18:02:16 RRSAgent has joined #social 18:02:16 logging to http://www.w3.org/2015/11/10-social-irc 18:02:16 present+ rhiaro 18:02:29 present+ akuckartz 18:02:37 eprodrom has changed the topic to: http://www.w3.org/wiki/Socialwg/2015-11-10 18:02:37 present+ jasnell,tantek,sandro,wseltzer,rhiaro 18:02:45 present+ csarven 18:02:48 Present+ Rob_Sanderson 18:02:55 present+ 18:02:59 present+ wilkie 18:03:15 present+ ben_thatmustbeme 18:03:25 ben_thatmustbeme: see https://www.w3.org/wiki/Socialwg/2015-11-10#Agenda 18:03:46 present+ cwebber2 18:03:46 somehow i missed it 18:04:02 present+ Benjamin_Young 18:04:03 Rene sent regrets to the list i believe 18:04:24 I can scribe 18:04:26 I haven't in a while. 18:04:27 Arnaud, can I ask a favor? 18:04:43 Can you copy over those agenda items? 18:05:41 Present+ eprodrom 18:05:46 zakim, who is here? 18:05:46 Present: Arnaud, csarven, rhiaro, aaronpk, shanehudson, sandro, elf-pavlik, kevinmarks, wilkie, eprodrom, jasnell, ben_thatmustbeme, cwebber, tantek, hhalpin, james, tsyesika, 18:05:47 I'll scribe 18:05:50 ... wseltzer, akuckartz, shepazu, Rob_Sanderson, Shane_, rene, cwebber2, Benjamin_Young 18:05:50 On IRC I see RRSAgent, akuckartz, KevinMarks, csarven, azaroth, the_frey, tantek, jaywink, eprodrom, jasnell, bblfish, melvster, Arnaud, wilkie, cwebber2, ben_thatmustbeme, 18:05:50 ... bitbear, shepazu, bret, tommorris_, tsyesika, jet, aaronpk, raucao, Loqi, bigbluehat, ElijahLynn, tessierashpool_, rhiaro_, rrika, Zakim, dwhly, pdurbin, rhiaro, sandro, 18:05:51 ... trackbot, wseltzer 18:06:36 trackbot, start meeting 18:06:38 RRSAgent, make logs public 18:06:40 Zakim, this will be SOCL 18:06:40 I do not see a conference matching that name scheduled within the next hour, trackbot 18:06:41 Meeting: Social Web Working Group Teleconference 18:06:41 Date: 10 November 2015 18:06:48 scribenick: cwebber2 18:07:06 eprodrom: unfortunately I have limited connectivity, may be slow on IRC 18:07:14 eprodrom: first item is to review/approve last week's minutes 18:07:20 https://www.w3.org/wiki/Socialwg/2015-11-03-minutes 18:07:44 eprodrom: I'd like to ask to review the minutes 18:07:48 +1 18:07:50 cwebber2: +1 18:07:56 +1 18:07:57 +1 18:08:01 Proposal: approve minutes for 2015-11-03 18:08:04 eprodrom: any objections or -1, please say so now 18:08:11 +1 18:08:17 eprodrom: I'd like to call this resolved 18:08:30 eprodrom: and we're ready to move right along. 18:09:09 eprodrom: one note about this week's agenda is that we had some agenda items from last week that did not get handled, so according to the FIFO procedure we're following, we have now moved them all to today's agenda? 18:09:14 eprodrom: is that right Arnaud ? 18:09:23 Arnaud: reload your page and I think it should look reasonable 18:09:40 audio arrives here somewhat broken 18:09:48 eprodrom: first item at top of page is simplified microformats 2 json format, JS2 18:09:51 aaronpk is not here 18:09:54 s/js2/jf2 18:09:58 eprodrom: I think we've discussed it, maybe we can review quickly 18:10:01 Alehors made 1 edit to [[Socialwg/2015-11-10]] https://www.w3.org/wiki/index.php?diff=86760&oldid=86758 18:10:02 Rhiaro made 1 edit to [[Socialwg/2015-11-03-minutes]] https://www.w3.org/wiki/index.php?diff=86761&oldid=86650 18:10:06 Arnaud: aaronpk is not here, we need to roll over 18:10:22 eprodrom: thanks for reminding me, we'll roll over to next week and address AS 2.0, is james on? 18:10:24 jasnell: I'm here 18:10:29 eprodrom: great let's work through our list 18:10:39 eprodrom: a couple have come from elf, is elf on the call? 18:10:55 Arnaud: the first few are from last week because elf was not on 18:11:01 jasnell: I'm pulling it up now to look at the list 18:11:07 Arnaud: if elf is not on, we will skip again 18:11:19 The Blog object type probably rolls in to the extension vocab discussion 18:11:19 tantek: didn't see him dial in, don't see him on present list from Zakim 18:11:46 eprodrom: I'm not sure we need the proposer to be here to deal with these. should se address now or wait till elf is here? my feeling is to wait, but we may pass off for multiple meetings 18:12:00 tantek: if james has a succinct resolution to propose, otherwise I'm ok with postponing 18:12:08 jasnell: I have my own biased opinion, best to postpone 18:12:16 eprodrom: I think that makes sense 18:12:36 eprodrom: so we wil pass over 221, 223, 208, that brings us to 52 18:12:51 eprodrom: since this is a pretty contentious one, let's start with 52 18:13:11 jasnell: the short summary is that the AS spec currently defines an activity media type to identify docs 18:13:31 jasnell: the argument in the proposal is to provide the json-ld type 18:13:37 FYI: documented responses on this issue https://www.w3.org/wiki/Socialwg/Media_type_for_AS2 18:13:39 https://github.com/jasnell/w3c-socialwg-activitystreams/issues/52 18:13:39 jasnell: with profile 18:14:05 jasnell: my proposal is to treat them as equal, that the AS2 type SHOULD be treated as equivalent 18:14:14 jasnell: generally +1'ed, with 2 dissenteres, on the thread 18:14:14 but the "must" support is for "application/activity+json" right? 18:14:20 q+ 18:14:34 sandro: there's no requirement on server to say "I want activitystreams" 18:14:52 q+ to disagree with Sandro 18:14:53 jasnell: to use json-ld media type you have to provide a profile parameter 18:15:01 sandro: you can't set one on a ? header 18:15:03 q+ 18:15:14 jasnell: you can but many implmenetations don't, it's an optional feature 18:15:19 s/? header/accept header/ 18:15:27 sandro: the spec says you can't put it there, unless I'm wrong 18:15:43 sandro: so unless I'm wrong, you can't ask for activitystreams from someone who doesn't know that media type 18:15:49 I believe you can put whatever you want in Accept, but it isn't standard 18:15:57 sandro: you shouldn't do this, but oif you really want to, you're in bad shape 18:16:12 jasnell: my proposed solution is to treat them as equivalent 18:16:18 Q? 18:16:22 as an implementor i don't really want to look for 2 possible types of rel-alternates 18:16:28 jasnell: with the AS type being a MUST, and equivalency is a SHOULD 18:16:33 as long as it is: ; something = "some string" 18:16:40 tantek: just to understand.. oh ok james may have just clarified my question 18:16:43 Ack tantek 18:16:57 I like the compromise here 18:17:02 "application/activity+json" is a MUST, proposing a SHOULD for profile equivalence 18:17:05 +1 18:17:08 tantek: is that right? 18:17:12 jasnell: yes 18:17:15 http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html 18:17:16 tantek: I can live with that 18:17:19 +0 can live with that 18:17:24 q+ 18:17:24 ack azaroth 18:17:25 eprodrom: azeroth? 18:17:26 azaroth, you wanted to disagree with Sandro 18:17:40 azaroth: this is the updated HTTP/1.1 for that bit https://tools.ietf.org/html/rfc7231#section-5.3.2 18:17:57 azaroth: I disagree with sandro's interpretation... there's clearly a parameter in the media range, don't see why you couldn't fit in a parameter. If that is true though, that's something to keep in mind though 18:17:58 +q 18:18:08 https://tools.ietf.org/html/rfc7231#section-5.3.2 18:18:11 azaroth: if we can't do that, there's no point defining an equivalence 18:18:25 sandro: I think you're right 18:18:32 q+ 18:18:32 sandro: I wouldn't have guessed that was media range 18:18:34 q- 18:18:43 "The media-range MAY include media type parameters that are applicable to that range. " 18:18:48 cwebber2: fwiw I think manu said they specifically added profiles for this purpose 18:18:51 the implementation support for content negotiation using parameters is sketchy at best 18:18:53 ack akuckartz 18:18:57 akuckartz: my problem is we've been discussing this for some time now 18:19:04 akuckartz: in favor of media type and against profile 18:19:19 akuckartz: all the time one argument was refuted... and once again we have a new argument for resolving this 18:19:32 akuckartz: we created a wiki page to collect all arguments we have 18:19:33 q+ 18:19:35 there are popular implementations, for instance, that do not properly handle case sensitivity of comparison in parameter values 18:19:41 there are others that ignore parameters entirely 18:19:45 akuckartz: I suggest we provide new page with reasons for profile 18:19:46 the wiki page mentioned https://www.w3.org/wiki/Socialwg/Media_type_for_AS2 18:19:53 akuckartz: collect for or against 18:20:02 Alehors made 1 edit to [[Socialwg/2015-11-10]] https://www.w3.org/wiki/index.php?diff=86762&oldid=86760 18:20:22 akuckartz: last argument was put forward against json-ld media type was problem of list of lists... even that is not against json-ld + profile 18:20:24 I honestly do not care what the media type is so long as it is unique, and preferrably only 1 media type 18:21:00 the_frey_ has joined #social 18:21:00 eprodrom: it sounds like you have a no-compromise idea, what you proposed as as a process is that there is a list of reasons for having a media type for activitystreams, and those will bve refuted one by one, and when th full list is completed, that's when the wg handles it 18:21:12 akuckartz: I'm not against it, I don't see the reasons in support of a new media type 18:21:19 I don't think this is worth more time frankly, so much time has been wasted on such plumbing. Perhaps we can do a strawpoll to see where the group is overall? 18:21:24 akuckartz: I'm not against comporomise if necessary 18:21:47 eprodrom: could you sketch out to see what way you think the w3c should go forward? would it be just json-ld with profile? 18:21:58 and wait what about for people who aren't using json-ld..? 18:22:11 q? 18:22:12 akuckartz: yes, then we could have just json-ld... it's more effort than to handle profile 18:22:14 there's nothing preventing the use of the application/ld+json media type being used by those who want to...the document could still use the AS2 @context value 18:22:19 q- 18:22:30 eprodrom: what I'd like to do is to continue this discussion in middle of discussion 18:22:32 this was already about a compromise where people didn't want to see explicit json-ld 18:22:38 eprodrom: I for one would like to get it completed 18:22:40 perhaps a straw poll A, B, C? of the options? 18:22:49 eprodrom: I'd like to get at est of the working group 18:22:50 q? 18:22:52 ack eprodrom 18:23:12 ack eprodrom 18:23:14 eprodrom: to ack myself, I'd like to ask james if we're dealing with multiple fallback media types, should we put application/json on that list 18:23:16 q+ to note application conformance re list of lists 18:23:29 eprodrom: do we support that in our list 18:23:32 jasnell: we don't have to 18:23:42 jasnell: that's defined as part of the rfc in support of the suffix 18:24:05 jasnell: I believe anything using the +json suffix automatically has application/json as fallback 18:24:08 eprodrom: oh great 18:24:11 jasnell: making sure right now 18:24:14 jasnell: yup 18:24:18 jasnell: *reads spec* 18:24:27 that certainly makes sense 18:24:28 jasnell: it's built into the rules 18:24:34 eprodrom: great we get that for free 18:24:46 KevinMarks has joined #social 18:24:54 q? 18:24:54 ack cwebber 18:24:57 scribenick: rhiaro 18:24:57 I'll scribe 18:25:00 oh ok! 18:25:03 https://tools.ietf.org/html/rfc6839 18:25:11 cwebber: I think we're as close to consensus as we're going to get on this 18:25:25 ... It seems fromt he LD side we're going to be able to have a guarantee that things can be in an LD format 18:25:29 ... and having an alias to a profile is great 18:25:40 ... the microformats community doesn't necessarily want things to be JSON-LD which is fulfilled by this 18:25:47 ... Seems like we're in a good space of both sides being happy 18:26:05 ... In response to the previous questiona bout the justification, there has been a lot of conversation in thsi group who don't necessarily want to view things as LD 18:26:14 ... and want to have a fallback where they can guarnatee that things are simple json 18:26:26 ... and being able to provide a way to say that things are handled here and that is handled as the profile 18:26:31 +1 to the proposed comprise and for moving forward 18:26:33 +1 get this over with 18:26:34 ... I'd prefer not to lose this opportunity to have the group agree 18:26:35 sandro: 7231 is actually clearer than 2616 (yay, progress!) https://tools.ietf.org/html/rfc7231#section-5.3.2 18:26:36 woot 18:26:43 eprodrom: well said! 18:26:44 q? 18:26:49 ack azaroth 18:26:49 azaroth, you wanted to note application conformance re list of lists 18:27:02 azaroth: I'm also +1 to the proposal from James 18:27:02 q? 18:27:02 ack azaroth 18:27:09 scribenick: cwebber 18:27:22 azaroth: one concern is the list of list for geojson list of lists 18:27:35 q+ 18:27:35 azaroth: that one can't just be a profile of json-ld, because json-ld won't provide those things 18:27:37 please document concerns 18:27:46 cwebber2: jasnell: that's just an extension anyway to AS2 though right? 18:27:53 uhhhhh, but then those wont work in json-ld at all, so they break other things....? 18:28:07 eprodrom: I have one more question for james, which is we're using this we're using this for AS1 and AS2, is this compatible 18:28:22 jasnell: it's independent of that. activity+json is specifically AS2 18:28:28 eprodrom: what I'd like to do is put up a proposal 18:28:51 is there a media type for AS1? I don't think I'm using one. 18:28:59 from earlier: jasnell: with the AS type being a MUST, and equivalency is a SHOULD 18:29:00 cwebber2: wilkie: :) 18:29:31 the proposal is to say that implementers SHOULD treat application/ld+json; profile="http://www.w3.org/ns/activitystreams" as being equivalent to application/activity+json 18:29:45 but that application/activity+json is the media type for AS2 documents 18:29:51 jasnell: I just sent it 18:30:02 Sandro made 1 edit to [[Socialwg/Media type for AS2]] https://www.w3.org/wiki/index.php?diff=86763&oldid=86653 18:30:05 eprodrom: that looks good, can you copy/pasta that with PROPOSAL: 18:30:09 jasnell: mmmmhmmmmm 18:30:19 cwebber2++ 18:30:22 cwebber2 has 54 karma 18:30:25 PROPOSAL: the proposal is to say that implementers SHOULD treat application/ld+json; profile="http://www.w3.org/ns/activitystreams" as being equivalent to application/activity+json, but that application/activity+json is the media type for AS2 documents 18:30:35 cwebber2: +1 18:30:43 ? s/PROPOSAL/PROPOSED 18:30:46 cwebber2++ 18:30:48 +1 18:30:49 +1 18:30:50 cwebber2 has 55 karma 18:30:52 +1 18:30:52 +1 18:30:53 +1 18:31:01 +0 can live with as noted previously 18:31:08 -1 (not convinced) 18:31:13 0, i would prefer one type, but that acceptable 18:31:15 +0 can live it, not excited 18:31:23 cwebber2: akuckartz: is that -1 or -0? 18:31:26 q? 18:31:26 +1 18:31:40 it is a -1 18:31:58 sandro: would you rather have this spec never be published 18:32:02 sandro: than go forward 18:32:09 sandro: you'd like to have several more dozen person hours on this 18:32:11 q? 18:32:11 ack akuckartz 18:32:18 akuckartz: well the problem is there's several sequential arguments put forward 18:32:25 akuckartz: the list of list is invalid 18:32:36 0 may not be a stakeholder so wont state a view, tho I've not seen arguments in favour of application/activity+json so it seems a strange vote 18:32:47 akuckartz: you can overrule my objection but I'm not convinced and I think it's the wrong decision. You can overrule 18:33:00 sandro: do you have a proposal 18:33:05 akuckartz: I'm not against a common solution 18:33:12 cwebber2: that does sound like you are 18:33:17 cwebber2: by -1 18:33:24 akuckartz: what if we reverse it 18:33:29 cwebber2: oh okay 18:33:38 cwebber2: that's a clearer proposal 18:34:00 PROPOSED; the media type for AS2 is application/ld+json; profile="http://www.w3.org/ns/activitystreams" and implementation SHOULD treat the media type application/activity+json as equivalent 18:34:08 q? 18:34:10 +1 18:34:11 +1 18:34:11 +0 18:34:11 +1 18:34:20 +0 18:34:20 +1 18:34:21 -1 requiring JSONLD is a non-starter 18:34:24 +0 18:34:25 +0 18:34:26 -1 18:34:28 cwebber2: I could live with this, but I feel like it's not as good. 18:34:36 jasnell: does the effect as1 people? 18:34:42 s/:/,/ 18:34:46 No, this has no effect on as1 18:34:46 -1 agreed with Tantek 18:34:49 cwebber2: it's obvious that there's stronger objections from many parts of the group 18:35:01 -0 jsonld seems more required. that'll cause implementation schisms I'd think 18:35:05 -1 18:35:06 cwebber2: whereas previously there was closer to a happy compromise between many parts of the group 18:35:16 +0 18:35:16 q? 18:35:26 If json-ld can't support lists of lists it needs to be fixed 18:35:30 cwebber2: again I think we're going to lose an opportunity 18:35:39 also what wilkie said - it becomes much harder to go pitch / ask developers to implement something supposedly simple JSON but says MUST support JSONLD mediatype. 18:35:39 sandro: do you think you understand where these -1s are coming from 18:35:44 sandro: can you change their mind 18:35:57 azaroth: there's a third option which is to reduce it to one 18:36:06 azaroth: which is to reduce to one option 18:36:08 q+ 18:36:12 s/azaroth/akuckartz/ 18:36:13 s/azaroth/akuckarz 18:36:17 sandro: no -1s would go away 18:36:20 q? 18:36:21 similarly there's another option to *only* use application/activity+json 18:36:21 azaroth: some not all 18:36:21 KevinMarks If List of Lists is a requirement? One can raise arbitrary structures and say that it is a non-starter. 18:36:42 any user stories for lists of lists? 18:36:49 csarven: Yes, for GeoJSON 18:37:07 List of lists is valid json. Json-ld should be able to handle them or stop calling itself json 18:37:08 A list of list is valid json so it's ridiculous to specify a media type that disallows valid json 18:37:12 azaroth Is GeoJSON-like data in the UC? 18:37:39 scribenick: rhiaro_ 18:37:49 cwebber2++ 18:37:50 cwebber: we should take the opportunity to find a compromise that more people are happy with 18:37:52 cwebber2 has 56 karma 18:37:54 KevinMarks That argument has 0 barring on what's being discussed. 18:37:57 alternatively we could propose a straw poll on *only* "application/activity+json" for the folks who prefer "just one mediatype" if they would like to consider it 18:38:03 ok, I change to -0 18:38:03 csarven: Good question, but I've heard it asserted that it's in scope. 18:38:05 rhiaro++ 18:38:06 Maybe we should start calling it JSONish-LD 18:38:07 scribenick: cwebber2 18:38:08 rhiaro has 182 karma 18:38:14 q? 18:38:17 -q 18:38:34 eprodrom: note ^^^ akuckartz above 18:38:40 eprodrom: to mf2 people, is there any way to rephrase #2 proposal where you'd be happy 18:38:46 azaroth If that's in the scope, a URL would be nice. When was it initially asserted? 18:38:51 tantek: I think you got the compromise you're aksing for if you look agove 18:38:53 sandro: what's that 18:39:01 eprodrom: which is to just overrule it? 18:39:06 tantek: no, he changed it to -0 18:39:27 eprodrom: I see it, no if that's the case, if there are no other objections, I'd like to mark it as resolved 18:39:30 cwebber2: yes hold on 18:39:36 honestly i dont think saying "json ld is a non starter" is any kind of argument 18:39:44 when its a W3C REC 18:39:54 RESOLVED: the proposal is to say that implementers SHOULD treat application/ld+json; profile="http://www.w3.org/ns/activitystreams" as being equivalent to application/activity+json, but that application/activity+json is the media type for AS2 documents 18:39:58 melvster lots of W3C RECs are non-starters today. 18:40:05 melvster, not every W3C Rec is welcome in every context 18:40:09 melvster++ 18:40:11 eprodrom: great, I'd like to give my congrats to the group 18:40:13 melvster has 19 karma 18:40:18 eprodrom: if nothing else is done today, this is a huge step forward 18:40:25 cwebber2: yes I agree, congrats all 18:40:32 eprodrom: I'd like to move on to the extension vocab 18:40:38 akuckartz++ for making a clear argument, but also being willing to compromise 18:40:42 akuckartz has 3 karma 18:40:43 jasnell: there's a proposal to drop story / folder / alboum 18:40:59 jasnell: my proposal is to set up an extension vocab to include those and some other extensions, like Blog 18:40:59 +1 keep dropping things 18:41:01 If some W3C Recs are not welcome, I think it is reasonable to say that mf2 is a non-starter. 18:41:10 +1 drop all named collections, honestly 18:41:11 jasnell: album, blog, story, form(?) and wiki 18:41:21 http://ns.jasnell.me/socialwg/ 18:41:22 jasnell: something that's not core but should be defined, basically 18:41:29 jasnell: here's a link to a strawman draft 18:41:39 q? 18:41:55 melvster++ 18:41:58 melvster has 20 karma 18:42:07 melvster, in voting mostly what counts is if someone is willing to lie in the road. That's what matters, not what people think on the wiki. 18:42:10 eprodrom: what about making this as any kind of ? for the core 18:42:18 jasnell: i could address that if I understood it 18:42:35 eprodrom: instead of 3 docuemnts, we have 2 documents, one with core classes and types, secondary doc is the extension 18:42:41 q+ to suggest dropping all the named collections as wilkie said, and leave it up to those that support those collections to put forth an extension draft for consideration as an ED 18:42:41 jasnell: I'm not sure that ? very much 18:42:52 s/?/buys us/ 18:42:52 jasnell: the core is the minimal core that folks should understand 18:42:57 tantek++ 18:43:01 tantek has 257 karma 18:43:15 q? 18:43:16 jasnell: what the proposed extensions are... we'll have an extension vocab anyway 18:43:24 jasnell: and still have a home for these other things 18:43:28 eprodrom: ok I buy that 18:43:41 tantek: I want to put forth a proposal wilkie brought up, which is to drop all the named collections 18:43:45 tantek: I strongly support this 18:44:07 tantek: and rather than leave this a responsibility for jasnell, leave this as a separate spec for someone else to do 18:44:10 ack tantek 18:44:10 tantek, you wanted to suggest dropping all the named collections as wilkie said, and leave it up to those that support those collections to put forth an extension draft for 18:44:13 ... consideration as an ED 18:44:13 jasnell: that' thes same proposal 18:44:28 tantek: I'm saying don't commit to it here, let someone else do it with any commitment to take it to rec trac 18:44:44 tantek: I think that's a bit different than what was on IRC and the minutes 18:44:46 my proposal is for the extension to be a note 18:44:48 not a rec track 18:44:53 q? 18:45:00 tantek: I think that burden should not fall to you, but to the strong advocates of it 18:45:02 I am an advocate of it 18:45:03 :-) 18:45:06 :) 18:45:09 :) 18:45:15 eprodrom: if we have no more discussion, then maybe we can resolve this 18:45:16 kevinmarks has joined #social 18:45:18 and I've already started a draft ... http://ns.jasnell.me/socialwg/ 18:45:23 KevinMarks_ has joined #social 18:45:36 eprodrom: maybe along the lines of a simple... accept that it.... well move story/folder/album to a new extension vocab 18:45:40 cwebber2: hold on 18:45:48 cwebber2: what are all the terms 18:45:54 I'll write it... one sec 18:45:58 cwebber2: story/folder/album/wiki? 18:46:10 cwebber2: ok I'll let jasnell do it :) 18:46:11 PROPOSED: move Folder, Album, Story out of core into separate extension vocabulary 18:46:12 jasnell, yes a note-track I think would be a better start 18:46:19 and we should be explicit about that 18:46:20 cwebber2: +1 18:46:25 +1 18:46:27 +1 18:46:28 +0 18:46:32 +1 18:46:53 eprodrom: great 18:46:56 +1 with caveat that the "separate extension vocabulary" is at best a note (we can discuss more later) 18:47:06 +1 18:47:07 eprodrom: I think that makes sense tantek 18:47:08 +1 18:47:16 eprodrom: barring any objections, marked as resolved 18:47:19 +1 18:47:27