 q? 15:32:21 q? 15:32:25 ack kaniini 15:32:36 scribenick bengo 15:32:56 kaniini: One of the larger frustrations I've noticed amongst implementors who try to interact with the SocialCG (which led to litepub becoming a thing) 15:33:21 q- 15:33:30 kaniini: Traditionally there have been these meetings that people go to every once in awhile. People are supposed to open a GitHub issue. But then there are no facts or figures about when consensus is going to be reached. 15:33:32 q+ to suggest how github should be used vs how socialhub should be used 15:33:50 kaniini: e.g. someone files a bug, but there is never a response like 'we will discuss this issue at this time' 15:34:18 kaniini: Now that there is this socialhub forum, I think that's a better way to go. What you can do is to create the issue on GitHub, which tracks the issue. Then we can have a corresponding discussion on the forum that references it. 15:34:45 kaniini: That way there is a paper trail about what has been discussed, when consensus is made, and that way the entire reasoning behind each conslusion is clearly documented. 15:35:04 kaniini: If we can leverage socialhub to coordinate how issues are addressed, I think that would be optimal. 15:35:11 q? 15:35:13 ack hellekin 15:35:43 hellekin: I want to add to that. Discourse comes with GitHub integration. It's possible to allow discourse to write into GitHub 15:35:44 +1 hellekin 15:35:47 q? 15:35:50 ack cwebber2 15:35:53 ack cwebber2 15:35:55 ack cwebber 15:35:55 cwebber, you wanted to suggest how github should be used vs how socialhub should be used 15:36:15 cwebber2: For me, socialhub.activitypub.rocks hopefully points to a much happier future for the issue tracker. 15:36:38 cwebber2: I think we do want to aim for 'issues zero' and address/resolve all issues in some way 15:36:42 cwebber2: (as an ideal, if not in practice) 15:37:04 cwebber2: I think it was a mistake in this community that we kind of said 'now that the specs done, the issue tracker is both for discussion and issues we want resolved in the spec' 15:37:16 cwebber2: It made it very hard to sort through issues that need action vs open places for discussion 15:37:23 +q 15:37:49 cwebber2: Now we have a big opportunity with socialhub forum to move not-actionable discussions to the forum. Before we worry about issue-zero, we should try to dequeue all conversational issues and move them to socialhub.activitypub.rocks 15:37:58 cwebber2: I think it would be a clean separation 15:38:34 q? 15:38:39 ack kaniini 15:38:43 cwebber2: It would be helpful to have volunteers to filter open issues and identify conversational ones that might be best moved to the forum 15:39:32 q+ to respond to crickets in the socialcg 15:39:39 kaniini: As implementors of large deployments of AP (e.g. Mastodon/Pleroma), we have at times attempted to engage the SocialCG for guidance, and it's basically crickets. Some issues are more urgently actioned than others. 15:39:53 q+ 15:39:55 kaniini: I'd like there to be an 'urgent, actionable' status on these issues. 15:40:07 kaniini: Sometimes when we've engaged the socialcg for guidance, it's because we have ongoing problems. 15:40:31 kaniini: which means that we have users that are rightly angry because there is some defect in the way that the federated network works. 15:40:43 kaniini: ocap, key rotation are things we're pushing for because people are asking for answers. 15:40:56 kaniini: Ultimately what we need is some workflow for urgent items. 15:41:29 kaniini: The best thing we could get out of socialcg is support for getting these things accomplished. 15:41:45 q+ 15:42:21 ack cwebber 15:42:21 cwebber, you wanted to respond to crickets in the socialcg 15:42:23 kaniini: In liu of that support, we sometimes resort to quick-fixes. And having a priority queue would help us avoid quick fixes. 15:43:00 cwebber2: I think you're right that socialcg has not been the most responsive to issues people want to address. In some ways it's possible for the community to help us do better. You're right there could be better workflow, which we could discuss on the forum. 15:43:21 cwebber2: Here is my interim workflow proposal: If it's a conversational issue, post on socialhub.activitypub.rocks. It's an issue with a spec, raise it on the issue tracker. 15:43:39 cwebber2: If one of those things looks like we need to have a conversation, raising it for these calls would be helpful, and say that explicitly so it gets on the agenda. 15:43:50 puck has joined #social 15:44:19 cwebber2: We've had these meetings in the past, and the agendas are empty. I think not being able to edit the wiki has been a barrier. But we can work around that with people saying on issues or forum posts as 'please put on agenda for next call' 15:44:31 cwebber2 +1 15:44:40 cwebber2: Was that a sufficient response kaniini ? 15:44:41 q? 15:44:57 kaniini: yeah i think we're on the same page. As long as there's a workflow for something when things are on fire we need a solution now. 15:45:03 kaniini: Traditionally we haven't had that. 15:45:07 cwebber2: Let's work on it 15:45:10 ack sl 15:46:49 ack bengo 15:47:24 +q 15:48:01 bengo: there's a lot of issues that have a lot of conversation but don't move quite to something actionable 15:48:28 bengo: most of them aren't things that are quite at a point to be able to be yes/no voted on a spec change 15:48:30 q? 15:49:07 ack kaniini 15:49:52 kaniini: I agree many issues aren't actionable. That's one thing that makes litepub group different from socialcg. It's more close-door, but when we release something it's more in the form of a full specification. 15:50:38 kaniini: That's kind of due to the background of how that group formed. But I think we can document a lot of what we've learned in litepub (e.g. how to write a good specification), and that help us move toward actionable specification on the issue tracker. Like how IETF does RFCs 15:50:45 q? 15:50:46 q+ 15:50:58 kaniini: I think if we can shift toward that model for SocialCG, it could be really productive 15:51:02 kaniini +1 15:51:09 kaniini: (without the closed-door aspect) 15:51:27 cwebber2: closed-door wrt socialcg or litepub? 15:51:52 kaniini: litepub. The way litepub has made most specs, some developers in the group get together, talk privately, and someone makes a spec (usually me) 15:52:12 kaniini: e.g. pixelfed and pleroma work well on many topics. Mostly because we get together, bang out interop, and release something 15:52:19 kaniini: That's how our work on OCAP got released 15:52:45 kaniini: That's what I mean about closed-door: one or two people get together and bang out a draft and get it done 15:52:59 kaniini: socialcg seems to be the opposite: conversation, consensus, etc. I think we need the best of both worlds. 15:53:09 [lost connection] 15:53:09 First off and don't get me wrong 15:53:09 I do understand the anger of "great implementations" 15:53:09 did i understand right that implementations having most users should 15:53:09 be treated first? 15:53:10 kaniini: COnversation and consensus discussions a solid draft, not vague ideas. 15:53:10 then it raises a broader question to me which is about democracy 15:53:10 [ the same than with one very great very white man currently ] 15:53:10 I just mean : Diversity means to put weaker on the same level and 15:53:10 base democracy is important for the AP community. 15:53:11 And we should treat Issues important for protocol itself first … 15:53:15 q? 15:54:27 kaniini: There are people who come in and are idea people, but someone also has to turn ideas into specs. If you let idea people dominate the conversation without having actual spec-work, without being able to prove things out in an actual document that has revision/thought-process, then you ultimately hit the ground but the wheels are spinning instead of actual traction 15:54:43 kaniini: What I think is necessary is balance between ideas and hardened polished drafts. 15:55:14 kaniini: I think that socialhub could come into that. What we can do is have threads on socialhub for ideas. Then people who know how to turn ideas into specifications can do so. 15:55:56 cwebber2: 15:56:25 +q 15:56:32 ack cwebber2 15:56:35 ack cwebber 15:57:04 cwebber2: I think the call to make things more actionable is being said all around. Sounds like there in consensus. And that everyone also wants a place for conversations. 15:57:20 https://github.com/swicg/general/issues 15:57:40 q? 15:57:42 cwebber2: I would help, but need access to the repo 15:57:42 cwebber2: Any volunteers to identify conversational issues and propose they be moved to the socialhub forum 15:58:07 cwebber2: for Oauth access from Discourse 15:58:18 ACTION: give hellekin authority to open/close issues on activitypub 15:58:47 q? 15:58:59 ack kaniini 15:59:02 Can we start a socialhub forum thread for a megaproposal to close a list of github issues an dencourage refiling as forum threads? People can then raise concerns (if any) and we can vote all at once 15:59:14 kaniini: I want to address what sl007 said about democracy 15:59:19 bengo, I am completely +1 on that. can you do it? :) 15:59:22 bengo: sure 15:59:44 q+ emacsen 16:00:08 kaniini: In practice I agree. But also we're trying to make something that people depend on in the real world. So a balance is necessary. Can't have complete democracy without the actual userbase [having a say]. 16:00:34 bengo: you may paste the URL to the issues list on Github in the topic title, then describe the topic. There's a #todo tag ;) 16:00:57 kaniini: Implementations of all sizes and maturity level should be equally at the table. At the same time, when you have bad actors in the fediverse going around harassing people utilizing security vulnerabilities in the design of the protocol or caused by unclearness, I do believe that those issues do need more urgent triage 16:01:37 kaniini: Because the fediverse is based on trust, but if users cannot trust that their data is being processed in a secure way, then there is no fediverse anymore. Or maybe there is, but people wont use it if they don't trust it. 16:01:37 About "security issues" is what I would strongly consider as "we should treat Issues important for protocol itself first …" 16:02:19 kaniini: In terms of implementations needing to come to consensus on an extension, I think they can just follow the new socialcg process we're working on. 16:02:25 kaniini: do you think we need a private space to discuss these security issues urgently? 16:02:36 q? 16:02:48 PROPOSAL: extend meeting by 10 minutes to wrap up this conversation and then schedule next meeting 16:02:51 +1 16:02:52 +1 16:02:53 q+ 16:02:53 +1 16:02:54 +1 16:02:54 +1 16:02:55 +1 16:02:58 +1 16:03:01 emacsen: +1 16:03:06 +1 16:03:06 hellekin: i don't. because by the time these security issues are being discussed, they are likely being exploited in the wild already 16:03:06 RESOLVED: extend meeting by 10 minutes to wrap up this conversation and then schedule next meeting 16:03:18 ack emacsen 16:03:25 q- 16:03:48 sounds good to me 16:04:01 emacsen: To volunteer to help triage issues, should I coordinate with hellekin 16:04:05 cwebber2: Yeah and bengo 16:04:09 q? 16:04:23 (and besides, there is litepub group for mature implementations to privately discuss security matters) 16:04:55 TOPIC: Next meeting scheduling 16:04:58 cwebber2: We had a discussion about litepub how and how to work on specs and stuff, and think we might continue that discussion in next meeting. For now let's talk scheduling next meeting 16:05:14 cwebber2: A couple weeks ago we discussed doing this every other week, not just monthly. 16:05:32 cwebber2: We had also discussed switching between wednesdays and saturdays. Because some people can't make some days 16:05:38 cwebber2: I'd like a vote on those topics 16:06:06 PROPOSED: Continue bi-weekly (every second week) meetings of the SocialCG for the interim 16:06:09 +1 16:06:09 +1 16:06:09 +1 16:06:12 +1 16:06:13 ¿1 16:06:14 +1 16:06:14 +1 16:06:16 +1 16:06:29 emacsen: +1 16:06:37 RESOLVED: Continue bi-weekly (every second week) meetings of the SocialCG for the interim 16:06:46 yeah, wrong keymap :P 16:07:35 PROPOSED: Continue with rotation between wednesday and saturday meetings, with next meeting occuring on October 26th 16:07:37 +1 16:07:40 -0 16:07:43 +0 16:07:49 +0 16:07:56 emacsen: +0 16:08:00 +1 16:08:02 +1 16:08:15 +0 16:08:20 +0 16:08:26 jesopo, there are now events on the forum 16:08:32 RESOLVED: Continue with rotation between wednesday and saturday meetings, with next meeting occuring on October 26th 16:08:35 does that send out emails? 16:08:38 or so 16:08:43 if we put the meetings on socialhub, there's an agenda 16:08:52 not that I know of - hellekin ? 16:09:01 cwebber2: This raises a meta-conversation about how do we schedule meetings in the future across wiki/forum/irc. 16:09:10 q? 16:09:15 https://socialhub.activitypub.rocks/c/meeting/socialcg/l/agenda only shows upcoming meetings 16:09:16 cwebber2: Let's discuss later. Until then, people should feel free to propose topics on the forum, and I will move them to the wiki 16:09:21 q? 16:11:01 Zakim, bye 16:11:01 leaving. As of this point the attendees have been cwebber, jesopo, melody, hellekin, rigelk, bengo, emacsen, sl, kaniini 16:11:01 Zakim has left #social 16:11:09 i would also like to see a working group established for security / anti-abuse concerns (and mitigation) in the protocol itself 16:11:09 RRSAgent, create minutes 16:11:09 I have made the request to generate https://www.w3.org/2019/10/09-social-minutes.html cwebber2 16:11:12 RRSAgent, bye 16:12:04 for example, spam is a serious problem right now 16:16:15 rigelk: Historically, there are 'topics' section on this wiki (if you have an account): https://www.w3.org/wiki/SocialCG/2019-10-09 16:16:20 *amateur and rare porn lover liked this* 16:19:51 um 16:22:44 lain_soykaf: well, that spam issue is pretty easily resolved by just blocking sinblr/humblr ;) 16:23:15 created next meeting page: https://www.w3.org/wiki/SocialCG/2019-10-09 16:23:18 oops 16:23:20 hellekin: how do i get this pleroma swag that lanodan has on socialhub? 16:23:24 https://www.w3.org/wiki/SocialCG/2019-10-26 16:23:25 :P 16:27:40 https://www.w3.org/2019/10/09-social-minutes.html 16:29:29 Sorry, Insufficient Access Privileges 16:29:30 :) 16:31:29 https://socialhub.activitypub.rocks/t/identifying-activitypub-github-issues-that-are-not-really-actionable-should-maybe-be-closed-as-issues-and-converted-to-forum-topics-here/96 16:32:22 https://socialhub.activitypub.rocks/c/meeting/socialcg 16:33:02 I put up https://www.w3.org/wiki/SocialCG/2019-10-09/minutes based on my irc logs again because insufficient access privileges :P 16:33:13 I need to email w3c staff anyway so 16:33:16 I'll ask about it when I do so 16:33:37 tantek has joined #social 17:17:14 cwebber2: I missed discussion last week about 'evergreen' spec status. If there isn't already a forum post about it, will you please make one? 17:19:28 bengo: yes 17:39:48 hellekin: ping again about socialhub pleroma forum thingy 17:39:49 (: 18:18:32 kaniini: I'm on it 18:19:45 kaniini: you got it 18:22:41 thx 19:15:19 sl007 has joined #social 20:49:49 sl0071 has joined #social 22:14:47 dmitriz has joined #social 22:53:48 hi has joined #social 22:54:12 hello! 22:54:15 is https://test.activitypub.rocks broken? 23:19:21 Guest84 has joined #social