From W3C Wiki
Jump to: navigation, search

Social Web Working Group Teleconference

16 May 2017

See also: IRC log


tantek, aaronpk, ben_thatmustbeme, eprodrom, cwebber, cwebber2
Ben Roberts


<Loqi> eprodrom: tantek left you a message 3 days, 17 hours ago: as ben_thatmustbeme pointed out appears to be unresponsive / not found - any chance you can fix this? The AS2 PRs (hopefully soon to be RECs) link directly to (as I'm sure you know :) )

  • sigh*

<scribe> scribenick:ben_thatmustbeme


approval of minutes form last week

<eprodrom> PROPOSED approve as minutes for 9 May 2017 telecon

<scribe> scribe: Ben Roberts

<cwebber2> +1

<eprodrom> +1

<aaronpk> +1

<tantek> I think I was chair, other than that +1 :)

<tantek> who is SV_MEETING_CHAIR?!?


<eprodrom> RESOLVED approve as minutes for 9 May 2017 telecon

RESOLUTION: approve as minutes for 9 May 2017 telecon

charter extension and rechartering

eprodrom: we voted to request 6 month extension and we feel we would do a better quality of work

sandro: its not going to be as smooth as i was hoping, last week i had the impression it would be a straight forward thing, turns out not to be the case in that...
... A couple years ago there was a rule added that groups cannot be extended past 6 months without talking to the AB

tantek: some other groups have hit this as well in the past

sandro: if you could find examples, that would be helpful

<Loqi> Tantekelik made 1 edit to Socialwg/2017-05-16

<Loqi> Tantekelik made 1 edit to Socialwg/2017-05-09-minutes

sandro: bottom line is we have to ask the AC for this, there is no specified response on how this works

tantek: in the past other groups just rechartered instead of asking for another extension for the AC
... the only other I can think of is HTML media and that was a whole other issue
... we might be the first group to try to do this without a charter scope change and without any formal objections in our history
... it doesn't normally make a lot of sense to just do this short extension again

sandro: i don't feel like its really proper for us to recharter right now as it would mean we have to figure out new scope in the next week

tantek: whats driving this is really increased interest in our existing scope, we've got 2 RECs out, 2 more that are about to become RECs, we're seeing developer interest spiking this calendar year, multiple different specifications
... so we want to extend our charter to take that developer input into consideration for our specs that are in CR and our specs that are in PR / REC there is a chance to do revisions where developers are evolving that technology
... i believe that is all in our current scope
... the specs with additional extensions that developers are working on as well
... thats how i would explain it to W3M (?) myself
... the first extension was that we needed time to finish wrapping up our specs, but the situation has changed now

sandro: i have no trouble convincing W3M about that, whats left is we need to convince AC, but there is no precedent for what we need on that
... unless there is some issue the AC has with it, i think it seems like this will happen

tantek: i think if we can craft the message I think we will mostly see positive response from the AC

sandro: i don't want to take up more WG time on this
... you and i can work out the details
... i do feel like with the uncertainty here we should TRY to get to CR exit by the time our charter is up
... and make our little back-log of things if we get more time
... i think thats the responsible way to approach it

tantek: i think thats true

sandro: i should be able to get it to the AC, maybe tomorrow

eprodrom: it sounds like we have a plan going forward, is there anything else needed from the WG?

sandro: i don't think so

meeting next week

<cwebber2> +1 to meeting next week

eprodrom: any objections to meeting on the 23rd?
... i cannot make it, but tantek would be chairing anyway
... with no objections, lets plan on doing next week


sandro: AS2 and Micropub, amy and i sent of the transition request for those yesterday

i don't see any issue with as2 going to REC next week

micropub might have some small issues, not a lot of response, and one minor, non-formal objection

sandro: i'll let you know if those turn out to be issues

aaronpk: i just had a quick quesiton, in terms of not a lot of responses, is there something we can do for that? how can we help?

sandro: if it turns out that its not enough, i will go out and try to get more people to respond
... in both these cases, if theres a problem, i'll let you know

tantek: its odd in that we don't see much use in member companies, there are things like launching with this as their main client API.
... i've never seen this happen at W3C before

sandro: again, we'll deal with it if it becomes an issue

tantek: its more just a question, for me, of is it good for the web

eprodrom: we have one outstanding issue on AS2, it feels like an either/or issue about breaking out a section of the vocab document. are we at a point where making editorial edits is not worth the work?

sandro: its ok to still do changes, sort-of, its also kind of annoying, so i don't know how important it is

eprodrom: i feel like its useful for adoption, its interesting but not crucial so i will try to resolve this with amy

sandro: if you can do it sooner, the better, i'd like to publish next tuesday

tantek: we don't usually see editorial changes between PR and REC, so its probably better to err on the side of safety and it may be better to leave it to the errata

eprodrom: basically its that amy suggested a primer and one section of the doc has a lot of implementation notes, she suggested we edit those out and move it to a primer

<Loqi> Cwebber2 made 1 edit to Socialwg/2017-05-16

eprodrom: the other option is that if we want to create a primer after this, it would be duplicating some of this in a seperate primer

tantek: it think it would be better to not make edits that are not in response to AC comments

eprodrom: i'm comfortable with that, i will discuss it with amy


tantek: i haven't been able to make any edits on PTD

cwebber2: i have been working on the test suite, i've gotten pretty far on the client to server side stuff, but the test suite succeeded at being a test suite and i found some parts i was missing
... lets have the goal be to send out requests for people to test out that portion of the test suite next week

eprodrom: next up is websub


aaronpk: for websub, we have 1 issue that i want to talk about, its here (irc link)

<Loqi> [aaronpk] #102 Why should subscribers return 2xx on invalid signatures?

the spec says to always return a 2xx even with invalid signatures

scribe: no one had any issues with allowing it to return anything, 4xx for issues for example
... it doesn't effect interop
... i went ahead and made the change in the spec already

eprodrom: is this change reflected in the test suite?

aaronpk: yes

eprodrom: have you seen any implementors that depended on 2xx reply?

aaronpk: if it has a correct signature, it still requires 2xx, it only effects bad signatures which you can't really test for

and some people wanted it to be able to return 410 gone for example

eprodrom: it makes a lot of sense to me, i've never been in the position of not being able to return accurate error codes, so this makes sense

tantek: makes sense to me too

aaronpk: i don't think we need a full vote, just having it noted in the minutes

tantek: do we need to make a new CR?

aaronpk: this was merged in 7 days ago, the question is it a normative features

sandro: because its loosening the constraints, it could technically be checking it for that exact error code...

aaronpk: but it would have to be sending a bad request

sandro: yeah, this seems to be a non-substantive change

<sandro> sandro: normative but couldnt-break-anything

tantek: this is non-breaking, non-substative, but still a normative change

aaronpk: there are 2 other issues in that discussion, there are some added text for how to migrate subscribers


<Loqi> [Alkarex] #106 Suggestion: Use HTTP 410 Gone

aaronpk: the other potentially problematic one is about (link in irc) which is that hubs recognize 410
... it comes out of the previous issue of how to handle specific responses from the error after it gets a 410

this would change logic

<aaronpk> this would change what hubs are required to do

but it can be an extension as the spec would now allow 410

eprodrom: aaronpk just to be clear
... are we looking to a resolution to this?

aaronpk: no, i just wanted to point out that it is an open issue
... is it even something we should consider to the spec or just leave it as an extension

<ben_thatmustbeme> i'm +1 extension

tantek: again, if this is going to reset our clock perhaps we should leave it as a spec extension until we know for sure we have a group extension

sandro: that makes sense, i'm still waiting to hear, there are a lot of pubsubhubbub implementations out there, i'm waiting to see how many of them are compliant
... i'm looking for ones like github or mastodon

lots of people have used off the shelf pubsubhubbub for years

our goal was to not break things there

eprodrom: GNUSocial would also be a good one here

cwebber2: would this be something good to bring to the community group friday

sandro: the testing against websub, definitely

tantek: i think we should ask ALL implementations showing up to submit implemention reports, since i think they all support it

eprodrom: anything else on websub?

tantek: if we want a new CR, we need a resolution


tantek: do you have a changes section?

<eprodrom> PROPOSED publish new CR for WebSub based changes listed at

<ben_thatmustbeme> can we add bit about not resetting the clock

<sandro> WG believe and this request is contingent on it not restarting CR clock

<cwebber2> +1

<sandro> +1

<aaronpk> +1

<tantek> +1 noting that informative guidelines reflect current webdev discussions of migration being important, and the only normative change is non-subtantive

<ben_thatmustbeme> +1

<eprodrom> +1

<eprodrom> RESOLVED publish new CR for WebSub based changes listed at

RESOLUTION: publish new CR for WebSub based changes listed at

<aaronpk> 🎉

eprodrom: any other items?
... any other doc statuses?

tantek: assuming we get REC published next week, we should ping amy to update SWP accordingly
... i'm fine giving a blanket, SWP can be updated any time a document status changes



cwebber2: we have a meeting this friday (link in irc)


we have also not settles on a weekly time

people voted in the last meeting they want it weekly, but i'd encourage you to fill it in

this would be for every week

eprodrom: that leaves us with, i have a note about cwebber2's representation

cwebber2: i joined the verifiable claims WG, but that does some legal things
... their call is at the exact same time as this group

sandro: i know we picked this time to allow for amy, but she is on the move now
... she is in japan now then to europe next week, then completely unknown

<cwebber2> +1

eprodrom: we need to set up a poll for our weekly meetings

<cwebber2> thanks!

<eprodrom> trackbot, end meeting

<tantek> ben_thatmustbeme++ for minuting!

<Loqi> ben_thatmustbeme has 70 karma in this channel (220 overall)

Summary of Action Items

Summary of Resolutions

  1. approve as minutes for 9 May 2017 telecon
  2. publish new CR for WebSub based changes listed at