#bear.imcreated /container (+223) "Created page with "{{stub}} '''<dfn>Container</dfn>''' is a application management tool used to bundle the application, configuration and operating system into a single self-contained bundle...."" (view diff)
#bear.imcreated /kubernetes (+159) "Created page with "{{stub}} '''<dfn>Kubernetes</dfn>''' is a container orchestration environment used to manage container environments. == See Also == * https://kubernetes.io/"" (view diff)
KartikPrabhu and j12t joined the channel
#bear.imcreated /helm (+179) "Created page with "{{stub}} '''<dfn>Helm</dfn>''' is a tool to manage the kubernetes configuration files for continuous integration and deploy within kubernetes. == See Also == * https://helm..."" (view diff)
#tantek^^^ seeing people that I've never heard of post about getting #indieweb things working amongst themselves is pretty cool (and evidence of it not just being an echo chamber)
#Loqiwillnorris has 1 karma in this channel (21 overall)
#tantek.comedited /code-of-conduct (-38) "Similar Codes of Conduct - since all examples are based on IW CoC, note that explicitly instead of as exceptions" (view diff)
#tantek.comedited /acquisition (+433) "add self with a couple of examples, found a 2016-02 post for a rhiaro acquisition as a good enough at least since, lowercase generic red see also" (view diff)
#ZegnatHmm. I wonder if Micropub clients can do something about the Twitter policy change. But I guess not.
#LoqiZegnat: [eddie] left you a message 12 hours, 23 minutes ago: my concern about using repost for “save later”. Most existing Micropub clients are just going to send the url under repost-of. But then Aperture doesn’t currently parse extra urls in posts. So my “save later” feed would end up with the repost url but no content (no title, content, etc) from the initial page I wanted to read later
Kaja_, [kevinmarks], petermolnar, yoroy_, strugee, cweiske2 and aaronpk_ joined the channel
#Zegnat!tell [eddie] I actually didn’t check what current Micropub clients send for repost posts. I actually expected nested h-cits, but that expectation might have been wrong!
petermolnar, petermolnar_, globbot, [kevinmarks], leg, ScalaWilliam, eli_oat and [gerwitz] joined the channel; mblaney left the channel
#[gerwitz]Hello, IndieWeb… I’m asking [manton] to use webmentions to inform my site that content from it has been syndicated.
#[gerwitz]Syndication is not quite “mention” so I’m curious if anyone things this is [in]appropriate
[manton] joined the channel
#[manton]Repeating what I said in the Micro.blog Slack... The main issue I can think of is that the Webmention would probably create a comment in WordPress (for example) that had the same content as your original post, since it's kind of like you linking to yourself. To look good, I think the receiving code would need to display it differently.
#sknebelWe are talking about micro.blog following the feeds and creating syndicated copies?
#sknebelI agree with Manton that it'd be odd as a comment
#sknebelBridgy gives the link in return to the webmention triggering it and the sending code can then handle that, but general micro.blog syndication doesn't have that
snarfed joined the channel
#sknebelSince syndication links are a property if a post, micro.blog could make a micropub request editing the post, but I don't think there is prior art for that and it's kind of bad giving out full ability to post to the blog
#sknebelAlthough I guess you have to be prepared to get webmentions from syndicated copies anyways, since you can't stop them from being sent by someone
[tantek] joined the channel
#[gerwitz]I’m in shock this isn’t already a “thing” … certainly others have been annoyed by syndication targets that don’t return a URL?
#[gerwitz](My content updates are async, so I *like* the microblog passive syndication. I just want to be informed of the resulting location)
[kevinmarks] joined the channel
#[kevinmarks]I'm a bit ocnfused - do you want something like bridgy, where replies to the syndicated copy show up on the original?
#[eddie]I think there are potentially two different issues from your original question on micro.blog [gerwitz]. I believe one issue is you want to be able to respond from your site to someone whose post url is NOT a micro.blog post URL. The other issue is you want to allow people to find your micro.blog post from your site if they want to reply to you. Does that sound accurate from your initial request on micro.blog?
#[gerwitz]Not the former. (I was only diagnosing the existing webmentions behavior.)
#[gerwitz]…and also, eventually, other syndication targets.
#[eddie]Passive syndication is definitely a new area in the IndieWeb. Most syndication has been done either through Micropub forwarding or Webmentions
#[eddie]Both of which respond with the syndicated url
#[eddie]So I think there definitely must be a solution, the challenge will be thinking through how passive syndication works and the least obtrusive way to integrate that both for people adding support for it AND for not interfering with existing expected behavior.
#[eddie]I think the two most obvious solutions (which have already both been mentioned above) is 1: Adding rel=“canonical” to the micro.blog page on the link to the original post. Then sending a webmention. The expectation would then be if a webmention is received with the target url being part of a rel=canonical link, then it is considered a “syndication notification”
#[eddie]2: Micropub Update, this would mean micro.blog the hosted service holds an access token to your website with “update” scope. Whenever it parses a feed and creates a new post, it would send a micropub updated request adding the syndication value containing the URL of the micro.blog copy (example: https://www.w3.org/TR/micropub/#add)
#[eddie]The downfall of #1 is that every existing webmention implementation might have to be adjusted (affects existing behavior)
#[eddie]The downfall of #2 is that micro.blog holds a token that can edit any post on your entire site AND your micropub endpoint has to support both JSON posting AND the update ability, which not a lot of sites support
#[eddie]It’ll be good as more IndieWeb veterans hop on throughout the day read this conversation and provide their input because the dangers we have to balance are making the bar too high for people to support versus messing with existing functionality 🙂
#aaronpki kind of like [eddie]'s description of #1 there
#aaronpkthe only trick is it would be surprising to most people to get a webmention from their syndicated copy right now, so maybe it could be an opt-in feature
#aaronpkit does feel like the least obtrusive option though
#aaronpki've been wanting to set up an indieweb aggregator for a while now, where it crawls everyone's site listed on /irc-people and publishes a new feed from everyone's sites any time they mention indieweb keywords. so that is a sort of automatic syndication as well.
#aaronpkthis mechanism of linking to the canonical with rel=canonical then sending a webmention feels appropriate for that as well, especially since rel=canonical is already a thing
#sknebelYes, and any syndication can already send a mention of it links back
#sknebelLike, I could go through people's tweets and send mentions through their bridgy equivalent
#ZegnatIf I get a webmention and I parse the page the webmention is from, do original post discovery, if it turns out it is my own post take action accordingly?
#[eddie]That’s true, that would probably work: “in the h-entry that represents the POSSE copy, look for a link with “u-url” and “u-uid” - use that href as the original post URL.”
#[eddie]The only issue is most webmentions currently would generate a comment
#[eddie]Would we then say if a post matches the original post discovery algorithm, it should be considered a syndicated copy and not be used as a response?
#[eddie]It does seem to match up with what we were just discussing because what we discussed would be step #3 of the original post discovery algorithim
#sknebelSounds like a solution. Display would be a question (I'd probably limit them showing up to approved sources, so people can't randomly create the impression on my site that I syndicated to them
#ZegnatLots of people are alreadyd filtering out likes and reposts, some even do content parsing to filter out reacji. If you are already running the mf2 parser, you can easily implement parts of the original post discovery alg
#ZegnatAnd since link relations are parsed as part of an mf2 parser, for the first two checks, you can rely on the mf2 parsing you are (maybe) already doing. No need to implement your own HTML parsing there.
#ZegnatLots of discussion still on going on that discovery alg though, not sure how mature it is or if tantek was thinking of converting it to a W3C Note like Post Type Discovery. But something to keep in mind :)
#[eddie]We could at least follow the first couple of steps (check for u-url/u-uid and rel=canonical because those are pretty flawless
#[eddie][manton] How does that general direction of this sound to you. Opt-in Webmentions to avoid throwing Webmentions at people that might not be ready for this kind?
#[manton]Catching up on this discussion. I like where it's going. And yeah, all Micro.blog conversation links have u-url. Happy to add canonical too, but maybe it's not necessary.
#[eddie]It looks like my replies to Stephen Hackett’s posts didn’t go through
#[eddie]so it looks like any replies to an initial post don’t work
#aaronpkoh, i think it's when i find something from the micro.blog timeline and the URL is a canonical URL and then I can't get my reply syndicated to micro.blog because their site doesn't link to their micro.blog URL
#[manton]The weird thing to understand about Micro.blog conversation links: it shows the same thread for any URL that links to the same conversation, but depending on the post ID in the URL, it changes the markup to apply h-entry/etc. to the post being linked to.
#[manton]That way parsing the page when you get a Webmention works, even if each reply doesn't strictly have it's "own" page.
#[eddie]That’s a good solution. The other one is fragment ids, but selectively adding h-entry works too
#[manton]Yeah. It would make more sense in the current implementation if the page visually changed to highlight the linked post, which I want to do.
#[eddie]So as far as replies to micro.blog, a cross-posted reply (original post on aaronpk’s site, reply post on my site, attempting to syndicate reply to micro.blog) is on GitHub here: https://github.com/microdotblog/issues/issues/75
#Loqi[EdwardHinkle] #75 Adding support for cross-site replies
#LoqiDropbox is a file hosting silo that provides js;dr URLs for the files hosted there that display just file name and a spinner unless you load some number of scripts from various domains https://indieweb.org/Dropbox
#gRegorLoveHm, it is in the "Public" folder and worked before. Guess they changed/broke URLs at some point. It shows "Only you" in Dropbox now, and an option to create a link.
#[eddie]So the idea is since with passive syndication you don’t have a direct response, we had to figure out how to get the syndicate-to url back into our sites
#dgoldYes, I see that now with your most recent two issues
#[eddie]For my site m.b is like syndication because I actually created an internal syndication target that when selected adds the post to my feed that m.b reads
#[eddie]But me (and everyone else) is still missing the piece of having the syndicated link on our pages
#[miklb]I think I’m going to be moving to a similar model [eddie]
#[eddie]miklb I would definitely recommend it. I love how it’s been working for me
#[eddie]It allows me to not spam my m.b feed when I have a lot of watch posts or read posts
#[eddie]But important ones (like Starting or Finishing a book or an exciting TV episode) I can still post
#[eddie]In fact, I currently store a u-syndicated property on the post with the uid= my micro.blog url. This allows until I get the actual syndicated url back per above discussion, people can see the post is syndicated to micro.blog and at the very least get to the m.b timeline (see the globe on https://eddiehinkle.com/2018/02/22/4/note/)
#Loqi[Eddie Hinkle] I built my first bot a year ago today, I’d definitely love to work on more bots but haven’t had the right project to work on yet.