2014-09-02 UTC
bblfish, nicolagreco, Arnaud1, Arnaud, bblfish_, deiu, barnabywalters, cmhobbs, elf-pavlik, Shane, dret and jasnell joined the channel
tantek joined the channel
Zakim joined the channel
RRSAgent joined the channel
# 16:58 RRSAgent trackbot, access must be one of public, group-strint-submission, offline-webapps-workshop-program-committee, group-webmobile-chairs, group-rdf-val-pc, alumni, group-payment-workshop-submissions, wstar, group-digipub-chairs, member, memberSearchers, group-csv-chairs, wsridirectors, i18n, valid, group-strint-pc, webcrypto, offices, w3f, mlw, group-wot-workshop-pc, team, webandtv-moderators, ab, group-share-psi, group-payment-workshop-pc, memberEditors, w[CUT] # 16:58 Zakim ok, trackbot; I see T&S_SOCWG()1:00PM scheduled to start in 2 minutes # 16:59 Zakim I am sorry, deiu; I do not know a number for deiu # 16:59 Zakim I am sorry, deiu; I do not know a number for matt ShaneHudson joined the channel
# 17:01 jasnell thinks Zakim partied a bit too hard over the U.S. 3-day weekend # 17:01 Zakim T&S_SOCWG()1:00PM has not yet started, Arnaud # 17:01 Zakim On IRC I see ShaneHudson, RRSAgent, Zakim, tantek, bblfish, jasnell, elf-pavlik, cmhobbs, deiu, Arnaud, wilkie, shepazu, rektide, KevinMarks, Tsyesika, bryan, mattl, tommorris, # 17:01 Zakim ... Loqi, rhiaro, bret, aaronpk, kylewm, trackbot, botie, sandro, wseltzer_igf, oshepherd MattMarum joined the channel
# 17:01 Zakim T&S_SOCWG()1:00PM has not yet started, deiu # 17:01 Zakim On IRC I see MattMarum, ShaneHudson, RRSAgent, Zakim, tantek, bblfish, jasnell, elf-pavlik, cmhobbs, deiu, Arnaud, wilkie, shepazu, rektide, KevinMarks, Tsyesika, bryan, mattl, # 17:01 Zakim ... tommorris, Loqi, rhiaro, bret, aaronpk, kylewm, trackbot, botie, sandro, wseltzer_igf, oshepherd # 17:02 Zakim T&S_SOCWG()1:00PM has not yet started, bblfish MarkC joined the channel
# 17:02 Zakim On IRC I see MattMarum, ShaneHudson, RRSAgent, Zakim, tantek, bblfish, jasnell, elf-pavlik, cmhobbs, deiu, Arnaud, wilkie, shepazu, rektide, KevinMarks, Tsyesika, bryan, mattl, # 17:02 Zakim ... tommorris, Loqi, rhiaro, bret, aaronpk, kylewm, trackbot, botie, sandro, wseltzer_igf, oshepherd # 17:02 Zakim ok, Arnaud; that matches T&S_SOCWG()1:00PM # 17:03 Zakim On the phone I see ??P1, ??P2, Arnaud, +1.303.204.aaaa, jasnell, aaronpk, deiu, ??P7, [IPcaller], MarkCrawford, [IPcaller.a] # 17:03 Zakim the conference code is 7625 (tel:+1.617.761.6200 sip:zakim@voip.w3.org), ShaneHudson # 17:04 Zakim sees on the phone: ??P1, ??P2, Arnaud, +1.303.204.aaaa, jasnell, aaronpk, deiu, ??P7, [IPcaller], MarkCrawford, [IPcaller.a], ??P6, Sandro, ??P10 # 17:04 Zakim sorry, ShaneHudson, I do not recognize a party named 'P10' Dan_Romascanu joined the channel
hhalpin joined the channel
# 17:05 Zakim deiu, listening for 10 seconds I heard sound from the following: ??P6 (5%), jasnell (9%) # 17:05 Zakim On the phone I see ??P1, ??P2, Arnaud, +1.303.204.aaaa, jasnell, aaronpk, deiu (muted), ??P7, [IPcaller], MarkCrawford, [IPcaller.a], ??P6, Sandro, ShaneHudson, +1.412.370.aabb, tibor_katelbach joined the channel
# 17:06 Zakim sees ??P7, [IPcaller.a] on the speaker queue # 17:06 hhalpin hey everyone, I am in border control at an airport due to flight delay so dialing in late # 17:06 Zakim sorry, MattMarum, I do not recognize a party named '?PP7' # 17:06 Zakim sees ??P7, [IPcaller.a], ??P2 on the speaker queue # 17:06 Zakim sorry, MattMarum, I do not recognize a party named '??PP7' # 17:06 Zakim sees ??P7, [IPcaller.a], ??P2, [IPcaller] on the speaker queue # 17:07 Zakim sees ??P7, [IPcaller.a], ??P2, [IPcaller] on the speaker queue # 17:07 Zakim I see [IPcaller.a], ??P2, [IPcaller] on the speaker queue # 17:07 hhalpin trackbot, start meeting # 17:07 RRSAgent trackbot, access must be one of public, group-strint-submission, offline-webapps-workshop-program-committee, group-webmobile-chairs, group-rdf-val-pc, alumni, group-payment-workshop-submissions, wstar, group-digipub-chairs, member, memberSearchers, group-csv-chairs, wsridirectors, i18n, valid, group-strint-pc, webcrypto, offices, w3f, mlw, group-wot-workshop-pc, team, webandtv-moderators, ab, group-share-psi, group-payment-workshop-pc, memberEditors, w[CUT] # 17:07 Zakim ok, trackbot; I see T&S_SOCWG()1:00PM scheduled to start 7 minutes ago # 17:07 Zakim On the phone I see ??P1, ??P2, Arnaud, +1.303.204.aaaa, jasnell, aaronpk, deiu (muted), MattMarum, [IPcaller], MarkCrawford, [IPcaller.a], ??P6, Sandro, ShaneHudson, wilkie, rhiaro # 17:07 Zakim sees [IPcaller.a], ??P2, [IPcaller] on the speaker queue # 17:07 hhalpin just making sure! # 17:08 Zakim On the phone I see ??P1, ??P2, Arnaud, +1.303.204.aaaa, jasnell, aaronpk, deiu (muted), MattMarum, bblfish, MarkCrawford, [IPcaller.a], ??P6, Sandro, ShaneHudson, wilkie, rhiaro # 17:08 Zakim sees [IPcaller.a], ??P2, [IPcaller], bblfish on the speaker queue # 17:08 Zakim sees [IPcaller.a], ??P2, [IPcaller], bblfish on the speaker queue # 17:08 Zakim I don't understand your question, Arnaud. # 17:09 Zakim I see [IPcaller.a], ??P2, [IPcaller] on the speaker queue # 17:09 Zakim I don't understand your question, Arnaud. # 17:09 Zakim I see [IPcaller.a], ??P2, [IPcaller] on the speaker queue # 17:09 Zakim sees [IPcaller.a], ??P2, [IPcaller] on the speaker queue # 17:09 Zakim On the phone I see ??P1, elf-pavlik, Arnaud, +1.303.204.aaaa, jasnell, aaronpk, deiu (muted), MattMarum, bblfish, MarkCrawford, tommorris, ??P6, Sandro, ShaneHudson, wilkie, rhiaro # 17:10 hhalpin looks like quora to me # 17:11 Zakim sorry, tantek, I do not recognize a party named 'P19' # 17:11 Zakim On the phone I see tibor_katelbach, elf-pavlik, Arnaud, Tsyesika, jasnell, aaronpk, deiu (muted), MattMarum, bblfish, MarkCrawford, tommorris, ??P6, Sandro, ShaneHudson, wilkie, # 17:12 hhalpin quick note - if anyone is unclear of their invited expert status, just email me. # 17:13 hhalpin there was a systeam error when group was first set up so myself and chairs werent notified for a while. daveskiba joined the channel
# 17:13 hhalpin thus folks can easily slip through # 17:14 Zakim sorry, sandro, I do not recognize a party named '?P6' # 17:15 Zakim Not knowing who is chairing or who scribed recently, I propose MarkCrawford # 17:15 Zakim Not knowing who is chairing or who scribed recently, I propose wilkie # 17:15 jasnell I'll volunteer for either next week or the week after # 17:16 tantek reminder, when you speak, please say your name first, at least the first few times - it helps the scribe(s) a lot! # 17:16 ShaneHudson I will volunteer soon too, once I've got to grips with how it all works :) # 17:17 deiu quick reminder, use scribenick instead of scribe (easier for the parser) # 17:20 wilkie tantek: have minutes copy/pasted on wiki to retain history; people will fix/maintain them # 17:20 wilkie Arnaud: common scribe is a better tool for doing this. tell us of any better options. # 17:20 jasnell doubts that it's really worthwhile to debate this too much. # 17:21 hhalpin I recommend Skype dial out over SIP in general # 17:22 wilkie proposed: scribe copy/paste minutes to social wg wiki # 17:22 Arnaud RESOLVED: scribe copy/paste minutes to social wg wiki # 17:22 wilkie resolved: scribe copy/paste minutes to social wg wiki # 17:23 rhiaro I'll prepare myself a Skype account for next week # 17:23 hhalpin you can make munutes text by adding ,text to end of URI # 17:23 tantek regrets for next week's meeting. I will be on a flight. # 17:25 wilkie Arnaud: you need to register for TPAC, face-to-face meeting # 17:26 bblfish ah yes, I was looking to see what it would look like. There is no link to that from the irc channel info... # 17:26 wilkie Arnaud: we are scheduled (TPAC) to meet on 27-28 (October). the chairs will try to set up a schedule/agenda. # 17:27 hhalpin notes that registering early saves on flight and hotel # 17:28 wilkie Arnaud: it would be helpful to have a joint meeting with Social-IG # 17:28 tibor_katelbach I'm an observer # 17:28 tibor_katelbach I submitted request though # 17:28 tibor_katelbach wiating for validation # 17:29 MarkC given the large pool of folks involved in social in the bay area, we might need to have some control over the number of observers # 17:29 wilkie MarkC: social-IG will coordinate and come up with topics of interest to bring to social-wg for discussion/collaboration # 17:29 Zakim sees tantek, elf-pavlik on the speaker queue # 17:29 ShaneHudson Will the meetings have remote access for those of us that cannot make it? # 17:29 Zakim sees elf-pavlik, tantek on the speaker queue # 17:29 tantek btw that's how you put yourself at the end of the queue # 17:30 tantek if you want to defer to the next person in line # 17:30 jasnell As far as TPAC agenda is concerned, it would be good to give an overview of the AS2 draft at that time. I can walk people through the spec. # 17:31 jasnell Also, it would be good to discuss the Actions proposal and how it relates to alternatives such as Hydra and schema.org/Actions # 17:32 wilkie tantek: consider topics that are difficult to discuss online (email, phone) to discuss in the face-to-face. such as UI and visual discussion. # 17:32 MattMarum OpenSocial is looking to organize a separate event around same time most likely at UCSF. Still in discussion. # 17:33 wilkie tantek: let's avoid things such as 'here is a powerpoint' which could be done just as well in email # 17:33 wilkie tantek: had some success with working groups with 'unconference' or 'barcamp' type formats # 17:34 wilkie MarkC: agrees with tantek. james wants to do a walk-through of the specs. focus will be on the use-cases and how best to present those. # 17:34 tibor_katelbach maybe subjet proposals can be submitted in advance , to motivate interest # 17:35 tantek +1 to providing text summaries on the wiki *before* bringing up subjects in meetings # 17:36 wilkie Arnaud: we will set up a page for the face-to-face meeting and people can propose topics to discuss # 17:36 wilkie Arnaud: barcamp is good so that topics can be flexible, bad when schedules overlap # 17:38 wilkie Arnaud: last week we set up an agenda to separate sections of the AS specs into dedicated specs # 17:38 wilkie jasnell: 'url' is ubiquitious and should be there, was not in AS1 # 17:39 wilkie jasnell: some disagreement about 'id' vs 'uid'. discussion on github led to some reasonable compromise # 17:39 wilkie jasnell: people should read through the issue page on github and make comments either there or on the mailing list # 17:41 wilkie jasnell: between as1 and as2 changes were made to how links were handled. as2 wants to have a more generalized view of links. is this appropriate or do we need media links. # 17:42 oshepherd The fundamental thing is are we defining just an interchange format or a data model # 17:42 oshepherd Merging media links and objects make use as a data model more complex # 17:42 tantek any guidance we can take from AS2 implementations? # 17:43 oshepherd I know from AS1 impls, Pump.io, Impeller, Hubub (the latter two are Pump.io related apps fwiw), Idno/Known(?) use AS1 as their internal data model # 17:43 wilkie jasnell: (acking tantek in irc) not sure we can take much from them yet # 17:43 tantek I heard that there are only 2 AS2 implementations that could be used to inform (prev issue) and that neither really helps in that regard. # 17:45 wilkie jasnell: (wrt issue #12) html5 and microformats have taken a looser take on link relations, rfc5988 is stricter, which do we point to? jasnell prefers stricter rfc5988. # 17:45 tantek jasnell - thought your strict intersection "compromise" made sense # 17:45 tibor_katelbach +1 microformats # 17:46 bblfish yes, step backward on the higher picture seems a good idea to me. # 17:46 oshepherd I volunteer to raise data vs syntax model to mailing list? # 17:46 tantek and yes +1 for suggesting use of the microformats rel-registry (like HTML5) does # 17:46 wilkie jasnell: some of the larger issues (wrt AS spec) we should frame out and discuss (ie json-ld) # 17:47 bblfish we just saw that some of this needs to work on HTML5, it may need to work in JSON-LD, then also work on HTTP layer. # 17:47 wilkie MattMarum: when we looked at AS in the past, it was a data format that didn't specify much wrt how to retrieve the streams # 17:47 jasnell tantek: ok, I think it makes sense also so I think we're close on this one # 17:48 wilkie MattMarum: to me, we need to think about how these APIs work for getting AS and social business use-cases around these activities # 17:48 hhalpin zakim, dial hhalpin # 17:48 wilkie MattMarum: That's what the social api means to opensocial # 17:49 Zakim sees tibor_katelbach on the speaker queue # 17:49 Zakim sees tibor_katelbach, shepazu on the speaker queue # 17:49 oshepherd I raised the question a while back as to whether embedded "widgets" and the "API" are the same spec # 17:49 Zakim sees shepazu, aaronpk on the speaker queue # 17:49 wilkie tibor_katelbach: why was AS chosen compared to schema.org/Hydra # 17:49 jasnell I think we need to step back. The OpenSocial API approach works ok but is certainly not ideal by any stretch. # 17:50 tantek tibor_katelbach: anyone can bring alternative proposals to the group # 17:50 tantek right now, AS2 seems to have the most critical mass and active open editing / participation in this WG # 17:50 Zakim deiu, listening for 13 seconds I heard sound from the following: Hhalpin (11%) # 17:51 wilkie Arnaud: the WG has agreed to proceed with the publication of the spec. we can still choose to not go further with it beyond publication. reconsideration of previous decisions is valid. # 17:51 tantek jasnell, want to take this one as an FAQ to write-up? # 17:51 oshepherd I think tantek could give you some good reasons as to why not Schema.org :-) # 17:52 wilkie tibor_katelbach: I was using schema, and then found AS and found it quite complete. I have no problem with it, just wondering why # 17:52 bblfish yes, it's true the group only formally started meeting last week. # 17:52 tantek Arnaud, jasnell volunteered to write up FAQs for these questions. # 17:52 Zakim sees shepazu, aaronpk on the speaker queue harry joined the channel
# 17:53 wilkie Arnaud: anybody wanting to do a comparison and present a recommendation can do so, but we have to keep a timeframe in mind # 17:53 oshepherd shepazu: the OpenSocial API submission link is in the charter # 17:54 jasnell Sorry, but IMHO, the OpenSocial API is a great example of a bad example to follow. # 17:54 tibor_katelbach why jasnell ? # 17:55 jasnell unfortunately too much to go into right now tibor... # 17:55 wilkie aaronpk: micropub builds on top of oauth for authentication. mirror-image of h-entry. uses form requests to make, for instance, status updates. # 17:55 tibor_katelbach :) I'd love to here those thought maybe we can catch up later ? # 17:55 Zakim harry, listening for 19 seconds I heard sound from the following: aaronpk (89%) # 17:56 wilkie aaronpk: micropub is a way to make social posts which people are currently using. half of a dozen implementations for creating posts, thousands of users # 17:56 Zakim shepazu, listening for 10 seconds I heard sound from the following: Arnaud (34%), aaronpk (4%), Hhalpin (12%) # 17:56 tantek +1 to micropub as a building block for Social API # 17:56 jasnell as far as API is concerned, there are three key areas to address: Vocabulary, Actions and Federation. We're looking at Vocabulary and Actions already. There is overlap we need to look at with regards to various ontologies, including schema.org. # 17:56 tantek open spec and 10+ interoperable micropub implementations is a great start # 17:56 wilkie harry: while there is this opensocial spec submitted to w3c, is this the latest version? # 17:57 tantek jasnell, regarding overlap need to look is more like h-entry than schema.org (as far as API is concerned) # 17:57 oshepherd harry, or somebody else: Whats the procedure for bringing a proposal to the WG? # 17:57 tantek no one is using an API with schema vocab AFAIK. where as 10+ publishers / consumers are using micropub with h-entry and h-card # 17:58 jasnell for Federation, we're largely talking about discoverability and distribution... which is a fancy way of saying GET and POST :-) nicolagreco joined the channel
# 17:58 wilkie ??: The specification is indeed based on the latest version # 17:58 harry In the WG, you just ask the Working Group to adopt it as an Editors Draft # 17:59 Zakim sees bblfish, jasnell on the speaker queue # 17:59 harry and if the WG agrees, then we'll set you up a Respec.js and a space to edit in w3.org # 17:59 oshepherd harry: OK. I have a draft I made a few months back I'd like to clean up and propose. I guess I'd just mail in a link? # 17:59 harry Then if the WG moves forward with the spec, we can put in on what's called "Rec Track" # 17:59 tantek Arnaud, could you pick a chair for next week (Evanpro?) - I cannot make it - I will be on a flight. # 17:59 harry Yes, mail over mailing list, link to wiki, and we can discuss next call # 17:59 jasnell I'll post a detailed note about the Embedded Experiences and Action Handlers evolution by next call # 17:59 Loqi harry meant to say: Make sure int's in scope though! # 18:00 wilkie bblfish: we have a bunch of apis, and I want to know how far one can one go building off of that and how many people would be interested in that # 18:00 wilkie Arnaud: that is something that would be good to discuss on the mailing list # 18:00 harry As regards whoever was bringing up schema.org, until recently schema.org was incompatible with W3C's patent policy # 18:00 harry As regards Hydra, just point us to implementations and I'm sure folks can look at it. # 18:00 tibor_katelbach bblfish +1 # 18:01 harry There's been lots of Semantic Web "ontologies" in this space, all with very little uptake but we understand that can chage. # 18:01 wilkie Arnaud: we should gather our use-cases. people should look at the agenda and add anything that one thinks is missing. # 18:01 Loqi harry meant to say: There's been lots of Semantic Web "ontologies" in this space, all with very little uptake but we understand that can change. # 18:01 Zakim As of this point the attendees have been Arnaud, deiu, +1.303.204.aaaa, jasnell, aaronpk, MarkCrawford, Sandro, rhiaro, ShaneHudson, wilkie, MattMarum, bblfish, tommorris, # 18:02 Zakim ... elf-pavlik, Tsyesika, tibor_katelbach, tantek, Dan_Romascanu, oshepherd, Doug_Schepers, Hhalpin # 18:02 RRSAgent I'm staying, trackbot; no access has been specified for the meeting record # 18:02 Zakim Attendees were Arnaud, deiu, +1.303.204.aaaa, jasnell, aaronpk, MarkCrawford, Sandro, rhiaro, ShaneHudson, wilkie, MattMarum, bblfish, tommorris, elf-pavlik, Tsyesika, # 18:02 Zakim ... tibor_katelbach, tantek, Dan_Romascanu, oshepherd, Doug_Schepers, Hhalpin # 18:02 Zakim sorry, trackbot, I don't know what conference this is # 18:02 RRSAgent I'm staying, trackbot; no access has been specified for the meeting record # 18:03 harry hehe - sorry, paranoid due to a meeting missing hitting generate minutes # 18:03 Arnaud I think whoever is chairing should be responsible for starting the meeting and ending it # 18:03 harry So here is minutes in text for scribe to post to wkiki: # 18:04 Arnaud otherwise we don't know who's doing it and we may end up with no one doing it :) # 18:05 Arnaud wilkie, we should decide on where to put the minutes for each meeting # 18:05 Arnaud either at the end of the agenda where have a link for now or on a peer page like xxx-minutes # 18:06 rhiaro I don't suppose there's a way to join with Skype that doesn't require credit? # 18:07 elf-pavlik rhiaro, ekiga worked fine for me ... but i needed connect over VPN since Uni Leipzig blocks some ports :( # 18:08 Arnaud I wish I did though, I always pay skype credits when I'm abroad jasnell joined the channel
# 18:09 harry There's rumor W3C may be upgrading Zakim at some point, but timeline is unclear. I recommend Skype credit insofar as its the cheapest I can find. # 18:09 wilkie rhiaro: if skype phone-dialing is right now your best option, it's probably your BEST option. :/ # 18:10 rhiaro I was connected to the call today, just no audio # 18:10 Arnaud just no audio? not the best for a phone call :) # 18:11 wilkie I had to dial in today. I need to open ports on the other end of a VPN for next time. # 18:11 wilkie amazing that the better method of calling people involves routing through several machines :/ # 18:12 rhiaro I'll be back in Edinburgh in October though, where my I can just use a phone # 18:12 wilkie ok. that's technically 2 votes vs 1 for year-mm-dd-minutes # 18:13 Arnaud I guess "no audio" qualifies as "poor audio quality"? :) # 18:17 Arnaud tantek: for next week we can ask Evan if he wants to chair, otherwise I'm available jasnell joined the channel
# 18:20 wilkie wonders if it would be easier to have scribe.perl output wikitext barnabywalters joined the channel
# 18:20 aaronpk I meant an operational possiblity rather than a technical one # 18:20 wilkie right now it is a mess to just copy and paste it with all of the arbitrary indentation it uses # 18:21 aaronpk another possiblity is parsing the html to convert to wiki syntax # 18:21 harry if someone wants to make a perl script to convert w3c's text (made for email) to wikitext, that would be cool # 18:22 aaronpk harry: I looked at the plaintext output, it seems a little cumbersome to try to parse it, lots of structure is lost in the text version # 18:22 aaronpk what would be really cool is if scribe.perl were able to add a few class names to the HTML output that turns it into a microformats2 doc. then it'd be really easy to convert to wiki syntax # 18:23 wilkie barnabywalters: fixed time: 1300 EST every tuesday # 18:23 tantek barnabywalters: weekly telcon time has been set # 18:23 tantek aaronpk - I believe #sysreq is in charge of scribe.perl # 18:23 tantek might make some progress with making requests to improve scribe.perl there # 18:23 wilkie aaronpk: easier, yes, but at that point *shrug* just add wiki output heh # 18:23 aaronpk nah wiki syntax is completely different from HTML # 18:23 aaronpk adding mf2 class names is a minimal change to the existing code # 18:24 wilkie and then there is a script to convert that to wikitext? # 18:24 tantek wilkie - likely be easier to update the HTML than provide a completely different URL with different syntax # 18:24 aaronpk like my $prePhoneParagraphHTML = "<p class='phone'>"; would change to my $prePhoneParagraphHTML = "<p class='phone h-entry'>"; # 18:24 wilkie how do we have the arbitrary scribe run the script? # 18:25 aaronpk ideally this would be a small enough change that the main scribe.perl script could run the updated version to generate mf2 markup # 18:26 wilkie what do you use to generate the wikitext from the markup? # 18:26 aaronpk I would write a script that takes the mf2 JSON output and converts it to something that looks good in the wiki # 18:28 Arnaud "wilkie: right now it is a mess to just copy and paste it with all of the arbitrary indentation it uses" # 18:28 Arnaud tantek: it's hard for me to resist to saying "didn't I tell you so?" ;-) # 18:28 Arnaud wilkie: the good news ia accordding to tantek someone will fix it for you... # 18:29 Loqi Arnaud meant to say: wilkie: the good news is accordding to tantek someone will fix it for you... # 18:30 wilkie haha true, but I have some pesky pride getting in my way # 18:32 tibor_katelbach I had a few questions on actual use cases being developped on the Working group # 18:33 tibor_katelbach is there a common repository , how do we present our work in progress to get feedback ? # 18:34 tibor_katelbach we are a group of projects # 18:34 tibor_katelbach getting together building common goods , with interoperability and ldp in mind # 18:34 tibor_katelbach it's a citizen social network # 18:34 tantek I was not aware that LDP had become self-aware yet ;) # 18:35 tantek does this citizen social network have a specific name or URL? # 18:36 tibor_katelbach it's still in very early stage , we advanced on an citizen toolkit, that helps us build common good project # 18:36 tibor_katelbach right : pixelhumain.com # 18:36 tibor_katelbach but the work is still in QA state not published and all in french for the moment # 18:38 tantek tibor_katelbach: what do you foresee as intermediate steps between the current state of http://www.pixelhumain.com/ and when you wish to "present [y]our work in progress to get feedback" ? # 18:38 tibor_katelbach I build bricks for helping build forms based on ontologies # 18:38 tantek that sounds at least three degrees separated from the user! # 18:38 tibor_katelbach the project has many sides to it # 18:39 tibor_katelbach there a whole technical discussion part # 18:39 tibor_katelbach that's for us devs # 18:39 tantek tibor_katelbach: even devs should (MUST) start at solving user problems # 18:39 tantek I don't believe there is anything just "for us devs" # 18:40 tibor_katelbach the user part is comming , we hope end of the month , with mobile + web interface # 18:40 tibor_katelbach no we are fully user consient # 18:40 tantek tibor_katelbach: then that's a good time to think about presenting, *after* you have the "user part" integrated # 18:40 tibor_katelbach conscient # 18:41 tibor_katelbach it's the chicken and the egg, since I wanted to make sure the best practises are put into it # 18:41 tibor_katelbach before # 18:41 tibor_katelbach our data is LD'd harry joined the channel
# 18:41 tibor_katelbach and it's building quite nicely # 18:42 oshepherd tibor_katelbach: How does it differ from what AS2 provides, and how is it better? # 18:43 tibor_katelbach it's more about technical interactivity , we have 3 projects running in a similar direction , and willing to interoperate # 18:43 oshepherd "technical interactivity", those are words with no meaning # 18:43 tibor_katelbach I must admit I discovered AS2 last week, I jsut found you guys 3weeks ago # 18:43 tibor_katelbach and have some catching up to do # 18:44 tantek tibor_katelbach: best practices is whatever is minimumally necessary to provide the "user part" # 18:44 Loqi tantek meant to say: tibor_katelbach: best practices is whatever is minimally necessary to provide the "user part" # 18:45 tibor_katelbach sure but sometimes it's good to show stuff as you go to gather precious feedback # 18:45 tantek tibor_katelbach: to show stuff you need to have a "user part" # 18:45 tibor_katelbach very intereting # 18:46 tibor_katelbach I understand that's where we're heading anyways # 18:47 tantek if you're not heading for something with a "user part", then what's the point? # 18:47 tibor_katelbach osheperd:I could really say , I discovered AS last Week and socialWG the week before that , through openApps # 18:47 tantek tibor_katelbach: how did you discover indiewebcamp.com? # 18:47 tibor_katelbach we defintitly are # 18:47 tibor_katelbach it's full on user centric # 18:48 tantek then I'm a bit confsued by "the user part is comming , we hope end of the month" # 18:49 tantek since user-centric would imply the user part is at least there in some part at the beginning and always, not "end of the month" # 18:49 tibor_katelbach couldn't say , bouncing of all these social websites # 18:50 tibor_katelbach we have the API architecture set up as a citizen toolkit , and the 3rd application using it , 1st official one will be out in end of sept # 18:51 tibor_katelbach I have a lot of catching up to do that's why a quick tchat could help me keep the proper focus ,while profiting from the cllective knowledge # 18:52 tibor_katelbach I started a french version discussion of the technical interperability ,w ith henry story # 18:52 tantek tibor_katelbach: IMO there should be *no* API architecture without first figuring out "the user part" # 18:53 tibor_katelbach and we'll be starting the same ahngout with elf pavlik within 10days # 18:53 tibor_katelbach the user part is really clear # 18:53 tantek without the user part, it is likely any "API" or "architecture" will both fail to satisfy the user part, and be needless complicated # 18:53 tibor_katelbach the api is cut up into micro reusable modules # 18:54 tibor_katelbach login, create user, create group, send msg ...etc Shane joined the channel
# 18:54 tibor_katelbach I hear you, but it's working nicely # 18:54 tantek if the user part is really clear, could you point to a URL documenting this "clear user part"? # 18:54 tantek tibor_katelbach: *what* in particular do you consider "working"? # 18:54 tantek e.g. are you posting things on your own website using this API/architecture? # 18:55 oshepherd So my protocol proposal/the Pump IO API does all of what you said (except login, which I've diked out, because there are open issues there) in ~400 lines of ReSpec HTML definition :-) # 18:55 tibor_katelbach we have a scenario based system # 18:55 tantek tibor_katelbach: please provide a URL documenting your scenario based system # 18:56 tibor_katelbach this was the very first version # 18:56 tibor_katelbach it's not ld'ed # 18:56 tantek oshepherd: +1 the Pump IO API is quite powerful, have seen it demonstrated by evanpro in person running an actual site / UI> # 18:57 tibor_katelbach it's used on a mobile project for people observation of marine realted issues # 18:57 oshepherd tantek: Yeah, its' basically "You have an inbox and an outbox, and all you need to know" # 18:58 oshepherd And, aside from some corner cases, pretty much everyone who interacts with the client-to-server part of it at least goes "Wow, this is simple, clean and elegant" # 18:58 tibor_katelbach I'll lok into pump io , it this you ,building it ? caseorganic joined the channel
# 18:59 tibor_katelbach we've gonea long way from that first application , I could show it as a live session # 19:00 tibor_katelbach or I'd need to push it online , I made a network map^ping tool # 19:00 tibor_katelbach or participative ticketing system # 19:00 tibor_katelbach loads of use cases , with user interfaces # 19:01 tibor_katelbach this is our discussion session with henry story and the other projects # 19:01 tibor_katelbach it's a monthly thing # 19:02 tibor_katelbach we'll be doing the same in english soon # 19:02 tibor_katelbach anyone is welcome to join in # 19:04 tibor_katelbach pump io looks really interesting thx bblfish joined the channel
# 19:15 tibor_katelbach this is a 4 month old version of the api scenario # 19:15 tibor_katelbach it's a lot cleaner now # 19:16 tibor_katelbach we build a scenario for any citizen based application # 19:17 tibor_katelbach then we solve teh scenario with brick from the api toolkit melvster1 joined the channel
# 19:17 tibor_katelbach if a part of the scenario is missing # 19:17 tibor_katelbach we add it and keep it generic # 19:17 tibor_katelbach to be able to reuse # 19:18 tibor_katelbach this builds the api scenario of an application , we then simply plug it to any front end melvster1 joined the channel
# 19:20 oshepherd You see, I'd say what pump.io does was bult the other way around: Define something simple which encompasses all needs from the start # 19:21 tibor_katelbach I'll definitly look into it # 19:21 tibor_katelbach the toolkit modules are growing with each project using it # 19:21 tibor_katelbach and since 2 projects we've started linkingdata # 19:22 tibor_katelbach it's opening great new perspectives # 19:22 tibor_katelbach I have to look into AS as well looks quite rich as well # 19:23 tibor_katelbach I can plug to this tool for building Ontology based forms # 19:23 tibor_katelbach forhe moment only based on schema.org ontologies but that's jsut a switch jasnell joined the channel
# 19:24 tibor_katelbach sorry for the bad sound # 19:24 tibor_katelbach it's really alpha state # 19:25 tibor_katelbach but you'll see the idea # 19:26 tibor_katelbach yes , since a long time in fact # 19:26 tibor_katelbach i didn't quite like the way it is integrated into html though # 19:26 barnabywalters and that the use of “microformat” on that page has nothing to do with the existing standards and community? # 19:27 tibor_katelbach I found microformats interesting for the generic structures representing things # 19:28 tibor_katelbach you're right I've been using the word for a very long time , the became a trend # 19:28 barnabywalters it seems a little disrespectful and unwise to knowingly use an existing term completely out of context # 19:29 tibor_katelbach not disrespectfull , I was using it long back # 19:30 tibor_katelbach project open atlas # 19:30 tibor_katelbach yep , i'd have to dig for it # 19:30 tibor_katelbach it was a real coincidence # 19:31 tibor_katelbach then I saw the mf come out it correspond quite nicely # 19:31 tibor_katelbach open atlas became an NGO since then though # 19:32 tibor_katelbach that's quite a while back # 19:32 tibor_katelbach no there's no more url to it # 19:33 tibor_katelbach it died when Google Earth came out # 19:33 tibor_katelbach and Google maps # 19:33 tibor_katelbach back then I was builing that , and GMaps came out in 2001 or so # 19:34 tibor_katelbach more or less # 19:34 tibor_katelbach you're making my hair go grey # 19:35 tibor_katelbach if you look into Gmaps chat there's surely references in there # 19:35 tibor_katelbach I switched to there api for 2 years or so and lost the mojo # 19:36 tibor_katelbach the project turned into the humanpixel # 19:36 tibor_katelbach it's evolved into an empowermetn movemetn # 19:37 tibor_katelbach with a strong technical approach to it # 19:37 tibor_katelbach but remains user centric # 20:08 Zakim excuses himself; his presence no longer seems to be needed jasnell, jasnell_, deiu and bblfish joined the channel
# 21:08 tantek oshepherd: "While Salmon works for this purpose, it is itself highly complex." yeah. # 21:08 oshepherd tantek: Those bits were written back when the charter was first being drafted up and I saw suggestions that it would be based upon PUSH and Salmon and went "Ye god no!" harry joined the channel
# 21:10 oshepherd I think we are going to have to deal with nominally similar complexity in the auth system (i.e. some kind of public key signing), but at least we can avoid the mess that is XML signatures and such # 21:11 tantek oshepherd: I'm not so sure. We're going to have to deal with the usability of all this stuff for sure. And implementability. The best laid spec plans are no match for things that are actually implementable by more people (and thus implemented by more people). # 21:12 oshepherd tantek: Yeah, but here I'm thinking of "Client A needs to authenticate itself as user@server1.com to server2.net" :-) # 21:12 tantek Salmon (and others) fell into the uncanny implementation valley of "we got a few implementations, but no more" # 21:13 tantek oshepherd: we already have this working: "Client A needs to authenticate itself as user1.example.com to user2.example.com" with IndieAuth # 21:13 tantek much more easily than any of the Salmon stuff. Also no Webfinger needed. # 21:13 oshepherd tantek: Without the user having to visit user1.example.com? hhalpin joined the channel
# 21:15 oshepherd (A usecase I'm thinking of is "A friend has shared post server2.com/~alice/note/983 with me and now my mobile client want to fetch its' comments" # 21:18 oshepherd tantek: I don't see any protocol there (besides the one which /trusts/ IndieAuth.com, which is an unacceptable centralization) # 21:18 tantek right, it's a default implementation, all open source # 21:19 oshepherd It seems to depend upon the silos for identity. Or is there another protocol there? # 21:20 tantek it has the *option* of using silos (user-choice), for *authentication* # 21:21 tantek a heck of a lot simpler than *either* Salmon or Webfinger # 21:22 oshepherd Basically, whats needed is some way for a client to prove to a server other than the user's home server that (A) it is acting on behalf of the user, and (B) that it has permission to make the request it's doing # 21:22 wilkie I want to prove my identity (strong integrity) (and authenticate, I guess) with ONLY my own server. not clear how that works with this indieauth page tbh. # 21:24 tantek wilkie - well we can start with setting up your site to support simple IndieAuth and go from there - it helps to do it yourself to learn how it works. jasnell joined the channel
# 21:24 wilkie I'm not sure why I would do that outside of mere education heh # 21:25 tantek wilkie: the simple use-case is to be able to use your domain to sign-into the indiewebcamp.com wiki and edit things there # 21:25 wilkie right. what do I *get* by using it... not really sure # 21:25 tantek if you want to learn how it works, how the flow works # 21:25 wilkie I realize I'll need to use *your* authentication scheme to log into *your* suite of websites # 21:26 wilkie right, outside of educating myself about how you do things, not sure why I would want this as a member of the social web # 21:26 tantek wilkie: all you have to add is rel=me to the existing twitter link on your home page # 21:26 wilkie so people can subscribe to that feed in that way # 21:27 oshepherd tantek: I don't think any protocol is going to fly here if part of it is "just link to twitter" :-) # 21:27 tantek wilkie: rel=me is simply a markup declaration of what you just said # 21:27 wilkie it is indeed broken. I have to create a new SSL cert # 21:27 tantek indieauth is only *one* use-case that takes advantage of that declarative markup # 21:27 wilkie nothing you are saying is really answering my question # 21:28 tantek you could also add rel=me to your github link # 21:28 tantek because you linked to github on your home page # 21:29 tantek it makes a difference because you as a user have already asserted that that github profile is part of your identity # 21:29 wilkie tantek: the downfall of embedding links in html and always expecting them to work :) I lost that domain due to being poor # 21:29 tantek and linking to it in the sentence "I write code." # 21:29 wilkie why does it matter? does server B talk to github? did I suggest that I trust github as an avatar? # 21:30 tantek you have the *option* to choose that when you do IndieAuth yes # 21:30 tantek it gives you the user the option to choose from your rel=me links # 21:30 oshepherd tantek: I still don't see how any of this solves the problem I posted above :-) # 21:31 wilkie uh. I choose none of them. they should talk to my own server. # 21:31 wilkie and yeah, it doesn't solve oshepherd's original problem, which is a real and important one # 21:31 tantek wilkie - you can do that too - then you just need to setup your domain as an endpoint in itself # 21:31 tantek I was merely giving you the fastest path to understanding how it works # 21:32 wilkie ah, sounds good. it really doesn't come across on that page at all. # 21:32 tantek you're not going to find a simpler identity/auth solution than adding 6 characters to your existing home page ;) # 21:32 wilkie well I have to do more than that to avoid your server right? # 21:34 oshepherd tantek: (Also, if you checked the draft spec of mine, you'd see that WebFinger is used only to lookup acct: URIs, which are almost certainly required for backwards compatibility) # 21:34 tantek in terms of the auth implementation, like many open source solutions, you can either deploy it yourself on your own server, or you can use an existing deployed proxy # 21:34 tantek oshepherd: there isn't enough webfinger around to worry about backcompat # 21:35 oshepherd tantek: Sure there is. How are StatusNet, Pump.io, etc supposed to migrate if we remove that? # 21:35 tantek I should say, totally doomed to rot legacy, as it did with Google's deployment. # 21:36 tantek oshepherd: they have URLs of profiles, they don't need webfinger # 21:36 tantek Status.net users subscribed to my site tantek.com via PuSH updates no problem, no webfinger needed # 21:36 oshepherd tantek: Sure. But all of their existing relations are specified in terms of user@domain format # 21:37 tantek then its an internal matter for them to map those internal relations to be based on their profile URLs # 21:37 oshepherd tantek: What about when StatusNet users have connected with Friendica users or rstatus users? # 21:38 tantek and in their lists of "who am I following" they see lists of names linked to … profile URLs # 21:38 oshepherd What about the fact that most users seem to prefer user@domain as a user ID format? barnabywalters joined the channel
# 21:38 wilkie rstat.us can use a profile url in place of a webfinger and internally the webfinger is just a username and a domain. webfinger isn't really baked in. # 21:39 oshepherd webfinger is a protocol distinct from the acct: URI scheme, fwiw # 21:39 tantek oshepherd: we choose to disagree on what "most users seem to prefer", unless you say Facebook, in which case I agree. # 21:40 oshepherd StatusNet/GNU Social. Diaspora. Pump.io. rStatus. Jabber. SIP. E-Mail. There is a clear trend of "users are specified in user@domain format" # 21:41 tantek oshepherd: that's a pretty funny "trend" - more like dying legacy. If you want to talk user's specified in format of any, it's just "@-username". E.g. Twitter, Instagram etc. # 21:42 oshepherd Sure, but that doesn't exactly work when your username scope is the whole internet # 21:42 wilkie ugh the debate about whether or not you are a domain or you are user@domain # 21:42 oshepherd Per the draft spec. user IDs are URLs. Hell, it even defines some rules so you can omit all the cruft most of the time ("tantek.com" would be normalized to "https://tantek.com/") # 21:43 tantek oshepherd: sounds like good spec writing then! # 21:43 oshepherd somebody@somedomain.com would be normalized into acct:somebody@somedomain.com, etc # 21:44 wilkie what's the url-is-identity group's method of disambiguating usernames within, say, posts/statuses? # 21:44 oshepherd wilkie: Per my draft spec, it dereferences the ID (post WebFinger, if its u@d format), then looks at the AS2 objectType # 21:45 oshepherd But: you only need to implement a webfinger server if you want to use user@domain IDs; you only need to implement the client if you want to follow user@domain IDs (and you could work around that by manually working out where they point and your site would probably still work) # 21:45 wilkie you are typing out a status and you type in @Jeff and if there is only one @Jeff *you* know it knows to fill it in to whatever and can be smart about it # 21:46 wilkie and how the actual form of identity looks like doesn't seem to matter at all # 21:46 Loqi wilkie meant to say: and what the actual form of identity looks like doesn't seem to matter at all # 21:46 oshepherd wilkie: I think it should probably behave like Facebook does, i.e. @ really just brings up a combo box which lets you filter for a user # 21:47 tantek wilkie - there have been several different approaches to the UI - it's an area worthy of innovation! # 21:47 wilkie yes, a drop-down is what we already do. not sure how accessible it is, but it solves that problem well enough. what the user sees, who cares what it looks like... internally, just use a full url like we pretty much already do anyway # 21:48 tantek yes, FB drop-down UI for this does seem fairly good. barnabywalters joined the channel
# 21:49 Loqi wilkie meant to say: twitter does it too, but it is awful # 21:50 KevinMarks well, twitter does fuzzy matching as well- if you type the handle it matches, but it will auto-complete based on name too # 21:51 KevinMarks G+ fails at not showing the underlying ID (in their case email) which means I can accidentally cc one of my son's abandoned gmails instead fo his current one # 21:51 tantek interesting, what is this common UI widget called? # 21:52 tantek (I just made that up, hoping someone else has an actual name used by others) # 21:52 oshepherd Its' just an application of the autocomplete UI web brwosers use to other things # 21:52 tantek because good versions like FB use your network to do so, filter, sort etc. # 21:52 tantek FB is clearly using all kinds of other smarts # 21:53 tantek far beyond anything any browsers are doing with looking up a simple list # 21:53 oshepherd If I type "min" then I get the article "Prime Minister of the United Kingdom" on Wikipedia as a recommendation (random example) # 21:56 oshepherd I admit that Facebook has better results, but it also has a better ranking of things I might be referencing # 21:57 oshepherd (Or: My Facebook likes and friends lists, plus my friends' Facebook like and friends lists, are a pretty good estimate of "things I might be interested in talking about") # 21:57 tantek but specifically of *names* for which FB (and others) can and do do much smarter things # 21:57 oshepherd Sure, but its' a matter of the quality of the underlying data, not really any especially algorithms # 21:58 tantek oshepherd: I'd give FB the benefit of the doubt with having decent algorithms behind the "nice" drop down they have # 21:59 oshepherd Sure. Maybe a bit of recent history, prefer stuff I've interacted with directly, etc # 22:00 tantek Hmm - I'm going to go with autosuggest instead # 22:02 tantek and yes I grant the dropping of "name-" specifically # 22:02 tantek hopefully the context of discussion is sufficient to disambiguate # 22:13 tantek KevinMarks: I can add h-event to our WG wiki page bblfish joined the channel
# 22:15 tantek wilkie, thanks for the use-case / user-scenario of a user typing in a person's name and expecting to see it autocompleted - apparently there has been a bunch of discussion / work / analysis on this in the indiewebcamp community http://indiewebcamp.com/autosuggest bblfish joined the channel
# 22:16 tantek KevinMarks: no, h2vx blocks Google's web crawler # 22:16 tantek Google Calendar is too dumb to have its own UA # 22:16 tantek bad robot, pretending to be a different robot # 22:17 tantek since you're a user of Google Calendar, how do you report issues against in a place that is publicly trackable? grantmacken, jasnell, bblfish and nicolagreco joined the channel
elf-pavlik telecon in < 30min :)
RRSAgent logging to http://www.w3.org/2014/09/02-social-irc
ShaneHudson Zakim, what is the code?
hhalpin hey everyone, I am in border control at an airport due to flight delay so dialing in late
tantek bblfish this one: http://socialwg.indiewebcamp.com/irc/social/2014-09-02/line/1409678711975 :)
MattMarum Latest currently released OpenSocial specification: http://opensocial.org/documentation/opensocial-2-5-1/
tommorris Since last week, Linked Data Platform now has a Wikipedia article: https://en.wikipedia.org/wiki/Linked_Data_Platform
bblfish wilkie: yes, and especially how it ties into with LDP https://dvcs.w3.org/hg/ldpwg/raw-file/tip/ldp-primer/ldp-primer.html
Arnaud I was thinking about https://www.w3.org/wiki/Socialwg/2014-09-02-minutes but either way is fine with me
barnabywalters oh did I miss the telcon?
wilkie barnabywalters: afraid so. minutes are at: http://www.w3.org/2014/09/02-social-minutes.html atm
aaronpk I found the source code http://dev.w3.org/2002/scribe/scribe.perl
oshepherd tantek, very early draft: https://dl.dropboxusercontent.com/u/64151759/ActivityPump.html
KevinMarks G+ too