#social 2015-05-01

2015-05-01 UTC
bblfish, almereyda, elf-pavlik, the_frey and jaywink joined the channel
#
rhiaro
elf-pavlik: sunday 1900
#
rhiaro
^ in paris
#
rhiaro
but I think I should go straight to my CS host
#
elf-pavlik
rhiaro, i think so... we'll meet all together on Monday morning! :)
#
elf-pavlik
jaywink, i remember something about diaspora using hCard, do you know more details about it? https://github.com/w3c-social/social-vocab/issues/11
#
jaywink
it is the future diaspora federation ruby gem that has been extracted out of diaspora itself - so we can easily "switch" federation protocols :)
#
elf-pavlik
jaywink++
#
Loqi
jaywink has 1 karma
mib_u4n64y, bblfish, stevenroose and almereyda joined the channel
the_frey joined the channel
#
elf-pavlik
!tell tantek I hope you and Evan can try to work together on getting your systems to interoperate :) https://github.com/w3c-social/social-web/issues/1
#
Loqi
Ok, I'll tell him that when I see him next
almereyda, tilgovi and KevinMarks joined the channel
#
cwebber2
elf-pavlik: ++
#
bblfish
one should also look at the contact ontology http://www.w3.org/2000/10/swap/pim/contact#
#
bblfish
timbl uses that
#
elf-pavlik
rhiaro, what do you think about my comment on Microformats 'not using tripples'? https://www.w3.org/wiki/Socialwg/Social_syntax/Comparison#Entity_Attribute_Value
#
elf-pavlik
i may possibly miss something in my understanding
#
elf-pavlik
ben_thatmustbeme, aaronpk, kylewm ^
#
elf-pavlik
s/tripples/triples/
KevinMarks and tilgovi joined the channel
bblfish joined the channel
#
melvster
elf-pavlik: ^^
#
elf-pavlik
melvster thx! BTW bblfish uses React ...
#
melvster
elf-pavlik: are you staying with him?
#
elf-pavlik
ATM I play with converting microformats JSON of this channel to microformats JSON-LD
#
elf-pavlik
"rels": {
#
elf-pavlik
"pingback": ["http://webmention.io/indiewebcamp/xmlrpc"],
#
elf-pavlik
"webmention": ["http://webmention.io/indiewebcamp/webmention"],
#
elf-pavlik
"prev": ["http://socialwg.indiewebcamp.com/irc/social/2015-04-27", "http://socialwg.indiewebcamp.com/irc/social/2015-04-27"],
#
elf-pavlik
"next": ["http://socialwg.indiewebcamp.com/irc/social/2015-04-29", "http://socialwg.indiewebcamp.com/irc/social/2015-04-29"]
#
melvster
elf-pavlik: maybe before monday he can help you get a webid, then I can show you how to log in to some apps im making, if you are interested ...
#
elf-pavlik
this already looks closer to JSON-LD than AS links thing "links": [{"rel": "next"}]
#
elf-pavlik
sure, i can add turtle serialization to my deployment via https://github.com/zazukoians/rdf-ext/issues/6
#
elf-pavlik
I hope you can handle my HTTP 303 redirects :P
#
melvster
that would be great
#
elf-pavlik
but since LDP mandates JSON-LD WebID could also support it
#
elf-pavlik
specs from SocialWG can require JSON-LD but very unlikely will do anything with Turtle
#
elf-pavlik
still, for monday I'll support "Accept: text/turtle"
#
melvster
sounds good
#
melvster
if people want me to, sure
#
aaronpk
the demos should be focused around user stories
#
elf-pavlik
melvster, just so you expect it :) ^
#
aaronpk
that was the whole point of making the user stories, so we have a point of focus for our discussions
#
melvster
im only an observer so i may not demo anything at all, but ill be able to interact with the group and help if people see relevance, if there's a particular user story people would want solved, I could probably code it up while im there
#
elf-pavlik
i guess i don't need to explain reasoning for that...
#
aaronpk
elf-pavlik: it's described on the parsing spec page http://microformats.org/wiki/microformats2-parsing
#
elf-pavlik
BTW, in debate as:Link with as:rel or not, JSON-LD take similar approach as this "rels" object! while as:rel makes things more like JRD (webfinger profile)
#
elf-pavlik
aaronpk, in case that helps you relate to this discussion better ^
#
elf-pavlik
{ "pingback": "http://ex.foo", "webmention": "http://ex.bar" } vs. [{ "rel": "pingback", "href": "http://ex.foo" }, {"rel": "webmention", "href": "http://ex.bar"}]]
LCyrin joined the channel
#
elf-pavlik
just as "rels" in microformat JSON stays compatibile with HTML
#
elf-pavlik
!tell tantek last telecon you pretty much made an argument that Webfinger JRD stays compatibile with HTML while Microformats JSON "rels" NOT! http://socialwg.indiewebcamp.com/irc/social/2015-04-28/line/1430244146625
#
Loqi
Ok, I'll tell him that when I see him next
#
rhiaro
elf-pavlik: I agree with what you wrote about microformats matching a triple structure. Where does microformats claim not to be triples? I don't recall I've seen it mentioned anywhere either way. There's also a strong inclination in the community to make sure entities being described have URIs rather than nesting / blank nodes (sometimes URIs are implied to be the document the microformats are in, though sometimes it's explicit with a u-uid)
#
rhiaro
so on that front microformats aren't doing anything particularly different to other vocabularies
#
rhiaro
well, entity-attribute-value sure feels tripley to me :)
#
aaronpk
what is the entity in that case?
#
elf-pavlik
which case?
#
rhiaro
aaronpk: for a blog post, the h-entry would be the entity
#
rhiaro
which typically have their own URLs
#
aaronpk
oh okay
#
aaronpk
why is thinking of that as a triple useful?
#
rhiaro
in a feed they're in u-url or u-uid. If it's a h-entry page, it's implied by the URL of the page
#
rhiaro
it helps some people, doens't help others. The point is, there's no conflict
#
elf-pavlik
subject or entity
#
elf-pavlik
and both of them have paris of attribute: value or predicate: object
#
rhiaro
that looks straightfoward to me, if you assume u-urls to be identifiers for two different entities, that are linked by the p-org relation
#
elf-pavlik
10:01 tantek nah - RDF complicates the model unnecessarily with basing it on "triples" http://microformats.org/wiki/triples
KevinMarks joined the channel
#
elf-pavlik
09:59 csarven re: "all microformats are simply an object with a set of properties with values." from http://microformats.org/wiki/microformats2#Background . That's pretty much EAV model. Which is used by RDF as well.
#
rhiaro
I remember that conversation
#
aaronpk
well, regardless of whether you want to think of mf2 markup as triples, plenty of people do not, and get away with publishing and consuming it just fine, so i'm not sure I see the advantge
#
aaronpk
elf-pavlik: see above
#
rhiaro
I think this is a totally unnecessary thing to argue about
#
elf-pavlik
to my understanding each item sets new entity / subject
#
aaronpk
elf-pavlik: you still didn't clarify why thinking of it as triples is useful
#
rhiaro
aaronpk: In my mind querying by graph pattern matching would be easier than plucking things out of nested json objects, but I don't know anything about parsing json really, so it probably just depends where your area of expertise is
#
rhiaro
I do think the /triples page on microformats wiki is misleading; microformats markup doesn't need to be thought of as triples, but *could* be if it helps to query it, but otherwise it doesn't make much difference either way. Insisting the models are totally incompatible is unfair.
#
rhiaro
my solution to the 'arguement' would be for everyone to tone down their strong opinions a bit, accept there are different ways of thinking about things that don't conflict, and leave it there ;)
#
elf-pavlik
rhiaro, while I agree to big extent, to actually discuss technical topics sometimes we may need to make some effort to understand each other, otherwise we'll keep landing in arguing about terminology without even getting near the meaning we try to express with words we use
#
aaronpk
i don't have strong opinions about it at all
#
rhiaro
aaronpk: I didn't mean you
#
rhiaro
Can we find a way to pitch how beautiful and harmonious it is that rdf and microformats actually fit together so well? :)
#
rhiaro
Can we also get a big button someone can push whenever two people start arguing about terminology that indicates "you are actually saying the same thing, stop arguing"
#
elf-pavlik
rhiaro++
#
Loqi
rhiaro has 64 karma
#
rhiaro
doesn't have much energy
#
aaronpk
i like that button
bblfish joined the channel
#
rhiaro
I think I'm pre-exhausted in preparation for the f2f
#
elf-pavlik
what do you work on?
#
rhiaro
ooh we could make a giant poster that is some json-ld beside some microformats json to show how similar they look!
#
elf-pavlik
i can convert IRC logs of this channel from microformat JSON to microformats JSON-LD tomorrow
#
rhiaro
elf-pavlik: I'm setting up my chromebook so it's useful so I don't have to bring my thinkpad, tidying my desk, sending CS requests for Florence, and trying not to think about all the phd work I haven't done
#
rhiaro
elf-pavlik: great!
#
rhiaro
elf-pavlik++
#
Loqi
elf-pavlik has 18 karma
#
elf-pavlik
we still may need to deal with tensions after Shane vs. Manu incident :( http://manu.sporny.org/2013/json-ld-is-the-bees-knees/
#
rhiaro
is distracted laughing at that gif
#
rhiaro
so I think 'incident' is a bit strong, and also mention of 'different camps' in emails is unnecessarily divisive and probably doesn't help
#
rhiaro
if people start seriously arguing that microformats and rdf are fundamentally incompatible, I'm probably just going to leave
#
aaronpk
manu's snarky animated gif responses to shane's technical arguments makes me want to not have anything to do with the json-ld community :(
#
rhiaro
eh, I reckon judging a whole community on one blog post, or even one person, is a bit much
#
aaronpk
i agree, it's just certainly not helping
#
rhiaro
I really need to write my followup activities/objects blog post
#
bblfish
instead of thinking about formats why not just draw graphs of arrows. Then we can all agree and its visually more pleasing
#
rhiaro
bblfish: personally, I love graphs of arrows. But I think people who are already objecting to thinking about things as rdf might not see it that way
#
aaronpk
yeah i did not follow the arrows at all, no pun intended
#
rhiaro
I think we all need to make more effort to see things from the perspectives of people we think we disagree with
#
rhiaro
rather than repeating our own preferred models over and over
#
bblfish
yes, though its a bit difficult because you can always reduce every data structure to graphs of arrow
#
elf-pavlik
+1 and we *all* need to take some effort, especially expect more from oneself than from others
#
rhiaro
what would actually be fab is if a strong linked data advocate took a day to implement something with microformats, and a .. not linked data person... took some time to put together some data as a graph, and then both reported on their experiences
#
rhiaro
or maybe that wouldn't help at all, who knows
#
elf-pavlik
i think it would
#
rhiaro
another outcome would be "microformats isn't data-y enough, you can't do anything with this" and "rdf is complicated it's no use for anything" :p
#
bblfish
given that there is always a mapping from everything to a graph of arrows, there will of course be a mapping from microformats or microdata to that same graph. If that mapping can be automated, as it can with rdfa, turtle, json-ld , then reading those formats is not a problem
#
elf-pavlik
rhiaro, we also have a lot of similarities in API endpoints, but again talke about them in different terms
#
bblfish
But I am not really keen to enter the syntax wars - there are too many interesting things to do
#
rhiaro
my persepctive, as a user of both, is microformats are super easy to get started with and the tools for parsing them are ace, so they're useful, but I still get confused about what to nest where and whether I've got enough semantics to be useful; thinking about things as a graph comes most naturally to me, so modelling and storing and querying my data as rdf is awesome, but setting up a triplestore is a pain in the arse for most people
#
rhiaro
elf-pavlik: yes, also that with API endpoints
#
rhiaro
indieweb approach is totally FYN :)
#
bblfish
yes, the tools for rdf are much better but they can be improved a lot. In Scala there are some really cool things that are possible by way of improving the working with RDF. I mention that Scala now also compiles to Scala-JS . ( I show some of the features in my december london talk at Scala eXchange )
#
bblfish
But that is just the beginning.
#
elf-pavlik
bblfish, we need diversity with programing languages, if just one language has some cool tools we stay deep in a woods
#
rhiaro
also things that work on really basic servers, to enable people with little time or expertise to get social things up and running
#
elf-pavlik
aaronpk, where do i find more explanations about "value" as seein in JSON example of http://microformats.org/wiki/microformats2#h-event_location_h-card
#
elf-pavlik
KevinMarks, I guess you might also have link at hand :) ^
#
rhiaro
oh elf-pavlik I forgot to tell you, I published my notes on microformats as rdf: http://rhiaro.co.uk/2015/03/microformats2-rdf
#
elf-pavlik
rhiaro++
#
aaronpk
elf-pavlik: it should be on the parsing page, but not sure at the moment
#
Loqi
rhiaro has 65 karma
#
rhiaro
I had a couple of questions about properties I didn't understand or weren't well described
#
rhiaro
I suppose I should drop it in #microformats :)
stevenroose joined the channel
#
aaronpk
seems like a reasonable place to ask :)
#
rhiaro
elf-pavlik: 'article' is a bit strong ;)
#
elf-pavlik
has title so not a note (or i didn't remember this entilment corectly? ;)
#
rhiaro
yeah that's right. But semantically article says to me more coherant....
#
elf-pavlik
p-education - if it can have h-card which can have url woldn't it make it u-education?
#
aaronpk
the p- and u- and other single-letter prefixes are for parsing rules, and don't affect the vocabulary
#
elf-pavlik
so p-properties can all have nested entity with url?
#
aaronpk
the parsing syntax allows that
#
aaronpk
the vocab may not
#
bblfish
elf-pavlik: there is diversity in programming languages.
#
rhiaro
vocab seems pretty unrestrictive. I've been going with 'most things are fine' when marking up new things
#
aaronpk
e.g. "in-reply-to" may be a URL if the HTML class was "u-in-reply-to", but may be an h-cite object if the class is "p-in-reply-to h-cite"
#
rhiaro
(which may be reflective of my attitude more than the documentation, but.. *shrug*)
#
rhiaro
oh right yeah u and p prefixes get dropped after parsing right?
#
aaronpk
that's kind of the great part of mf2, is that the parser doesn't have to change but we can keep iterating on vocabularies separately
#
aaronpk
yes, the prefix doesn't appear in the parsed result
#
elf-pavlik
does some properties have p- and u- variants?
#
aaronpk
sure, the in-reply-to example above
#
rhiaro
If we get into heavy debates at the f2f can we have a couple of hours when everyone may only speak in haiku, to lighten the mood? :)
#
elf-pavlik
aaronpk, does only p-* and e-* allow nested microformats as values? (sorry for bit random question, just trying to catch up a bit on how it works)
#
elf-pavlik
rhiaro, let's take a closer look together next week on h-listing and h-product ?
#
aaronpk
elf-pavlik: excuse me for the seemingly impersonal reply, but this is a good writeup of how it works http://microformats.org/2014/03/05/getting-started-with-microformats2
#
aaronpk
jump to "Prefied Classnames"
#
aaronpk
"Prefixed Classnames"
#
aaronpk
the prefix tells the parser where to find the value
#
rhiaro
elf-pavlik: yeah, I expect to use them for requests/offers of items
#
elf-pavlik
"The reason comments are h-cites instead of h-entrys is that h-entry implies syndication — it’s something you’ve posted, or have re-posted, whereas a comment is a reference to a post on another site."
#
rhiaro
oh, I totally didn't know replies were supposed to be h-cites
#
elf-pavlik
rhiaro, does entity with certain url has 'type' h-entry or h-cite depending on where it appears?
#
rhiaro
h-entry or h-cite should be explicit in the markup
#
rhiaro
as far as I know
#
rhiaro
probably better off asking these questions in #microformats, I don't really know
#
rhiaro
I *have* read most of the wiki, but it hasn't all fused with my brain yet
#
aaronpk
also #indiewebcamp is a good place to talk about the usage of microformats vocabularies on personal websites
#
aaronpk
i don't understand the question
#
aaronpk
oh, the example on the wiki is citing tantek's post
#
elf-pavlik
"type": ["h-cite"],
#
elf-pavlik
"properties": {
#
elf-pavlik
"url": ["http://tantek.com/2013/104/t2/urls-readable-speakable-listenable-retypable"],
#
elf-pavlik
while on tantek site it parses to
#
elf-pavlik
"type": ["h-entry", "h-as-note"],
#
elf-pavlik
to me it looks like http://tantek.com/2013/104/t2/urls-readable-speakable-listenable-retypable has different type depending on context it appears in
#
elf-pavlik
rhiaro, quads? :D
#
rhiaro
elf-pavlik eeeeeeeeeep let's not. :p
#
rhiaro
one step at a time :)
#
aaronpk
h-cite is one of the lesser explored vocabularies right now, so don't read too much into it right now
#
rhiaro
but also, indieweb doesn't belive in explicit types ;)
#
aaronpk
start with the more popular ones like h-entry and the more useful in-reply-to like-of repost-of etc
#
elf-pavlik
rhiaro, ;P
#
rhiaro
elf-pavlik: that's actually really interesting
#
rhiaro
but I'm gonna have to try really hard not to think about it until I've dealth with a number of other things
#
rhiaro
(*really* interesting)
#
elf-pavlik
if i make request to u-in-reply-to i will get microformat with type h-entry not h-cite as when i nest it?
#
aaronpk
okay here is a relatively sane example of in-reply-to as an h-cite
#
aaronpk
also the comments are h-cites
#
elf-pavlik
there it has h-entry
#
aaronpk
yes his post on his site is an h-entry because it's on his site
#
elf-pavlik
when I think in triples and quads it makes perfect sense to me...
#
aaronpk
but when I show his post I give it h-cite because it's not my post
#
elf-pavlik
going from triple into quad brings context into the game: 'my site' vs 'his site'
#
elf-pavlik
yes, i would say that it has different meaning in different context
#
elf-pavlik
aaronpk, if you happen to reply in one post on your site to another post on your own site, you would also use h-cite in the reply?
#
elf-pavlik
last question for today, can you think of common cases where people give multiple types to entities?
#
melvster
aaronpk: having read through the use cases I think I can implement almost all using SoLiD
#
aaronpk
melvster: that's good, i don't doubt that. just make sure when you demo that you mention which user story you are demoing
#
elf-pavlik
melvster, all 90 or all from some of the top sections?
#
melvster
aaronpk: I will try, but im not a member of this group, I dont have access to the mail list discussions, or input on wiki I think, if I demo anything at all, it will be driven by what people ask me to do
#
melvster
elf-pavlik: is there 90? I thought I could have a go at all but 1
#
melvster
so 89/90
#
aaronpk
i think the mailing list archives are public anyway https://lists.w3.org/Archives/Public/public-socialweb/
#
melvster
aaronpk: I applied for access about a month ago, the application has yet to be processed, so ill just be an observer, any demonstrations would need to be driven by WG members
#
melvster
but given tantek's comments on the mailing list, it may be more productive to work over IRC
#
elf-pavlik
melvster, if you have demo of on of the top user stories from https://www.w3.org/wiki/Socialwg/Social_API/Sorting_user_stories using SoLiD almost certainly you can prestent it
#
elf-pavlik
please just don't give us "i could implement 89/90, but will do none since i have only invited observer status during this meeting" :D
#
melvster
elf-pavlik: let's talk about it on monday, in that case i dont plan to demo anything at all, ill just observe
#
melvster
my opinion is that SoLiD is capable of handling the use cases
#
melvster
most of them anyway
#
melvster
i could go into details how
#
melvster
if there's an overlap between my apps and the group's perhaps I could give a demo, but only if requested by a WG member
#
melvster
i could probably code up 1 use case over the course of a day too, so id have to see what was wanted
#
elf-pavlik
yes please demo some of the top user stories using solid :)
#
rhiaro
melvster: we've reached a point with the group where claims about what apis can aren't really entertained without demos of user stories. it would be awesome to see a demo of SoLiD performing a user story
#
rhiaro
s/apis can/apis can do
#
rhiaro
so far we haven't really had that, despite lots of people saying it's possible
#
melvster
ill just be an observer then
#
melvster
maybe during lunch or some minutes free i can show interested people what ive been working on
#
melvster
i can be coding during the sessions
#
melvster
if a wg member sees an overlap and requests more then I should be able to do that
#
melvster
it sounds like the main sessions will be about APIs
#
melvster
which is not that relevant to my work, as im using SoLiD for everything
#
melvster
if there's a particular use case you'd like demo'd perhaps I can do that on the day, tho
#
melvster
at the last meetup we coded up federated 'likes' during the session iirc
#
melvster
so i could possibly tackle 1-2 and explain how to use SoLiD to do that, if there's interest ... we'll see
#
melvster
obviously i wont be able to code up 90 use cases in a day
#
melvster
mainly looking forward to hanging out and seeing what you guys are up to
#
rhiaro
melvster: heads up - definitely will be interest!
#
rhiaro
elf-pavlik: do you think it would be useful for for us to add a side note on the agenda on wiki about which user stories melvster could demo with solid?
#
elf-pavlik
of course, as soon as he knows what he can demo
#
rhiaro
so melvster, could you let us know which user stories look most likely for you to be able to demo?
#
melvster
let's discuss on monday
#
Loqi
I added a countdown for 5/4 12:00am (#5678)
#
melvster
right now im setting up my new laptop and packing / getting ready
#
melvster
i got quite lucky
#
melvster
the laptop is also a tablet
#
melvster
which i didnt really know when i bought it
#
melvster
and all the apps work great on tablet too
#
melvster
in fact even better
#
melvster
touch is a big plus
#
melvster
when using material / responsive design
#
melvster
so been testing that side
#
melvster
my work is probably most closely related to payments group, but I hope there's some overlap with social
#
melvster
sure, but andrei will be there and wrote it, i guess i could do an integration
#
melvster
that's a good idea
#
melvster
in our eco system when one person writes a tool, all other systems can use it, it's an unusual paradigm in that sense, more like a web of apps, rather than, client API and server API
#
elf-pavlik
melvster, have you looked at http://indiewebcamp.com/micropub ?
#
elf-pavlik
you can try some of the micropub based apps http://indiewebcamp.com/micropub#Clients
#
elf-pavlik
i can authenticate using indiecert.net or indieauth.com for those which don't support distributed indieauth
#
elf-pavlik
you just need rel="me" on your homepage, please keep in mind http://indiewebcamp.com/indieauth#RDFa
#
melvster
demos are not till the evening on the last day
#
melvster
so plenty of time
#
melvster
elf-pavlik: i already support rel="me" but currently am testing with webid+tls, would be great if that could be supported too
#
melvster
doesnt work
#
elf-pavlik
do you have https?
#
melvster
waiting for letsEncrypt
#
melvster
but not sure i want to change my URI
#
melvster
cool URIs dont change
#
melvster
for sure not
#
melvster
startssl
#
melvster
doesnt even work in firefox iirc
#
aaronpk
it totally works in firefox, most of my certs are from there
#
melvster
chrome then
#
melvster
one of the browsers it didnt work in
#
aaronpk
also works fine in chrome and safari
#
aaronpk
and ios
#
melvster
without importing the ca?
#
elf-pavlik
melvster, if you don't make tiny effort of giving indieauth / indiecert a try, why someone would bother giving WebID-TLS a try?
#
elf-pavlik
you can always use indieauth with Twitter, Facebook and Persona since you can use all of them already
#
melvster
elf-pavlik: that's quite condescending given that I spent quite a bit of time working with indiecert giving feedback that was incorporated into the protocol, I make a lot of effort to help people, and am spending almost $1000 out of my own pocket to attend this meetup, my effort is generally not acknowledged or appreciated, that doesnt encourage me to do more
#
elf-pavlik
melvster, you can also use https://www.cacert.org/ just to get indiecert.net working
#
elf-pavlik
"my effort is generally not acknowledged or appreciated" what do you base it on?
#
melvster
elf-pavlik: you said, "you don't make tiny effort of giving indieauth / indiecert a try"