07:33:00 RRSAgent has joined #social 07:33:04 logging to https://www.w3.org/2023/09/12-social-irc 07:33:04 RRSAgent, make logs Public 07:33:05 present+ 07:33:05 please title this meeting ("meeting: ..."), tantek 07:33:06 present+ btsavage 07:33:07 present+ 07:33:07 plh has joined #social 07:33:11 present+ 07:33:17 present+ Dan_Appelquist 07:33:23 present+ 07:33:23 meeting: Social Web Incubation Community Group 07:33:56 scribe+ 07:33:56 present+ 07:33:56 cpn has joined #social 07:33:56 chair: eprodrom 07:33:56 present+ Chris_Needham 07:34:00 eprodrom: this is the first in person meeting we've had since the CG started. Thanks everyone for being here 07:34:12 ... dmitri and james our regular chairs are not available, so I'm going to chair 07:34:25 ... reliving old glories 07:34:40 ... and I have some fine grained questions I'd like to work on 07:34:43 present+ 07:35:28 ... our agenda is going to be very focussed on existing specs 07:35:28 ... I shared an agenda on the mailing list this morning 07:35:44 https://lists.w3.org/Archives/Public/public-swicg/2023Sep/0004.html 07:35:45 TOPIC: extension policy 07:35:54 eprodrom: this remains an open issue, there's a draft extension policy to discuss 07:36:02 angelo has joined #social 07:36:04 ... and there a number of issues in the repo which need input from the group 07:36:10 ... to me they seem relatively minor, editorial issues 07:36:25 ... but changes to the errata are up for the group to discuss 07:36:35 ... any additional agenda items? 07:36:47 s/TOPIC/extension policy/ 07:36:49 q? 07:36:51 TOPIC: intros 07:37:18 dka: Dan Applequist, cochair of the TAG, and one of the original chairs of the social web incubator group way back when before social web was cool 07:37:39 ... what's the current status of .. there's been a lot of talk on the fediverse about the need to recharter the WG. What's the status? 07:37:47 Agenda+ group status and rechartering 07:37:50 ... Curious from a W3C process perspective, and as a denizen of the fediverse 07:37:57 eprodrom: It has changed significantly in the last 10 minutes 07:38:09 tantek: queued for the agenda 07:38:43 pfefferle has joined #social 07:38:47 eprodrom: we can move this up in the agenda 07:38:54 ... any other topics? 07:39:20 pfefferle has joined #social 07:39:45 tantek: relatedly, I'd like to talk about possible rechartering as well, specifically in scope of a bunch of the specs we have produced have advanced, had adoption and patches and fixes and gained critical mass. We've also seen several other protocol stacks emerge, I think there's an opportunity for bridging and enhancing interop across those stacks 07:39:52 ... that's a proposed agenda item 07:40:09 Agenda+ recharter for updated specs and new protocol stacks interop: https://tantek.com/2023/254/t2/socialweb-cg-w3ctpac 07:40:14 timbl has joined #social 07:40:27 [preview] [Tantek Çelik] going to the #SocialWeb CG meeting @W3C #w3cTPAC tomorrow (2023-09-12) at 09:30 CEST. 07:40:27 Looking forward to seeing @evanp.me (@evan@cosocial.ca @evanpro) and many others! 07:40:27 So many advances in #ActivityPub, #Webmention, Micropub, #IndieAuth etc. that it... 07:41:16 eprodrom: reminder that we do have two breakouts tomorrow about data portability, and interop and testing 07:41:16 ... Any other open questions? 07:41:16 TOPIC: extension policy 07:41:40 eprodrom: background. When we originally created ActivityStreams Core and Vocab documents we included a section on extensibility 07:41:51 ... the section covers the technical aspects, and somewhat the process aspects 07:42:02 ... suggesting we maintain a registry of well known extensions to the activitystreams vocab 07:42:13 ... and also that we have a process for including those extensions into the context document for AS2 itself 07:42:49 ... Assuming some understanding of JSON-LD in this case 07:42:49 ... you can import and reuse other schemas within the context 07:42:57 ... we do that already with AS2 for a couple of different vocabs, and we said we'd have a mechanism for that in the future, but we deferred on creating that process 07:43:02 there is already a community process: https://codeberg.org/fediverse/fep 07:43:12 ... we've had two (I think?) major additions to the AS2 context since we finished the REC 07:43:27 ... the first was including the activitypub specific vocabularly and the second was addnig the alsoknownas from the did vocabularly 07:43:41 ... during this time AP has been actively used and developed on the social web, and consequently there are a number of different extensions out in the world 07:44:24 ... the idea is having a structured process for including those vocabularies within the activitystreams context doc to make it easier for developers to just use one context 07:44:24 tantek I already joined 07:44:24 ... Any questions about the goals of this discussion? 07:44:31 miriam has joined #social 07:44:34 dka: can you give me an example of a user need serviced by an extension so I have more context? 07:45:14 eprodrom: common use case on the social web - playing games across the network. Eg. words with friends across the social network using your identity, being able to make guesses at words etc. Being able to create an extension that wasn't built into activitystreams at the time that supports the new functionality in a new way 07:45:40 present+ 07:46:08 present+ 07:46:16 mro has joined #social 07:46:16 q? 07:46:20 eprodrom has joined #social 07:46:27 https://www.w3.org/ns/activitystreams/ 07:46:36 eprodrom: this is our namespace doc for activitystreams 07:46:53 q+ pfefferle to note there is already a community process: https://codeberg.org/fediverse/fep 07:46:55 ... the html representation is an ED of a NOTE that describes the terms that are in activitystreams 07:47:12 ... the core activitystreams vocabulary, we have the AP extensions and the DID Core extensions 07:47:15 [preview] [Amy Guy] ActivityStreams 2.0 Terms 07:47:16 q+ to ask how many in meeting AP implementers do we have? 07:47:19 q? 07:47:51 ... there's also a reference in this doc that says approval of extensions will be by the CG.. process and criteria is being finalised. The suggestion is to finalise that and link it here 07:47:58 q? 07:48:35 ack pfefferle 07:48:35 pfefferle, you wanted to note there is already a community process: https://codeberg.org/fediverse/fep 07:48:40 eprodrom: community process for creating FEPs 07:49:09 ... for creating extensions.. Fediverse Enhancement Proposals 07:49:12 bumblefudge_ has joined #social 07:49:22 ... a light community oriented process that mirrors a lot of programming language enhancement processes 07:49:28 ... I think it's based on PEPs the python process 07:49:41 ... there are some very interesting extensions built with this mechanism 07:49:46 ... there's a lot of interesting process here 07:50:03 ... it's loosely connected to this community group through the socialhub forum 07:50:22 ... suffice it to say I'm not sure there's a part of the FEP process that includes the process of including terms within the AS2 context document 07:50:35 ... because the loose affiliation of people involved don't have access to modify that document 07:50:45 ... we as the CG are responsible for maintaining that document and doing these extensions 07:50:58 ... I'm not sure that the idea that there is this FEP process is entirely related to how we included extensions in activitystreams 2.0 07:51:01 ... does that answer the questions? 07:51:31 I only wanted to mention it, to not have two different processes 07:51:54 ... JSON-LD lets anyone include thier own vocabularys, many implementers do. 07:52:01 I can hear you but its very noisy 07:52:05 ... Those are the major sources of extensions and we should encourage all of that 07:52:13 was there a question for me? 07:52:15 ... the question si when do we take those extensions and include them in the activitystreams context document? 07:52:26 ... What is the criteria for including them in the AS2 context document 07:52:33 https://w3c.github.io/activitystreams/draft-extensions-policy.html 07:52:48 ... this is a draft extensions policy 07:53:31 ... it's very light 07:53:49 tantek: how many AP implementers we have here that would be participating in this kind of process? 07:54:02 ... how many people today can contribute to this discussion? 07:54:12 bumblefudge_: how are you defining implementer and participation? I'm working on FEPs with people 07:54:30 tantek: we're not directly implementing AP per se but we have launched a mastodon instance at mozilla.social and we've been making lots of modifications to that implementation 07:54:41 ... not to its ap functionality yet, but we are touching code that is touching activitypub 07:54:44 ... anyone else? 07:55:05 btsavage: Ben Savage, Meta. We may well at some point need to add extensions for various things 07:55:30 bumblefudge_: I'm working on the test suite, and not on a specific implementation, but specifically on the FEP process 07:55:33 ... hoping to help this be the on ramp to extensions 07:55:41 eprodrom: I'm also implementing activitypub in various ways, clients and servers 07:55:55 I am implementing for Seppo.Social. 07:56:03 Casey: I'd potentially be interested in using extensions for postmarks which is early stage 07:56:13 q? 07:56:16 ack tantek 07:56:16 tantek, you wanted to ask how many in meeting AP implementers do we have? 07:56:16 ... and at glitch.com, we're exploring AP integration, we might be exploring extensions for that but don't know yet 07:56:59 timbl has joined #social 07:57:08 tantek: pfefferle is implementing ActivityPub for WordPress 07:57:15 ... that answers my question, some people here who can give feedback 07:57:21 eprodrom: I'll discuss the suggested process 07:57:51 ... Section 2. Several step process. the first is encouraging publication, so publish extension for review 07:58:10 ... Noted a few places including the FEP system, NOTES on socialcg. Also publishing anywhere else. 07:58:21 ... Publish so it can be implemented 07:58:23 ... It needs to be implemented, have to have it in use 07:58:34 ... third is that we have a registry where we list existing extension 07:58:39 ... there's no barrier for that whatsoever 07:58:42 ... any extension can be listed 07:58:47 ... the fourth step is proposing the extension for inclusion 07:58:59 ... justification why it's important for this extension to be part of the AS2 vocabularly document 07:59:02 ... having a vote within the CG 07:59:15 ... Our group would decide whether or not to include that document 07:59:20 ... create a draft of the context document, including the new terms 07:59:26 ... we test that draft, and then it becomes official 07:59:43 ... I think that this is a fairly standard mechanism for making changes to an existing document 07:59:49 ... and for building extensions that become part of the core system 08:00:18 q+ 08:00:51 timbl has joined #social 08:00:59 eprodrom: the criteria for including extensions in AS 2.0 .. the extension must have its own context, describes terms and usage, has an IPR policy that's compatible with w3c 08:01:41 ... asking for 2 independant publishers and consumers regardless of whether it's c2s or s2s 08:02:09 ... a publisher and consumer might be the same software 08:02:20 ... that's the minimum for interoperability. Hard to make a case for interop happening with only one implementer 08:03:42 bumblefudge_: this seems really straightforward. There's something a little implicit which is what happens if my work is msasively overwhelmingly successful and there's 20 FEPs that have been implemented twice that each have testable context files attached in the FEP repo 08:03:42 ... and the WG has to timeline them, prioritise them, pick which are worth foisting on all implementers 08:03:42 ... there's a gating thing that isn't mentioned at all 08:03:45 ... once lal the criteria is met, there's a deliberation process? 08:03:59 q+ to ask about test cases for extensions 08:03:59 eprodrom: good point. One option is we just put every extension into the context, and that we don't make any decisions 08:03:59 ... two reasons we wouldnt' do that 08:04:00 ack bumblefudge_ 08:04:17 ... conflicts of terms, terms that are used in multiple vocabs, that makes it difficult to do inclusions so we need to resolve those 08:04:17 ... second is size 08:04:26 ... the context document is not large. 10s of kb 08:04:31 ... but there i sa point at which it can get very large 08:05:23 ... and it's hosted on the web. There may be size issues to be concerned about 08:05:23 ... Third, the process. How do we queue things coming in from FEPs and other systems? 08:05:23 ... the fact that we would vote on it within the CG before including something, is that the point at which we'd use that discretion? 08:05:42 bumblefudge_: I can imagine a couple of of ways of doing it that have strengths and weaknesses. I don't have an answer I prefer 08:06:02 ... I've seen extensions in other processes are welcome, but merging into the main context is almost like making it a required feature 08:06:07 ... some processes do that only at major versions 08:06:15 ... extensions now are eligible to be in the context in the next major version 08:06:20 ... that's one form of governance I've seen for these kinds of processes 08:06:26 ... in some contexts the testability is debated for a while 08:06:42 ... you have to provide an extesnion to a test suite that proves you can work with someone who didn't implement the extension for example 08:06:52 ... Just positing that it maybe forces other issues abot what's a major version, what's the WG/CG division of power? 08:07:14 eprodrom: interesting question. Because we've already done this twice since the major version of AS2, we at least have some precedent for doing it between major versions 08:07:20 ... I would not see that we need a version change in order to do this 08:07:29 ... I also think this is a backwards compatible process if we're doing it additively 08:08:00 ... if we were to add the security vocabularly that we use for http signatures in AP and is widely used, but the vocab has to be included each time you create an AP document 08:08:04 ... that's a strong candidate 08:08:49 eprodrom: Any examples of other vocabs, system, with a similar process of inclusion of extensions? 08:09:05 ... where extensions are built and then filtered into a core system? 08:09:13 q? 08:09:18 ack tantek 08:09:18 tantek, you wanted to ask about test cases for extensions 08:09:42 tantek: this is a great minimal start to a process. The one suggestion I'd have is in the criteria ... 08:09:46 https://w3c.github.io/activitystreams/draft-extensions-policy.html#criteria 08:09:54 ... it mentions demonstrate implementation 08:10:12 ... I think for that .. how do we .. what are the test case requirements for that? 08:10:34 ... usually you demonstrate a test suite that can validate for a publisher, or a for a consumer a set of test cases that produces a user visible result 08:10:44 ... I think it would be useful for the criteria to be explicit about that, and give an example 08:10:59 q+ to ask about i18n, a11y, privacy, etc... 08:11:01 ... In the microformats community we came up with a change control process for adding vocabulary terms 08:11:18 https://microformats.org/wiki/h-entry#change_control 08:11:20 ... once people are depending on it and systems are interoperating it's a different process 08:11:24 ... it's very similar 08:11:32 ... the difference is we have a few stages. Proposed, draft, stable 08:11:45 ... we've found that's useful to give people an idea of where a proposal is in the process, and how to move it to the next stage 08:11:57 ... one difference is we settled on 3+ publishers/consumers 08:12:11 ... people would get excited about an extension, you could easily get 2, and those wouldn't necessarily last 08:12:23 q+ to differentiate vocabulary changes (which should be backwards compatible) from features or behavioral changes (which require test cases/suite support) 08:12:25 ... asking for 3 has been enough that 2 will continue existing over time 08:12:35 eprodrom: that makes sense 08:12:49 ... Test cases do raise the bar there 08:12:53 ... a testable behaviour 08:13:10 ... not necessarily a bad thing ,but it would slow down that process 08:13:19 tantek: would help to say is there an implementation 08:13:22 q+ 08:13:38 eprodrom: would it be worthwhile to .. "required" and "desired" eg in a job description 08:13:51 ... don't request without 2 implementers, better with 3 08:14:00 q? 08:14:07 ack dka 08:14:07 dka, you wanted to ask about i18n, a11y, privacy, etc... 08:14:17 dka: recognise that you're trying to keep the process lightweight 08:14:22 bengo has joined #social 08:14:29 ... in the TAG we do review of specs, and we ask for explainers. Document the user needs 08:14:35 ... have you done accessibility review, security and privacy review 08:14:41 ... i know some of these things are going to be quite small 08:14:53 qq+ to answer dka 08:14:56 ... but it does feel like criteria.. if you're expecting some to be part of the core spec they'll need to go through that level of rigour at that point 08:15:01 ... better to catch those issues ahead of time 08:15:07 ... especially around a11y, i18n, privacy and security 08:15:17 ... we have assets in the TAG that can help people reivew their specs 08:15:18 dmitriz has joined #social 08:15:29 eprodrom: it's a good point, but a heavier process 08:15:47 tantek: support the horizontal review in the TAG, that's important. the vocab in AS2 predates the registry process in w3c 08:16:01 ... the way evan is proposing is much closer to the way registries are handled in in w3c now 08:16:06 ... closer than a new feature 08:16:06 ... but it's kind of a mix 08:16:11 ... the AS2 context is kind of a registry 08:16:20 ... however there's usually some interesting piece of user functionality which comes with each function 08:16:23 https://www.w3.org/2021/Process-20211102/#registries 08:16:32 ... that user functionality would hep to have some kind of horizontal review 08:16:44 dka: eg. the words with friends example - there'd be a microsyntax, and does it take accessibility into account? 08:16:52 tantek: right now, WGs can add things to registries without any review 08:16:54 present+ 08:16:57 ... somewhere in the middle 08:17:00 present+ 08:17:02 q? 08:17:04 ack tantek 08:17:04 tantek, you wanted to react to dka to answer dka 08:17:15 ack bumblefudge_ 08:17:15 bumblefudge_, you wanted to differentiate vocabulary changes (which should be backwards compatible) from features or behavioral changes (which require test cases/suite support) 08:17:46 bumblefudge_: some of the things that are currently FEPs include .. if something is only adding something to the vocab... a vocab change should be backwards compatible, deosn't break implementations 08:17:56 ... but most of the time they're not just semantic extensions, they're semantic and behaviour 08:18:12 ... we're dancing around the fact that the semantic extension part should always be backwards compatible. But there's a behavioural extension motivating adding the extension 08:18:13 timbl has joined #social 08:18:21 ... the context is just a registry, but there's almost always a behavioural extension attached 08:18:39 ... if the behavioural extension is big enough to maybe break implementations that haven't implemented it, then you need a test suite.... that's a totally different beast 08:18:45 ... maybe this is too complicated? 08:18:46 ... but some thoughts 08:19:00 bumblefudge has 1 karma over the last year 08:19:02 ... Also, there isn't currently a single robust test suite that is easy to extend as a fep is to write 08:19:19 ... if you coudl just write a fep and then add some test cases to an existing test suite it would be much easier to make a testability requirement 08:19:26 q? 08:19:27 ... but until we have that we're laying tracks in front of a moving train 08:19:36 ack btsavage 08:19:37 ack btsavage 08:19:42 btsavage: I like the suggestion from tantke that to b eincluded there need to be test suites 08:19:49 s/tantke/tantek 08:19:50 ... I think it's helpful for new activitypub implementers 08:20:00 s/to b eincluded/to be included 08:20:07 ... for all of the official extensions to have a test suite to work with 08:20:45 ... part of the incentive, to be part of the official list there need to be tests, seems like it would create a nice outcome 08:20:45 eprodrom: also one of the reasons we create test suites is so we don't have regressions or problems when we make additions 08:20:45 q+ 08:20:45 ack eprodrom 08:21:17 eprodrom: the question I have is how we move forward with this process 08:21:17 ... I'd like to propose we adopt the process document as presented as a draft to be maintained and developed by the CG 08:22:22 ... this would be a first draft note and we continue with modifications to this draft 08:22:22 tantek: does that include adopting it as process, or adopting it as something to publish? 08:22:22 eprodrom: the latter 08:22:22 ... not adopt the process yet 08:22:22 tantek: cg publish reports, but not NOTEs 08:22:59 timbl has joined #social 08:23:57 PROPOSED: use https://w3c.github.io/activitystreams/draft-extensions-policy.html as a draft of a report for an extensions policy for SocialCG 08:23:57 tantek: do you want suggestions for improvements filed as issues on that repo? 08:23:57 +1 08:23:57 eprodrom: yes 08:24:29 ... eg. versioning, test case requirements, etc 08:24:29 +1 to publish that as a CG report, and file our suggested improvements as issues on the repo linked in the draft 08:24:29 +1 08:24:29 +1 08:24:29 +1 08:24:29 +1 08:24:55 wrt "cg publish reports, but not NOTEs", I would like a citation on that 08:25:12 RESOLVED: use https://w3c.github.io/activitystreams/draft-extensions-policy.html as a draft of a report for an extensions policy for SocialCG 08:25:34 bengo, it's in the W3C Process, I bet plh can dig up a citation there 08:25:41 q? 08:25:47 agenda? 08:25:55 [preview] [Tantek Çelik] going to the #SocialWeb CG meeting @W3C #w3cTPAC tomorrow (2023-09-12) at 09:30 CEST. 08:25:55 Looking forward to seeing @evanp.me (@evan@cosocial.ca @evanpro) and many others! 08:25:55 So many advances in #ActivityPub, #Webmention, Micropub, #IndieAuth etc. that it... 08:26:04 TOPIC: group status and rechartering 08:26:08 wrt rechartering, that seems like something that should have been on the agenda 2 weeks ago 08:26:13 (per process doc) 08:26:22 dka: So... what's going on? 08:26:55 eprodrom: we were originally chartered in 2015 and worked for 3 years with some generous extensions, published 6 RECs 08:27:12 correction 7: https://www.w3.org/wiki/Socialwg#Recommendations 08:27:17 ... the off ramp for that WG was the idea that this cg would do ongoing maintenance of those documents 08:27:55 ... The current state of play is that the CG does some work in that area, we do meet fairly regularly 08:28:02 ... we do issue triage 08:28:06 q? 08:28:10 q+ to discuss whether the w3c process generally requires f2f meetings to have agenda items like rechartering to be published two weeks before f2f meetings (https://www.w3.org/2023/Process-20230612/) 08:28:15 ... however we're not able to update the rec track documents 08:28:16 --> https://www.w3.org/groups/wg/social/charters/ Social Web Working Group - Charters 08:28:17 q+ 08:28:29 [preview] [Elika J. Etemad / fantasai] W3C Process Document 08:28:31 q+ to ask where are the errata to ActivityPub and note living spec updates to Webmention, Micropub, IndieAuth etc. 08:28:42 ... one reason for us to recharter is for us to be able to move forward with those documents and take them to next versions 08:29:02 q? 08:29:10 ... There have been, since the publication of activitystreams and activitypub and the others, there have been a lot of development use, real world experience 08:29:20 ... there may be some value in applying those to iterations on those documents 08:29:27 ack bengo 08:29:29 bengo, you wanted to discuss whether the w3c process generally requires f2f meetings to have agenda items like rechartering to be published two weeks before f2f meetings 08:29:30 ack bengo 08:29:30 q+ to recommend dual CG/WG approach. 08:29:31 ... (https://www.w3.org/2023/Process-20230612/) 08:29:38 [preview] [Elika J. Etemad / fantasai] W3C Process Document 08:29:54 bengo: I noticed there was an agenda published before the meeting but it didn't include anything about rechartering 08:30:09 ... wrt to process it would normally the agenda for f2f meetings would be published 2 weeks in advance, especially to discuss something as serious as rechartering a group 08:30:15 eprodrom: I think this is a discussion, not a decision 08:30:19 ... but that's a fair point 08:30:28 qq+ 08:30:34 ... if we feel comfortable with it, we'll not make any decisions in this morning's meeting 08:30:47 tantek: the process doesn't have that requirement for CGs, only for WGs 08:31:10 ... second, a common process is to do agenda gardening as the first thing in the beginning of the meeting, people are given an opportunity to suggest or prioritise items 08:31:41 bengo: sounds good. Just worth having future discussions and not making resolutions without a bigger quorum 08:31:41 q? 08:31:42 q- 08:31:42 ack tantek 08:31:42 tantek, you wanted to react to bengo 08:31:48 ack plh 08:31:55 --> https://www.w3.org/2023/Process-20230612/#revised-rec-editorial Revising a Recommendation: Editorial Changes 08:32:21 plh: we don't need a WG to make editorial changes to docs, you can come to the Team for that 08:32:25 ... but for substantive changes, we need a WG 08:32:59 ... The way the current process.. the short version is the team propose a charter to the AC. We're looking for a group to tell us what they like for a charter, and we'll send it to the AC 08:33:43 q? 08:33:43 ... I'd be looking for, if we're going to create a WG to work on ActivityPub, I'd be looking to this community to tell the Team what you'd like to see in the charter because you know better than us 08:33:49 ... I'm very supportive of relaunching the social web WG for the simple reason that we have a rec and I'd like to have a WG to maintain it 08:34:02 ... when we closed the WG back then, we didn't have this process 08:34:15 ... Nowadays we dont' close the WG, they're in charge of maintaining the rec as a maintenance WG 08:34:22 ... Now there's interest in actively maintaining the rec 08:34:26 (for posterity, it is very hard for remote attendees to hear) 08:34:26 ... for ActivityPub 08:34:29 +1 plh 08:34:37 tantek: agree with plh 08:34:50 ... we have seen a huge amount of substnative functional improvements to AP and AS2 by implementers 08:34:54 fwiw, so far there's been fairly significant community pushback to rechartering 08:34:57 ... which has been largely self governed and operating fairly well 08:35:10 ... getting some of that into the core rec would be a good thing 08:35:16 ... we've had 7 recs come out of social web wg 08:35:19 ... some have updates already 08:35:43 q+ 08:35:49 ... maintained as living standards by the indieweb community 08:36:05 ... eg. micropub. Would be great to get iprovements rolled into the w3c spec 08:36:16 ... we published a NOTE for indieauth because it was outside of the scope of the charter. We have a lot more implementations since then 08:36:25 ... I'd requst it was considered in scope for a rec track document based on the NOTE 08:36:37 ... lastly, there are a few new related specs that have been implemented, like microsub 08:36:43 ... multiple implemenations there 08:36:54 ... clients and servers 08:36:59 ... to add to the scope of a new WG 08:37:10 ... example - bridgy.fed - interesting unintended successes of the previous WG 08:37:28 ... evan and I worked really hard with the staff contacts and chairs to take 15 or so approaches to the social web and narrow them down to 2.5 to 3 08:37:53 ... we decided instead of trying to fight that out down to 1, we'd allow those communities to develop their technical approachs, and use the wg as a forum to develop interop across them 08:38:00 ... in terms of semantics, minimising friction when coverting between them 08:38:12 ... now we have services like bridgy fed that can connect between these different stacks 08:38:26 ... on tantek.com I publish html, I use webmentions, and I use bridgyfed to connect directly with the fediverse 08:39:11 ... I haven'tw ritten a single line of ap myself, and send bridgyfed a webmention, it delivers them ap inboxes across a number of people 08:39:11 ... there's a dashboard 08:39:11 ... I'd like to replicate that with the other technical approaches we're seeing, some are not in this room 08:39:11 ... eg. BlueSky 08:39:13 bsky is going to IETF 08:39:18 ... and Nostr 08:39:24 ... blockchain based, also growing in adoption 08:39:29 ... we're at this interesting moment for the social web 08:39:43 ... people want to invent a bunch of approaches, and w3c could serve a role here to bring together a bunch of different approaches 08:39:47 ... that's my suggestion for scope 08:39:54 ... to define it in such a way to be welcoming to those other groups 08:39:56 timbl has joined #social 08:40:00 ... to build bridges and allow for interop 08:40:06 nostr<>AP bridging is already happening in the wild (somewhat chaotically for the moderation system of AP that is, as yet, underspecified...) 08:40:43 dka: if you do recharter a WG that you keep the CG running as well - having that dual mode approach I've seen work very well with other things 08:40:43 and i have also heard from the bsky team that IETF is being targeted for some specs 08:40:43 ... eg. the immersive web 08:40:43 +1 08:40:43 queue+ 08:41:16 ... keeping the WG lightweight and focussed around what it needs to do to update the rec 08:41:16 ... and having most of the technical work happen in the CG is probably the right balance 08:41:16 ... you probably already know that 08:41:16 +1 dka 08:42:19 ack tantek 08:42:19 tantek, you wanted to ask where are the errata to ActivityPub and note living spec updates to Webmention, Micropub, IndieAuth etc. 08:42:19 ack dka 08:42:19 dka, you wanted to recommend dual CG/WG approach. 08:42:19 q? 08:42:19 ack eprodrom 08:42:26 eprodrom: I'm concerned about opening specifically AS2 and AP to normative changes fo r a couple of reasons 08:42:26 ... it's going through a period of explosive growth 08:43:30 ... if we establish a process for treating an AS2.0 potentially breaking changes that may put a damper on implementation 08:43:30 q? 08:43:30 ... why would I implement AP1 today? I should wait until AP2 comes out and it's not done and it's taking years.. 08:43:30 ... may have potential for inhibiting uptake right now, unless it was well messaged, backwards compatible 08:43:30 q+ to note we can be explicit about keeping AP compatibility in the Charter scope 08:43:30 ... I will also say that there are ... with aset of recommendations that exist right now that are available for developers to use right now, I'm less supportive of more protocols being built right now 08:43:33 ... I'd probably not think that adding BlueSky to socialcg makes sense 08:43:49 ... it's an effort to build a proprietary protocal separate from existing standards 08:43:55 ... I'm not sure having an array of standards is helpful for developers or users 08:44:02 ... it's a different world than it was in 2017 08:44:26 ... considerably more people, more implemenations, more infrastructure 08:45:05 ... I would not necesssarily encourage other experimentation at this point that isn't compatible 08:45:05 q? 08:45:05 [actually, the right answer is 2014 :) ] 08:45:05 q+ to inquire where people think bridging/cross-protocol interop work belongs in a WG/CG split like dka proposed? 08:45:05 btsavage: if we are going to charter a wg, my recommendationw ould be to keep the scope quite limited 08:45:13 ... this is based on my experience in the private advertising technology cg 08:45:22 ... we've been attempting to create a wg for nearly 2 years 08:45:25 ... faced a lot of challenges 08:45:29 ... one set has been pushback on scope 08:45:35 q+ to also note rechartering a previous WG is very different than chartering a new WG 08:45:35 q- 08:45:37 ... it could potentially slow us down 08:45:42 ... in getting a wg chartered 08:45:51 +1 08:45:51 ... I'd love any advice to avoid a situation where it takes years 08:45:54 q? 08:45:56 ack btsavage 08:45:59 q+ 08:46:08 is bsky pbllc even a w3c member company? 08:46:09 ack tantek 08:46:09 tantek, you wanted to note we can be explicit about keeping AP compatibility in the Charter scope and to also note rechartering a previous WG is very different than chartering a 08:46:12 ... new WG 08:46:24 tantek: sympathise with the potential problems with divergance in this space, we share a lot of values there 08:46:35 ... it took a lot of hard work to come to the balance that we did in the original social web wg 08:46:40 ... I think there is a social componenent to this.. 08:46:51 ... maybe we don't need to get those discussions into the cg - maybe a liason effort 08:47:00 ... someone mentioned bluesky going to ietf? Maybe a liason opportunity there 08:47:08 +1 to btsavage -- each additional function we assign to this maintenance group is additional risk of falling into the charter swamp 08:47:20 q+ 08:47:24 ... I strongly agree that we should keep AP compatible. Against breaking changes against what's been implemented 08:47:30 ... we might make changes to reflect what implementations decided to do 08:47:44 ... charters can say we'll work on this spec, but out of scope are breaking changes as far as what is implemented 08:47:47 ... tha twould be one guardrail 08:47:52 ... we could put other guardrails in place 08:47:56 ... to not break momentum 08:48:00 What if we made an activitypub CG? 08:48:01 ... a WG can help reinforce what you're asking for 08:48:28 ... to btsavage's challenge... it's a different istuation rechartering vs chartering a new wg, to update recs that have taken off in the market 08:48:32 ... agree a new wg should have a narrow charter 08:48:36 ... a lot of things to establish 08:49:00 ... wouldn't suggest adding a bunch of new things 08:49:06 ... but maintaining what we know has been widely adopted 08:49:16 ... taking the things with multiple implementations which are not recommendations yet 08:49:24 ... and including those 08:49:28 ... eg. indiauth 08:49:33 q? 08:49:36 ack plh 08:50:03 plh: we can make the charter of the wg about only fixing.. not allowing new features. Only modify normative aspects of specs if it's already deployed and reflects reality 08:50:07 ... we can make the charter as tight as possible 08:50:11 q? 08:50:17 ... we're not working on a new version of AP, just making sure the spec reflects reality 08:50:26 ... PNG is an example of that 08:50:39 ... and we want new implementers to implement reality 08:50:45 ack eprodrom 08:50:45 ... we can set that as a goal in the charter 08:51:19 eprodrom: strong goal 08:51:20 ... As ben pointed out, we're not going to try and make decisions today, but it sounds like from this conversation that there is... can we do a straw poll? 08:51:32 ... sounds like we have support for exploring a charter 08:51:43 dka: gotta go, but +1, feel like this is necessary 08:52:11 eprodrom: suggest we defer this to the next cg meeting in 2 weeks 08:52:11 q+ to suggest a proposal 08:52:22 ... consider a proposal for starting a charter process 08:52:32 tantek: a good idea for the wg to do refinements 08:52:40 ... but there's a much broader set of participants here at tpac 08:52:47 ... what belongs in a charter, what belongs in a w3c charter 08:52:59 ... even a straw poll from the people in this room and on the channel would be a useful opinion to take 08:53:05 ... we can keep iterating on that 08:53:11 ... we should iterate in socialcg meetings 08:53:32 a straw poll would be misrepresentative of the broader cant-afford-to-attend-TPAC-on-company dime ecosystem 08:54:04 PROPOSED: Explore rechartering the Social Web WG to do updates on existing RECs with wide implementations, consider adopting NOTEs with interoperable implementations as WDs, and consider incubations with interoperable implementations for WDs 08:54:35 q+ 08:54:41 ack tantek 08:54:41 tantek, you wanted to suggest a proposal 08:54:44 ack tantek 08:54:49 PROPOSED: Explore rechartering the Social Web WG to do updates on existing RECs with wide implementations, designate a core suite of testing strategiesconsider adopting NOTEs with interoperable implementations as WDs, and consider incubations with interoperable implementations for WDs 08:54:55 oops sorry 08:54:59 strike that 08:54:59 plh: are you proposing that the WG has new features in scope? 08:55:27 tantek: yes because implemenations have added new features that are beyond the scope of the spec right now. Eg. extensions. I'd like the WG to consider a similar process for improving AP and AS2 specs 08:55:36 q? 08:55:40 ack plh 08:55:42 ack plh 08:55:47 bumblefudge_: +1 on including testing strategies 08:55:49 +1 08:55:53 -1 not all RECs are worth working on hand-in-hand 08:56:25 -0.5 (needs broader conversation w community first) 08:56:29 bengo: I don't think it's a good idea to charter working on all of the different recs that came out of the WG in the past 08:56:43 q+ 08:56:53 ... people look in the minutes of the days we voted on those TRs that there were some surprise things inserted in and voted on things that were published before even AS2 08:56:57 tantek: which things? 08:57:05 q- 08:57:07 q+ I have an alternate proposal to run 08:57:15 bengo: would make sense to narrow it to only AS2 and AP if that's where by far the most adoption has happened vs other recs 08:57:17 (slight adjustment, not really alternative) 08:57:24 ... we could clarify what we mean by wide adoption and enumerate exactly what we mean 08:57:32 q? 08:57:35 q+ 08:57:40 PROPOSED: Explore rechartering the Social Web WG to do updates on existing RECs with wide implementations, designate a core suite of testing strategies, and integrate it with the FEP process and/or additional independent/community processes. Maybe adopting NOTEs with interoperable implementations as WDs, and incubation of interoperable implementations for WDs could stay in CG for now, until the test suite has consensus that would be t[CUT] 08:57:41 ack bumblefudge_ 08:58:08 q+ 08:58:17 bumblefudge_: laying tracks before moving train... if a maintenance group could take as its target defining a set of test tools that would be the basis for interoperable things 08:58:27 ... and bengo's comment about defining what's wide enough adoption 08:58:28 +1 defining a set of test tools 08:58:31 ... I feel like those are slightly downstream 08:58:44 ... you could maintain what there already is and have a foundation for deciding which specs can be declared provably interoperable 08:58:51 ... is there any way to stagger it, or do you have to recharter? 08:59:00 ... Maybe a layering thing. I'd love some time to have a test basis before moving testing into scope 08:59:08 ack bumblefudge_ 08:59:12 PROPOSED: Explore rechartering the Social Web WG 08:59:38 +1 08:59:49 eprodrom: proposal that takes out what we'd be exploring. Explore means we're going to spend time in the cg over the next multiple meetings now that we have this on the table, to discuss what that would actually entail, and work with staff and members to decide about what that is 09:00:01 +1 from dka as well who had to leave who gave general support for rechartering a Social Web WG 09:00:02 +1 exploration without specifying detail yet 09:00:05 +1 09:00:08 is this a proposal or a straw poll 09:00:08 +1 09:00:25 +1 09:00:32 tantek: as a CG we don't make formal decisions. This is the same thing 09:00:58 +0 (doesn't seem urgent, given the overall lack of interest in it in 09:01:06 the wider community) 09:01:12 eprodrom: seeing some concerns 09:01:20 0 - this should have more notice for more votes to be registered 09:01:46 q+ 09:01:50 ... will reach out to participants who aren't here today 09:01:51 ack eprodrom 09:01:51 ack eprodrom 09:01:58 yeah i thought i was +1ing defining a few scopes and strawpolling them at the next meeting 09:02:00 ... we can defer the rest of the agenda for today 09:02:21 plh: come back to me with the conclusion, I'm interesting in the outcome of this conversation 09:02:25 timbl has joined #social 09:02:34 eprodrom: what would we come to you with? 09:02:46 plh: bullet points, whatever you want 09:03:00 ... we can turn that into a charter, and send it back for iteration 09:03:11 tantek: I'd suggest listing what we want to work on standards track, updating or new 09:03:14 eprodrom: thanks everyone, ajourned 09:03:25 ... reminder of two breakout sessions tomorrow 09:03:37 bumblefudge_: at 1215 and 1430 09:12:36 timbl has joined #social 09:37:26 timbl has joined #social 09:40:52 tantek has joined #social 09:53:39 RRSAgent, make minutes 09:53:40 I have made the request to generate https://www.w3.org/2023/09/12-social-minutes.html tantek 09:56:13 present+ Juan 09:56:21 RRSAgent, make minutes 09:56:22 I have made the request to generate https://www.w3.org/2023/09/12-social-minutes.html tantek 10:01:03 Zakim, end meeting 10:01:03 As of this point the attendees have been Dan_Appelquist, tantek, btsavage, eprodrom, bkardell_, rhiaro, cwilso, Chris_Needham, plh, miriam, pfefferle, dmitriz, bengo, Juan 10:01:06 RRSAgent, please draft minutes 10:01:08 I have made the request to generate https://www.w3.org/2023/09/12-social-minutes.html Zakim 10:01:14 I am happy to have been of service, tantek; please remember to excuse RRSAgent. Goodbye 10:01:14 Zakim has left #social 10:03:02 plh has left #social 10:20:12 timbl has joined #social 10:28:14 timbl has joined #social 10:29:23 timbl has joined #social 10:58:13 timbl has joined #social 11:20:57 timbl has joined #social 12:05:25 cel[c] has joined #social 12:05:25 timbl has joined #social 12:07:30 feld65 has joined #social 12:07:45 timbl has joined #social 12:13:06 nightpool[m] has joined #social 12:50:33 xkr47 has joined #social 12:57:05 herman has joined #social 14:16:22 tantek has joined #social 16:07:37 Social Web / Fediverse meetup tonight adjacent to TPAC: https://cosocial.ca/@evan/111047903686906202 16:07:56 [preview] [Evan Prodromou] OK, this is happening Tuesday night from 19:00-20:00 local. We're having a fediverse meetup aka a #FedUp.If you're in Seville for #W3C #TPAC and you're on the fediverse, come hang out with us for drinks and snacks.100 Montaditos, C. San Fernando, 29I...