#rhiaroWelp because I'm great at adapting to timezone changes with my magic sleep-any-time powers, I'm entirely not jetlagged, so the call being at 4am for me is going to be mildly disruptive after all :p
#ben_thatmustbemeheh, rhiaro, since you are actively editing things, i won't edit now so as not to step on your toes, but i think you makde an error in that last change
#wilkietantek: well that's a pretty big set of hurdles to get pubsub through in terms of process related and naming is probably the most challenging thing to get through
#wilkierhiaro: we have one new issue since last week. this came from someone coming from i18n although not necessarily from that i18n group about the word 'inbox' and say you were an implementation of LDN related to emails and if you got a notification or error and used 'inbox' the meaning would be confusing.
#wilkierhiaro: the thread is long and has our argument and there isn't quite a good word to use
#wilkietantek: let's go to the queue as this is probably worth discussing
#wilkietantek: the only thing I'll raise as a related issue is there was a rather long thread about a user story about inbox that we changed but that wasn't a specification where there may be reasoning to draw from
#ben_thatmustbemewe had discussion going back a year at least on "inbox" being a poor term. I remember discussing this at F2F at MIT i think last year or two years ago
#wilkietantek: while I look for that issue I'll ack eprodrom
#wilkieeprodrom: on this topic, there don't seem to be great alternatives to 'inbox'. twitter calls it a 'home timeline' and on facebook calls it a 'feed' and other systems that call it an 'inbox'
#wilkieeprodrom: I understand there are tricky bits to it but it seems not worse than other names.
#wilkieeprodrom: what I'm say is that there aren't great alternatives.
cwebber2` joined the channel
#wilkierhiaro: right, that's what I thought. as 'inbox' is confusing for the context of emails, but other alternatives as 'feed' are equally confusing.
#wilkieeprodrom: the tricky part is if this is Amy's 'feed' the difference between what things Amy has published vs. what other's have published and that can be confusing to everybody
#wilkieben_thatmustbeme: if this is specific to LDN can we name it something related to LDN? like "notifications-inbox" or something to make it specific to that stack.
#wilkierhiaro: we are aligning with activitypub to align things with pump.io
#wilkieben_thatmustbeme: but this takes the place of that section in activitypub
#wilkierhiaro: but activitypub refers to inbox as well and has 'outbox'. cwebber2 also mentions that. I don't think it is worth that effort to change it.
#wilkierhiaro: and other people from the i18n group also replied and don't think this is an issue
#wilkierhiaro: it came up on their call in the last 5 minutes
#wilkiecwebber2: I want to weigh in and say we are in a space where it is hard to overload terms
#wilkiecwebber2: for instance, "object" is a very overloaded term we use in this space. and "actor" and you can say it is already defined in "actor model" and such.
#sandrorenames inbox to 7f486ef4-b534-4cfd-b1ac-85396da6e23c
#rhiaroPROPOSAL: Close LDN issue 52 without change as there isn't a better term that makes it worth changing at this point
#aaronpkthat's what flickr did with their early oauth prototype... they had "frobs" which were "flickr objects"
#wilkiecwebber2: avoiding this seems very difficult and our goal should be to define exactly what these terms mean and that's what these specifications do.
#wilkietantek: another thing you can add is a placeholder for implementation reports
#wilkiesandro: if they were both on github I could star them and be notified of changes. although if I had an LDN client I guess I would be! but I don't have a client
#wilkierhiaro: yeah, we'll get that all linked before the CR
#wilkietantek: those are the only things I'm seeing that are missing: the links to where the test suite will be and a rough statement "the test suite is coming" and the implementation reports and where they will be
#wilkieeprodrom: I have a question that may come up in the CR meeting... do we have something in there that links to social web protocols / other specs?
#wilkietantek: thank you csarven and rhiaro for your hard work. I know you wanted to take it to CR at the face-to-face and now what you are taking to CR is stronger.
#wilkietantek: that takes us to eprodrom and getting back to your issue or question about social web procotols about the relation to LDN and pubsub?
#tantekNext topic: relation between LDN and PubSub - is it documented in SWP?
#wilkieeprodrom: yeah, to answer questions about when we have more than one protocol in the use case area. if they are taken care of in social web protocols is that fine?
#wilkierhiaro: I still have to get social web protocols caught up with pub sub.
#wilkierhiaro: and adding LDN and subscribing to notifications. LDN only deals with delivery and not subscribing and we talked about PuSH as a method of subscribing and that helps us a bit.
#wilkierhiaro: there will be certainly a blow-by-blow alignment in social web protocols but not necessarily LDN
#wilkietantek: I think we could have a section in Social web protocols that mention this that would be good as something we could link to to help us in that transition call if we should need it
#wilkieaaronpk: it is under security considerations
#aaronpk"Receivers may periodically re-verify Webmentions and update them."
#wilkietantek: that sounds normative but optional. if you can add a note to the report stating what you just stated... that gets you to 2 implementations or just 1?
#wilkieaaronpk: just one but that feature doesn't affect interoperability
#wilkietantek: right. well, "MAY" is optional so we need to know somebody implements that so if you document that
#wilkietantek: you won't be able to publish updates to the recommendation so that's why we work to resolve these now
#wilkieaaronpk: let's say that errata are published on webmention.net after they go through github issues
#wilkietantek: ok, if you can pick a URL to say "this is where errata will be published" and link to that in the header like many w3c specs have a link to where they expect to find errata
#wilkietantek: in terms of processing them since we won't have a group to resolve conflicts you may want to document on webmention.net or elsewhere what your process is for handling issues
#wilkietantek: great question. this is an area that is evolving at w3c. many groups are having various success with varying methods. sandro, do you have opinions?
#wilkietantek: what is a good process for raising and processing issues after the group is closed? what are good processes you've seen?
#ben_thatmustbemewould this be a good time to discuss community group as a continuation for the group?
#wilkiesandro: I've seen two depending if there is staff or not. I've seen a wiki or repo that gathers issues and a community group... having a public list where a public space can comment on that and reflecting consensus when they see it
#wilkietantek: much like aaronpk had discussed and have github issues--
#wilkiesandro: as long as people aren't just closing issues
#wilkietantek: so as now issues can't be closed other than by the person that raised them or group consensus
#wilkietantek: I'm looking for explicit documentation about where errata issues are documented and discussed and a process. and if you think this is good, you use this or come up with something else and you write that down and link to that URL.
#wilkietantek: so people can look at that errata and know what has changed since the spec
#wilkieaaronpk: and that goes somewhere in the spec?
#wilkiesandro: if you look at existing specs they will have an errata section above the abstract
#wilkieeprodrom: I think what we are saying is that there will be an errata document that will be linked from the spec
#wilkieeprodrom: I guess I'm confused about access to github... can we put an errata document on github and link to that and still have access when the group closes?
#wilkiesandro: we can keep the access to github, yeah
#wilkieeprodrom: so linking to github will be sufficient for errata?
#wilkietantek: yeah. github seems like a reasonable place to put that. unless there are clear objections, I would put that choice to the editors
#wilkietantek: but once the group closes the errata doesn't have an official standing
#wilkieeprodrom: what has tradionally happened with errata in other specs?
#wilkieeprodrom: is there a treshold? where something is so clearly unimplementable that we need a 1.1 version of the spec or do they just pile up?
#wilkietantek: both things have happened. sometimes a working group starts back up to review and publish new documents and there are processes where you can amend a document
#wilkiesandro: I don't think so. it has to be approved by an advisory group so I think there needs to be a working group
#wilkietantek: I believe the staff can do that. it still goes to an advisory committee can do but staff can do that in absense of a group?
#wilkietantek: you don't want to do it for editorial things but as eprodrom said, severe things may be sufficient reasons to request w3c staff time to through this process
#sandro+1 since the WG is satisfied with the level of implementation
#bengoq+ to ask "What part of the Social WG chartered deliverables does Webmention fulfill?" It's mentioned as possible input to "A Web protocol to allow the federation of activity-based status updates and other data". Webmention alone isn't that. ActivityPub is closer, but doesn't mention activitypub
#Zakimbengo, you wanted to ask "What part of the Social WG chartered deliverables does Webmention fulfill?" It's mentioned as possible input to "A Web protocol to allow the federation of
#wilkiebengo: is webmention the federation part of our charter? where does it sit in regarding the charter deliverables
#wilkiesandro: my answer is that we have gone with multiple solutions because we didn't have consensus and that LDN and webmention are both solutions for federation
#wilkietantek: aaronpk, if you can start that page on the wiki and rhiaro and I can make sure we have all our 't's crossed before we take it to the director
#wilkieI had to really think about how to type 't's
#wilkieeprodrom: running through our checklist for PR status, the most important is the issues and any open normative issues
#wilkieeprodrom: we have only editorial issues except one normative issue we talked about already
#wilkieeprodrom: this changes a requirement from MAY to SHOULD and we talked about it and decided it won't take the same effort to redo the editorial process
#wilkieeprodrom: as far as issues are concerned we are doing well. I do need to clean out editorial issues
#wilkieeprodrom: I can do that over the next week or so
#wilkieeprodrom: we have a couple of comments that are still waiting for a reply from the commenter... I believe we said around 30 days for those if there is no response?
#wilkieeprodrom: so it may be time to start to wrap those up
#wilkieeprodrom: rhiaro has done a couple of those so I don't know if we should wrap those up or not
#wilkietantek: sooner is better... if we need a changes it would reset the clock
#wilkieeprodrom: we have a couple of requests that we decided, for example markup in the name where the poster wanted markup in the name but we decided in the group not to do that and we are waiting for their response
#wilkieeprodrom: we decided we would wait. I'm happy closing without the response.
#wilkieeprodrom: both the ones we have outstanding we decided not to implement are not normative changes
#wilkietantek: what we need with these issues is a comment "the working group decided <whatever the resolution was>" and a link to show our due dilligence to show to the director that we resolved this issue and the commenter didn't respond or disagreed
#wilkieeprodrom: I can do that. that's probably the best way to do it for now
#wilkieeprodrom: the other question is whether or not we had features not covered by test document
#wilkieeprodrom: everything we marked as a "feature" is covered by a test document and covered by the validator
#wilkieeprodrom: I think we have sufficient test coverage there to say we are testing all the features
#wilkieeprodrom: I didn't hold them to this to ask "how would this be used" but yes, it was incoming from Twitter engineernig
#wilkiesandro: I think it would be worth waiting if it gets them on board with the announcement
#wilkiesandro: and say "hey, would you like to be involved in the press for this?" and hold off a bit
#wilkieeprodrom: yeah. I'll be as straight-forward as possible; this was not from high-level people from Twitter. It was Twitter engineers saying yeah we want to implement this.
KevinMarks joined the channel
#wilkiesandro: traditionally, around the time your are at PR and go to REC, you get testimonials.
#wilkiesandro: even if they haven't implemented it but they are willing to say they are looking at it, that's still a win-win there
#wilkieeprodrom: I would be reluctant to get an annoucement from Twitter but would be happy to reach out to those who reached out to me and see if they want to give a testimonial.
#wilkiesandro: and if they tweet about it, we could just link to the tweet
#wilkietantek: and, yeah, try to get a course understanding if they want to implement this in a month or three months or just looking at it it helps us decide how to follow up on it
#wilkiecwebber2: still 19 issues open but I think they are all editorial with exception of #156 which was said to be a blocking for activity pub support on some existing implementations
#wilkiecwebber2: as in if patrick stewart publishes a mention to 1 million subscribers that servers don't get overloaded exchanging that
#wilkiecwebber2: we have a solution that diaspora seems to support but friendica hasn't responded
#wilkiecwebber2: someone in #155 mentions adding a history feature which sounds awesome but getting it right in this small timeframe would be difficult given the time and it would work as an extension
#wilkiecwebber2: I think we addressed the major issues. there is feedback I still want to record from the wide review.
#wilkiecwebber2: I don't have the test plan but I could have it by the CR call
#wilkiecwebber2: I could describe it briefly if that helps
#wilkietantek: I think we need a link to where the test suite will go and at that link describe the test plan description
#wilkietantek: there is confidence in the group that you can summarize what you would say there and point the director to that during the call and say we don't have it yet but this is how we would develop it in CR
#wilkietantek: I'm looking at the issues submitted to activitypub and congrats on the issues from outside the group. it is better to get those than not. good sign. congrats on that.
#wilkietantek: looking at the open issues, half of them seem editorial but the rest don't seem obviously editorial and that's something to resolve before moving forward
KevinMarks2 joined the channel
#wilkiecwebber2: a number of them *are* editorial but you're right we haven't proven they are editorial
#wilkiecwebber2: should we postpone moving to CR until next week?
#wilkietantek: just by looking at that, it would be hard to say to the director "hey we resolved these issues" and it would be hard to explain away these open issues that aren't marked editorial
#wilkietantek: I think the director would push back on that to get the issues marked accordingly
#wilkiesandro: we should have gone to CR last week!
#wilkietantek: the challenge here is to provide some reasoning. even if you pass the director and go to CR, the issues these would pose would lead us to go to CR again and slow us down at getting through CR
#wilkietantek: let's try to reduce that chance.. that's the goal
#wilkiecwebber2: I managed to churn through a substantial number of issues last week. so I should be able to get through these by next week. let's postpone.
#wilkietantek: do you have an expected number of implementations to know how many reports you'd get?
#wilkiecwebber2: my estimate is at least mediagoblin, the implementations I've done, rhiaro's implementation, someone else's implementation, and pump.io are at least 5
#wilkietantek: diaspora and friendica... any chance of reports from them?
#wilkiecwebber2: it is unlikely diaspora will implement within the short time of this group
#wilkiecwebber2: they are pushing hard on their own protocol. there are folks there that seem open to implementing it in diaspora and filing issues, and it is pivoting toward that but not fast enough for this group
#wilkiecwebber2: more likely in friendica because they implement everything
#wilkietantek: any other implementations you are seeing as potential outside of the working group?
#tantekalso congrats to both aaronpk on resolution to take webmention to PR, and rhiaro & csarven on resolution to take LDN to CR! thank you for all your diligent hard work.
#tantekrhiaro: I forgot to request during the call, could you prepare an update to Social Web Protocols to go out on Thursday since PubSub is going to FPWD?
#tantekI suppose that may have to just be an editor's draft update, until we can resolve to publish an update to SWP on next week's telcon
#ZakimAs of this point the attendees have been eprodrom, aaronpk, rhiaro, csarven, wilkie, ben_thatmustbeme, cwebber, julien, +, bengo, tantek, sandro, !, Benjamin_Young