Loqi[Ricardo Mendes] Gif or Jif ? "The Oxford English Dictionary accepts both pronunciations," - are you sure ?
1 min read
Gif or Jif ? "The Oxford English Dictionary accepts both pronunciation...
tanlaanI'm just thinking of taking this on with the Unix philosophy and making every part very small but functional for what it does to be able to piece it together in the end
[shaners]I added JSON Feed to the CrimethInc. site today. It took two hours and was pretty easy. Most of the time was spent digging through my own code. The spec is clearly written and stuff.
[shaners]I added it for the same reason I have an Atom feed and @adactio has AMP views. Because why not. It’s just another view of the same data. It’s an experiment. And it was quick and easy.
tanlaanI've been trying to figure this out by trial and error. On the install of Known, I had to change the permissions for my Uploads folder. I tried all the variations for permissions until I ended up with 777 which doesn't sit right with me. I figured even a 775 would have functioned if it was in the same group as the file itself. Anyone who has used i
miklb[colinwalker] looking, if you are using semantic linkbacks, I think you could add a commentmeta query to include semantic_linkback_type => reply to include those that aren't a webmention type.
miklbno, just authorize bridgy for what ever services, and then the bridgy publish plugin has options for which services to include (Twitter, FB, Flickr). you then get checkboxes on the new post screen which ones to send that post to.
Loqi[[shaners]] I added JSON Feed to the CrimethInc. site today. It took two hours and was pretty easy. Most of the time was spent digging through my own code. The spec is clearly written and stuff.
voxpellithe WebSub spec would need to be updated for JSON Feed to currently work with it as well – or JSON Feed to change so that WebSub doesn't require any changes
voxpelliI do think that something like JSON Feeds should have a generic link-property on both feed and item level though, to enable extensions like WebMention, payment etc to be easily and independently injected in there
aaronpk"the allowed values of additional members will be expanded (e.g. a collection link may support an array of values, whereas a self link does not)."
miklbpfefferle those functions in semantic linkback are great, I use them for my facepiles, but colinwalker was trying to get replies that came in as webmentions but loose the the type as they are now standard comments along with other webmentions
[colinwalker]Just had a thought (I know, careful) - as webmention replies have comment_type removed can't we just use semantic_linkbacks_type to check if a reply and then rewrite the value to the table?
ben_thatmustbemesigh, its a little disturbing to see people worried about backcompat for a spec so early in its infancy. But thats what happens when you publish something and give it a version and make it look like you are expecting it to be implemented before anyone outside of the editors have had a chance to review it
ben_thatmustbemei suppose the idea case is to have a simple, straight forward translation from mf2 - > jf2/jsonfeed (in some unified way) and back to mf2
barpthewire, mlncn and [manton] joined the channel
[manton]voxpelli and aaronpk: The way I read Julien's comments was that WebSub's Discovery section in particular might need to be updated. Not sure if there was something else... But I think the "Publishers _must_ implement" could be relaxed a little to "_should_ implement", giving room for something like JSON Feed's "hubs" (or the proposed "links"). I plan to write up an issue for this.
[manton]kevinmarks I don't think that would technically matter. If you strictly follow what WebSub says, it leaves no room for specifying a hub except in XML or HTTP headers.
[manton]aaronpk I totally get that. My first thought was "links" as well. But my view is that WebSub is critical to improving feed readers, and so kind of promoting it to a core part of the spec will give it more attention.
aaronpk[manton]: my only problem with "links" is that it's almost too generic of a term. that's why "rels" is nice because it's specifically just the links with a rel attribute. otherwise it could be misinterpreted that you're supposed to put *all* links in that array
[manton]It's really a trade-off with how generic something should be. There are many fields that could be abstracted to more of a type/value format, to be more flexible, but then you make the document more verbose. I don't think it's always clear. (I'm sure many folks would want next_url to be moved to links/rels too.)
ben_thatmustbemeas i realized doing so while keeping mf2 parsing in mind would mean there would be a simple way to go mf2 -> jsonfeed/mf2 in an automated way
voxpelli[manton]: to the credit of jf2 style is that it enables a parser/consumer to not just consume feeds, but user profiles, comments, events and more
voxpelli[manton]: and one thing that the post-blog social web has that the blogs didn't is that it has lots more types of data, not everything is articles anymore
voxpelliRight, but it can express events, profiles etc so it makes it easy to build a full Facebook on top of the format rather than be limited to just a Twitter/blog style feed of simple notes
[cleverdevil]I mocked out JSONFeed on my site last night and struggled a little bit with mapping different types of posts. I ended up adding in an extension to each item that contains some extra metadata. https://cleverdevil.io/?_t=json
ben_thatmustbemethe biggest difficulty jsonfeed will hit is that its going to be an uphill battle fighting existing standards that are quite well established
dougbeal|iOSI have a post from instagram via OYG and syndicated to Facebook. Facebook likes find the post fine, but the instagram ones say no webmention targets (via bridgy https://brid.gy/instagram/murderofcro.ws#)
aaronpkhehe yeah there are some extra pieces required if you are doing it from scratch. I have the advantage of already having continuous gps location on my server
LoqiJust generated this week's newsletter! You still have a few minutes to make changes, and I'll re-generate it 10 minutes before it gets sent out at 3pm Pacific time. https://indieweb.org/this-week/2017-05-19.html
ben_thatmustbemei have left my personal blog somewhat a mess as I have been working to figure out all of the indieauth stuff again for my rewrite to laravel
tantekI'm posting another RSVP, to exercise another special-case RSVP use-case that I have, to RSVP to the secondary copies, and to see how RSVP de-duping is handled :D
tanteksharing this here from chat channel because I think the person that said it would be ok with me doing so (i.e. not written by me) -----> "A hands-on conference focused on making your own website and connecting it to social media - so you don't have to store everything you make on Facebook"