tanteklet's do Eventbrite because I think (based on purely guessing) that it will notify more people "in our networks" that it is going on and that we've signed up to go to it
tantektito is likely good to lower the barrier for sign-up for people who don't have a domain, yet one of our challenges is getting better about getting the word out
tantekI think that's a great idea for /business-models for folks in the community - especially those who have setup their own site to run with as little hassle as possible.
tantekIn particular, given that there are whole businesses and independents who make money from setting up WordPress for businesses, I bet there is similar opportunity for someone to do so with setting up Known for businesses.
lukebrooker, j12t, nitot, shiflett, snarfed, cuibonobo, martinBrown and andicascadesf joined the channel
Loqi[shaners]: tantek left you a message 1 day, 6 hours ago: the reason I asked (and brought it up) is because in reading the CoC presuming good intentions, I have difficulty seeing anything explicitly against the letter of what it's stating. Hence looking for more intuitive/gut guidance what is the "right" thing to do, and how we can encourage that in CoC. http://indiewebcamp.com/irc/2015-11-02/line/1446451547265
ttepasse, snarfed, wolftune, sammachin, bret, shiflett, j12t and yakker joined the channel
kylewmhaving a little bit of a moral dilemma, I want to change my CMS kind of radically but don't want to worry about breaking it for my 2-3 users (at least until I figure out if my changes are good or bad)
davidpeachyer twitter I used guzzle. but with flickr i needed to do the whole create token request the have to be redirected to manually accept permissions page. So I thought Id give it a shot.
tantekkylewm: sometimes a change is so radical it's quicker to show what the end result would/should look like rather than attempt to argue every change there iteratively
tantekkylewm: in addition, it looks like jf2 will satisfy the 80/20, and then I'd expect that *someone* might develop a jf2->as2 proxy for folks that really want to go to the extra work of consuming as2
davidpeachso i added a comment to my test post on flickr. And bridgy has found it fine and has it in my bridggy dashboard, but it cant find a webmention endpoint it says.
tanteksingpolyma: correct. easiest for publishers is to publish mf2 which has a defined parsing model to jf2, which then API consuming code can trivially use
aaronpkthe current thinking is to link to the jf2 URL with a rel value. that lets you offload that conversion to a separate thing, like how pin13.net/mf2 works
tantekkylewm: depends on perspective. jf2 is much closer to the JSON return values from various silo APIs - if anything, jf2 is *less* revinvention, but rather based on existing JSON API patterns.
kylewmanyway, solving it from both sides; that does make sense. if the goal is to experiment to inform the development of a unified standard then i'm all for it
singpolymasnarfed: xml, understood properly, is more than just a generic format. it's a way to combine different specialised formats togther in a way that will let consumers that understand both benefit from the features of both
kevinmarksif let johnsStreet = john.residence?.address?.street { print("John's street name is \(johnsStreet).") } else { print("Unable to retrieve the address.") }
Loqisingpolyma meant to say: kylewm: I mean. some mods wanted more tools (and are getting them), but users can report posts to mods, and mods can remove posts or users
nitot, snarfed and [kevinmarks] joined the channel
voxpellithe design goals of JSON API are really focused on what makes an API good, not necessarily what makes a data format good – I like many parts of it though, just something to be aware of
voxpelli[shaners]: "A document MUST contain at least one of the following top-level members: data, errors, meta" – so your gist is for a JSON API resource object rather than a top level object, right?
voxpelli[shaners]: the only permitted top level attributes of a resource object apart from "id" and "typ" in JSON API is "attributes", "relationships", "links" and "meta" as I understand it
voxpelli[shaners]: In the spirit of JSON API I would probably add a self-link or similar to the relationship, pointing to the URL of the h-card, but that would perhaps not be a 100% representative of the data
aaronpk[shaners]: i'm expecting anyone who's publishing mf2 html will use a service to convert it to jf2. I also suspect some people will output jf2 directly.
aaronpki still don't really get the original question. I link to the pin13.net mf2 service because that's easier for me than also publishing JSON. I will do the same for jf2
voxpelliwhy should it be up to the publisher to pick a service to convert from mf2 to jf2? isn't it better for a consumer to decide whether it wants to parse the data itself or have someone else do it for them?
aaronpkalso it doesn't have to be a conversion from microformats to jf2, you can publish jf2 without microformats behind it, which i'm expecting some people will do
voxpelliI'm, very surpised to hear this when publishing data as RSS has generally been something that people have been fairly critical about due to eg. DRY-concerns. What differs jf2 from eg. RSS?
aaronpkdavidpeach: in contrast to oauth1, oauth 2 is actually simple enough you don't really need a library. also oauth 2 APIs tend to be different enough that a library won't help much
voxpelliok, I really shouldn't argue about this right now, too tired, but I'm really not convinced why people should suddenly start publishing jf2-data – as a format for API:s I'm pro it, as a format for publishing data on the web I'm against
aaronpkvoxpelli: it's still early for this for sure. but for starters, it can serve as a simplified json representation of microformats if nothing else.
voxpelliaaronpk: I'd say that it's especially too early to start publishing such links – if consumers start to widely adopt the jf2 format then it could make sense to have clients be able to publish it directly
[shaners]Actually, i think i’m doing it wrong. I think all the nested stuff is meant to go down at the bottom in an included section. Instead of nested inline.
voxpelli[shaners]: I wonder if one could consider the url-attributes redundant as the id:s of those types will always(?) be url:s? and if they are not URL:s then a self-link would probably be more representative anyhow?
voxpelli[shaners]: depends on whether you reference the same object multiple times in the same response – then you want a reference to a top-included object instead
voxpelliok, maybe lets keep it that way and have the discussion about exposing it go by another name as that's pretty separate from the actual format used? One could eg. expose the current mf2 JSON as well
voxpelli(if we should take notice from the OStatus / Activity Streams communities also – any mandatory switching of data formats is a pretty dangerous move – I have rewritten too much code because of such things)
sandroHmmm. Is anyone ( aaronpk? rhiaro? ) running something on their server that would let me be notified of new content without polling? I feel like I'm missing something obvious.
LoqiThis page is a description of a minimum viable subset of PuSH 0.4 that is sufficient for supporting the IndieWeb use cases of realtime publishing feeds of h-entry posts, and subscribing to & receiving notifications of updates from those feeds https://indiewebcamp.com/how-to-push