2017-01-18 UTC
# 01:01 Loqi KevinMarks: tantek left you a message 1 hour, 21 minutes ago: MozSF confirmed for HWC SF 2017-01-25 - go ahead and post the indie event and FB POSSE copy!
# 01:20 aaronpk hey ringmaster maybe instead of whining on twitter about how webmention doesn't work you should actually take a look at all the things that are built with it and have been working for the past several years
# 02:47 miklb I've known him for a very long time. Uncharacteristic.
# 02:48 miklb I need to do some debugging, because my notes are still posting as just site title and link. Not the text.Last test was explicitly set to not show link, and I know character count was < 140
KartikPrabhu1 joined the channel
# 03:06 miklb GWG that was from posting in the admin, not via micropub fwiw
marcthiele joined the channel
# 03:09 GWG I'm working to move the meta key from where I just put it to where the Micropub plugin stores it...in the microformat property format. mf2_syndicate-to
KevinMarks joined the channel
KevinMarks_ joined the channel
# 04:15 KartikPrabhu1 <phew> successfully moved site to Django 1.10 and sqlite3 and virtualenv!
KartikPrabhu, tantek, marcthiele and KevinMarks joined the channel
marcthiele joined the channel
# 06:48 Loqi miklb has 1 karma in this channel (15 overall)
marc_thiele, marcthiele, KevinMarks_, tantek and cweiske joined the channel
# 11:38 petermolnar !tell snarfed bridgy seems to be using the url as tag when it's marked up as <a class="u-category" href="/tag/tagname">tagname</a>, so when it's a POSSE to flickr the tag will look like http://domain.com/tag/tagname instead of tagname; how should I avoid this?
# 11:38 Loqi Ok, I'll tell them that when I see them next
# 13:13 Zegnat Your u- prefix implies that the URL is the identifier for the category, rather than the text string. So I don't think Bridgy is messing up
# 13:47 sknebel jonnybarnes: seems like your IRC Client doesn't cope well with japanese tweets
# 14:52 petermolnar !tell snarfed ping, because "Sorry, you've already published that page, and Bridgy Publish doesn't yet support updating or deleting existing posts. Ping Ryan if you want that feature!" - actually it's not an update, I wanted to post again because I deleted the posse copy.
# 14:52 Loqi Ok, I'll tell them that when I see them next
arush, Zegnat, singpoly1a, mindB and prtksxna joined the channel
tantek and miklb joined the channel
KevinMarks joined the channel
# 18:44 tantek ^^^ so that's the kind of thing that having a W3C REC can apparently impact
# 19:20 voxpelli Too bad so few initiatives around distributed social networks have been successful within Drupal :/
# 19:22 tantek also, Drupal adoption of Webmention support by default will put the head on getting Vouch solidiified, specified, and adopted
# 19:25 voxpelli tantek: core support won't really happen I think, I believe most ping mechanisms were removed from core many major versions ago
# 19:25 tantek makes sense to wait to strongly advocate that until Vouch support is included
# 20:13 aaronpk huh. what should the "name" of a repost be? I want it to be empty because all the content i'm reposting is in the repost-of object itself
# 20:13 aaronpk so right now it's using the implied name rules and resulting in a horrible mess of a name property
# 20:15 tantek aaronpk I thought that was all figured out / documented on /repost but perhaps not?
# 20:16 tantek I know KartikPrabhu and I did enough brainstorming at IWC NYC 2014 to figure all that out
# 20:16 aaronpk even when you include the reposted object as an h-cite?
# 20:16 tantek it's possible we didn't fully document from our session page to the repost page
# 20:17 aaronpk the only "how to" i see on the repost page just says "u-repost-of" and doesn't even mention h-cite
# 20:17 aaronpk searches for "repost" in IRC those days and doens't find anything
# 20:21 tantek hmm not sure how our markup didn't make it to the whiteboard or a photo
# 20:21 tantek tries to recall markup in his brain at the time he drew those whiteboard diagrams
# 20:22 aaronpk well for now i'm including an empty p-name tag to avoid the implied mess
# 20:22 tantek it should be the same name as the thing you're reposting
# 20:22 aaronpk i can't include the name of the thing i'm reposting without duplicating the content on the page (also not everything i repost has a name)
# 20:23 aaronpk ah yeah i guess everything does have a name because notes have names
# 20:23 aaronpk yeah, i include the reposted content in an h-cite
# 20:23 aaronpk so the name of hte thing i'm reposting is inside the h-cite
# 20:24 aaronpk which means in order for me to make it show up as the name of my own h-entry i'd have to *also* have that content outside the h-cite
# 20:24 aaronpk i don't have any examples of reposting articles yet, so far just tweets, plaintext notes, and a podcast episode (name + audio)
# 20:25 tantek hmm - what are other folks doing with theirs on IndieWeb Examples?
# 20:26 aaronpk kylewm uses "repost-of h-cite" and also doesn't set a p-name on his h-entry so has the same implied name prblem
# 20:30 martymcguire[m] hmm. no sure why Loqi didn't pick it up.
# 20:30 aaronpk omg it used parts of the URL that ben included in his post as the slug for his post
# 20:31 martymcguire[m] haha, oh, that could be a problem.
# 20:34 tantek aaronpk - either we duplicate the name from the repost-of destination into the h-entry of the repost, or we come up with some other way to mark up reposts
# 20:35 tantek and that being said, thanks for raising this issue
# 20:35 aaronpk so the text of the equivalent of our h-entry is modified to include "RT @username {original text}
"
# 20:35 tantek a-ha that makes sense! so that non-repost-aware consuming code will see the "RT @username {original text}
"
# 20:35 tantek and *that* is what I was looking for in our diagram
# 20:35 aaronpk for consumers that don't recognize the reposted object in the tweet
# 20:36 tantek we solved that fallback text for generic reposts
# 20:36 tantek and of course I left it as "an exercise for the reader" to determine the markup accordingly
# 20:36 aaronpk i'm just not sure how I would want that to *look* on my site
# 20:37 tantek if you're reposting a tweet, use the Twitter fallback text of the RT ...
# 20:38 tantek "♺ user.example.com: "--------..." user.example.com/permalink"
# 20:39 tantek the "what should the name property be?" question often has the same answer as the "what should the tweet POSSE text be?" question
# 20:39 tantek however that's "just" a brainstorm as neither KartikPrabhu nor I had implemented that - just figured out how we *would* implement it, including how it *should* look on our own sites
# 20:40 aaronpk for fallback purposes, i'd want to include the photo/audio/video properties from the repost on my h-entry as well
# 20:40 aaronpk so that readers that don't recognize the h-cite will still see the photo of my h-entry
# 20:40 tantek fallback is never expected to be as high fidelity as something that actually supports it
# 20:41 tantek you're not posting just the text, you're quoting it
# 20:41 aaronpk point is, if I'm showing the reposted post on my website, I want that to look the same in readers
# 20:42 tantek I'm not sure "look the same in readers" is a reasonable constraint (similar to cross-browser web design)
# 20:43 aaronpk we have very little publishing and consuming examples of summary. that's why i haven't added it to my site yet
# 20:43 tantek which should never be expected to be a complete substitute for the actual content or meaning being conveyed
# 20:44 aaronpk i might feel better about doing the "♺ user.example.com: "--------..." thing if it were in a "p-name p-summary" element
# 20:45 tantek note that the user.example.com is the original, not you
# 20:47 tantek (more UI / display brainstorming is good - what Kartik and I did was just a first attempt at something reasonable / non-sucky)
# 20:47 tantek if you can rethink / redesign something nicer presentationally, definitely interested in seeing it!
# 20:48 aaronpk well part of the problem is I'm happy with the presentation of reposts on my own website right now, so I have no intention of changing the way they *look*
KevinMarks_ joined the channel
tantek joined the channel
# 22:23 Loqi [Sebastiaan Andeweg] Day 3: relative urls
# 22:23 sebsel Don't know if it's worth another test, but he has a relative url in a folder.
# 22:25 sebsel also the updating of the redirect url... I don't know if I handle that correctly. I should indeed use that new url for solving relativeness, but should I also take the new url as target?
# 22:26 aaronpk whatever appears on your web page is what you should use as the target
# 22:26 aaronpk otherwise when the receiver checks if the link exists it would fail
# 22:27 aaronpk and it's up to the receiver to handle redirects of their own URLs
# 22:30 sebsel oh, I don't have to rewrite anything, because I was using that new redirected url only within ::discoverEndpoint() :D
# 22:36 Zegnat mumbles something about how he loves being on the edge of the spec and breaking people's implementations
tantek joined the channel