2016-05-04 UTC
jfrndz joined the channel
jfrndz, yakker and yakker_ joined the channel
# 00:51 Loqi [indieweb] "IndieWeb and Android Mobile
jfrndz joined the channel
yakker, yakker_ and [benatwork] joined the channel
# 01:20 [benatwork] I come from a land down under
# 01:20 [benatwork] It’s benwerd. I messed up my Slack bridge setup.
# 01:21 [benatwork] (Did that song make it to the US, or was that yet another dead reference?)
# 01:22 bear chuckles at a song reference that he actually got
[shaners] joined the channel
# 01:24 [shaners] benwerd: That song is one of like 6 points of reference that Americans™ have about .au
# 01:25 kylewm tantek: woodwind does not support deletes because it only fetches the h-feed, not the individual permalinks
# 01:25 [benatwork] Are the others putting steaks on the barbie, Crocodile Dundee, Mad Max, Kylie Minogue, The Rescuers Down Under and Outback Steakhouse?
# 01:25 tantek so what happens if something disappears from a refetch of the h-feed? do you dump the previous cache?
# 01:25 [benatwork] British people get to add soap operas to those.
# 01:26 [benatwork] shrimps makes sense
# 01:26 [benatwork] is that a sentence
# 01:26 kylewm tantek: no it does not remove items that disappear from the feed
# 01:27 kylewm I think that's why we talked at length about dt-deleted
# 01:31 tantek kylewm if you saw a dt-deleted, would you check the date and time to automatically hide the post when the dt-deleted time has passed? even without refreshing?
# 01:31 tantek that way a reader could respect an expired post too
# 01:32 kylewm I guess? It would definitely be more work than just yes/no
# 01:33 Loqi Expiring content is content that is only temporarily (ephemerally) relevant, and also part of a larger post, that can and should be (preferably automatically) removed once a particular datetime has passed (the expiration date) https://indiewebcamp.com/expiring
# 01:36 kylewm can someone explain to me what celebrities do on snapchat?
# 01:36 kylewm that is the part of that equation i can't get my head around
# 01:36 [benatwork] They broadcast
# 01:36 [benatwork] There are very well-subscribed channels
# 01:36 [benatwork] It’s a lot like TV
# 01:37 [benatwork] Snapchat started as person-to-person ephemeral messages but now it’s more like shows: you broadcast video to your channel and lots of people can watch it. It still goes away so they need to watch the next show.
# 01:37 [benatwork] Both Democratic presidential candidates have a Snapchat channel.
# 01:38 [benatwork] Looking at my list, the three people who have updated their channels today are Bernie Sanders, Hillary Clinton .. and Evan Promodou
# 01:38 [benatwork] (One of those is not like the other)
# 01:44 kylewm goes to snapchat something, probably cat related
# 01:49 kylewm accidentally records his phone vibrating which is SUPER LOUD
[kevinmarks] joined the channel
# 02:31 [kevinmarks] Put the tombstone in the feed. Though that does likely mean you need to preserve the slug
mlncn joined the channel
# 02:38 bear hmm, I just need to implement hide/unhide
yakker joined the channel
# 02:54 [kevinmarks] You make the post into a tombstone, then send out push, wait for crawl
# 02:54 tantek after an interval everything falls off the feed anyway right?
# 02:55 tantek kevinmarks do you know if Atom has a way of indicating deleted posts explicitly?
shiflett and yakker joined the channel
yakker and mblaney joined the channel
# 03:44 mblaney sorry after re-reading most of the logs on the day mentioned above, I realize there's still the problem of tombstone posts appearing in feeds.
# 03:45 mblaney so there needs to be a standard way to recognize a deleted post.
# 03:46 mblaney p.s. I've never heard of The Rescuers Down Under OR Outback Steakhouse :-P
# 03:57 tantek just skimmed rfc6721, seems a bit overdesigned.
# 03:57 tantek really don't need a whole another namespace, element, and such
# 03:59 tantek mblaney: agreed. and recommended publisher behavior that does at least something for backcompat with readers that don't support delete
# 03:59 tantek we can do that by adding a property as has been suggested before like dt-deleted or u-deleted
gRegorLove and snarfed joined the channel
# 04:09 mblaney tantek: I think I agree with the need for dt-deleted or u-deleted more now...
# 04:10 mblaney one other idea, a zero-new-markup option would be to have empty p-name and e-content, which implies a deleted post.
[kevinmarks] joined the channel
# 04:11 [kevinmarks] It makes sense in the feed especially. A 410 on the standalone post is clear, but knowing to remove it from the feed is key.
yakker_ joined the channel
# 04:19 mblaney my current criteria for replacing an item in a feed is just that the permalink matches (and obviously an old item would need to be reinstated to the feed).
# 04:22 mblaney tantek do you mean in terms of the permalink changing? that would create a new item in my feed.
# 04:22 GWG Does anyone have an example of a webmention source verification function that uses per-media-type rules as opposed to a pure text search?
# 04:24 tantek mblaney yes, like dropping the slug since it often has content
yakker and yakker_ joined the channel
# 04:26 mblaney I should probably support matching id's in preference to permalinks, but there's no concept of id's in microformats is there?
# 04:39 tantek interesting, I could have a u-uid of the post without the slug, and the u-url permalink with the slug
# 04:40 tantek so when I deleted the post I would redirect the u-url permalink to the u-uid
# 04:40 tantek but either way a reader would always see the same u-uid for the post
[kevinmarks] joined the channel
# 05:10 mblaney kevinmarks without either a consistent permalink or uid I'm not sure how else readers can match deleted posts.
# 05:16 gRegorLove Anyone have an implicit h-feed -- a stream with multiple h-entry but no h-feed mf2?
# 05:32 bear yea, I need to figure out how to redirect uid's to slugs and back for a static site
dogada joined the channel
# 05:33 bear oh poo, typed that and realized that I can just have uid files with a meta http-equiv redirect
KartikPrabhu and dogada joined the channel
loic_m joined the channel
yakker, cweiske, cweiske2, John_Duh, prtksxna, miklb, friedcell, Pierre-O and rMdes joined the channel
# 09:50 rMdes finally solved my dreamhost issue not forcing HTTPS :)
loic_m and adactio joined the channel
nitot_, nitot, cyberjar09 and hs0ucy joined the channel
BuffaloBill1 and mlncn joined the channel
snarfed joined the channel
hs0ucy, sknebel, wolcen_idle, dogada, shiflett, Pierre-O, hongpong, John_Duh, chrisbergr, wolcen and friedcell joined the channel
tantek, snarfed, friedcell and KevinMarks joined the channel
snarfed and j12t joined the channel
# 16:37 kylewm heh, noticed "rsvping" on the wiki. good name for indie rsvps
snarfed, Gold and [kevinmarks] joined the channel
snarfed, adactio and friedcell joined the channel
wolftune, snarfed and friedcell joined the channel
calumryan joined the channel
snarfed and tantek joined the channel
# 17:53 Loqi [Tim Baxter] Meaningful CSS: Style Like You Mean It
adactio joined the channel
# 17:58 KevinMarks Interesting as it is implying that styling microformat classes makes sense, whereas we have avoided that somewhat
Pierre-O, hongpong and gRegorLove joined the channel
# 18:23 tantek KevinMarks, no we have not avoided "styling microformat classes", we started with that approach, and then over time learned that that actually harms indepednent maintainability of form/content vs style/presentation.
# 18:24 tantek We *now* recommend avoiding directly styling microformats class names due to that longterm real world experience.
[shaners] joined the channel
# 18:40 tantek pretty sure I captured a bunch on the microformats wiki
# 18:40 tantek I mean that ^^^ is a pretty good summary of the biggest issue
# 18:42 tantek I'll restate to be clearer: styling microformat directly classes actually harms (makes it much harder to keep) independent maintainability of form/content vs style/presentation, which ends up causing the microformats to break, or the presentation to break when you change just one or the other, which can happen easily in larger teams where different people work on different pieces.
# 18:43 tantek so the one obvious exception is when you're just doing a one-off site that you don't expect to change (the content and/or presentation of)
snarfed, snarfed1 and hongpong joined the channel
j12t, snarfed and [shaners] joined the channel
# 19:07 [shaners] tantek: sounds like a team size / communication problem, not a technical
# 19:07 Loqi [shaners] meant to say: tantek: sounds like a team size / communication problem, not a technical one
# 19:07 tantek shaners - I have had this problem with a team size of 1
# 19:08 tantek but you're right that team size makes it worse
j12t_ and snarfed1 joined the channel
# 19:14 KevinMarks I certainly experienced it when adding mf1 calsses to projects, and having them removed later because nothing was styling them
# 19:14 KevinMarks though the real answer there is to add a parse and validate test tot he project
chrisbergr, marcthiele, KartikPrabhu, snarfed1, tantek and [kevinmarks] joined the channel
KevinMarks joined the channel
# 20:01 aaronpk now i don't have to hesitate to open twitter.com links anymore
# 20:02 GWG I have been working on the Webmentions plugin and Pingbacks in WordPress Core because of a common problem
# 20:02 GWG Without Mf parsing, what should a pingback or webmention display as content?
# 20:03 tantek GWG, pingback, undefined. Webmention? I'd say nothing.
# 20:03 tantek If someone is implementing webmention, they should also be publishing microformats2.
# 20:03 tantek if not, it's effectively an error, and better to let them know about it, and encourage them to fix it
# 20:03 GWG pfefferle solve this with the text, "This was mentioned on example.com
# 20:04 tantek sure you could do that too. maybe even with "Want your comment and author credit to show up? Use Indiewebify.me to validate your microformats"
# 20:05 GWG I am wondering if I should propose that the This was mentioned text replace the existing pingback text in WordPress Core
# 20:06 kylewm whoa, I did not expect them to ever change mobile.twitter.com
# 20:06 GWG My proposal is that it shows the human note, and that can be enhanced by the other changes I am making.
hongpong joined the channel
# 20:09 aaronpk it's still much lighter weight than main twitter.com
# 20:10 kylewm the situation before was the mobile site hadn't been updated in years, didn't show images, polls, etc.
# 20:13 [kevinmarks] There are different presentation styles appropriate for note and article posts
# 20:14 tantek sure, and that assumes microformats also. GWG was asking about some sort of default in the absence thereof.
# 20:15 GWG I am trying to get php-mf2 into WordPress as part of the SimplePie update
# 20:16 tantek how easy is it to update it with bug fixes once you get it in?
# 20:16 GWG So, I am trying to work on different fronts
# 20:17 GWG tantek, ticket in each version release.
# 20:18 GWG If not, then it continues to make sense to make microformats a plugin
# 20:18 GWG That is why I am working on improving the Pingback handler to work nicer with plugins
# 20:19 GWG At the same time I am trying to mirror those changes in the existing plugin code
# 20:20 GWG If I fail, no one will say that I didn't try
almereyda joined the channel
# 20:21 GWG They are working on early tickets for WordPress 4.6, so this is a good time to try and get eyes on patches.
# 20:24 tantek snarfed, FYI, economic reason NOT to upgrade your OS(X) - because doing so makes your computer not work with your printer that you paid for.
# 20:24 tantek "Free Update" is actually a lie. It's a "Assett loss update" since you can no longer use an assett you purchased.
# 20:26 tantek KevinMarks - Micropub has multiple interoperable clients for Creating, Deleting, Updating posts
# 20:27 GWG I proposed an update to 1.4 for the h-feed parsing
# 20:27 GWG gRegorLove, only if you also bundle php-mf2
# 20:27 hongpong i thought simplepie was abandonware
# 20:28 gRegorLove hongpong: Nope, our own mblaney has been making lots of updates!
# 20:28 hongpong cool that is good to hear gregorlove
[benatwork] joined the channel
# 20:29 [benatwork] Would simplepie with mf2 mean that feedwordpress reads h-feeds? I’d imagine so. And that’s used in far more places than you’d think.
# 20:29 GWG gRegorLove, ironically, I think the old maintainers were WordPress people
# 20:29 gRegorLove I'm surprised SimplePie is in WP. A bit unfortunate that php-mf2 has to be included separately with SimplePie though
# 20:34 aaronpk the bit of work i did on that for Monocle and a couple other small things made it really obvious how much variation there is in the way people publish h-feeds or lists of h-entrys
# 20:34 aaronpk it would be good to collect these into a test suite, since otherwise each implementation is going to only successfully handle a subset of feeds
# 20:35 aaronpk i've started to collect tests in XRay that capture those
# 20:36 aaronpk right. but hearing someone say "I added h-feed parsing to X" makes me nervous because there isn't an actual definition of what an h-feed is
# 20:38 aaronpk one of the issues is determining the author of a post in a feed
# 20:38 aaronpk the other is how to determine whether a list of h-entrys is actually a feed or not
# 20:39 KevinMarks we found the other day that atom is worse than h-feed for authorship
# 20:39 aaronpk there could be a per-post author, a feed-level author, or if there is no h-feed and only a list of h-entrys, then there might be a sibling h-card that is the author
# 20:39 gRegorLove Based on what I've done with picoFeed so far, I think I'm content to just use the domain name as the fallback author if there is no h-entry author or h-feed author
# 20:40 aaronpk i'm guessing it will fail pretty bad because I 1) don't have a top-level h-feed, and 2) am step 7 on /authorship
# 20:41 gRegorLove I've got it parsing implicit h-feeds, so that part isn't a problem.
marcthiele and friedcell joined the channel
loic_m joined the channel
# 21:14 gRegorLove Next thing I'm thinking about is parsing the language of an h-feed. Guessing that's going to be tricky.
# 21:19 gRegorLove It should be easy to add everything except for #4 since I can mostly use the code I wrote for ProcessWire Webmention
# 21:19 gRegorLove This also depends on the project accepting my PR. They seemed open to the idea, though.
shiflett joined the channel
wolcen_, miklb and mlncn joined the channel
# 22:28 gRegorLove I think Spotify and Pandora (probably others) have been doing similar
[shaners] joined the channel
# 22:32 [shaners] aaronpk tantek et al: Any idea how / where the current IWC logo is licensed?
friedcell and [benatwork] joined the channel
# 22:35 [benatwork] KevinMarks: the sooner someone topples the iTunes podcast db the better
# 22:35 aaronpk IIRC it's CC0 by virtue of being uploaded to the wiki directly
# 22:38 [shaners] aaronpk: That’s certainly implied. But unless Crystal uploaded it to the wiki, it’s technically *not* CC0.
# 22:41 [shaners] aaronpk: Can you ping Beasley to see if she has any of the original authoring files for the logo. For posterity.
# 22:45 aaronpk i believe the svg on the wiki is the closest thing we have to the source
KartikPrabhu, mlncn, tantek and cyberjar09 joined the channel
# 23:44 gRegorLove FB messaging is so weird to me. The whole adding to conversation / leaving conversation stuff
# 23:47 gRegorLove I got invited to one apparently after it started, so can't see the original comments, just everyone being excited about something.
# 23:48 gRegorLove And I'm one of 60 people. It's like an annoying reply-all email thread, but I have no context.
# 23:50 gRegorLove Maybe these things work better on the apps; I've never tried them. I find the web version a clumsy interface
# 23:51 tantek as far as webchat UIs go, I've been pretty impressed
# 23:51 tantek it has the whole (...) typing indicator and everything
# 23:56 Loqi gRegorLove meant to say: I have "Chat (Off)" in the lower right, a collapsed popover
# 23:56 tantek ah ok - if you turn that on, it becomes a realtime UI to the same messages
# 23:57 gRegorLove I think I've seen the ... indicator you're referring to in messages, if we both happen to be on at the same time.
# 23:57 tantek unless they're have a very specific topic (e.g. #indiewebcamp :) )
# 23:58 tantek the FB Messenger iOS mobile app is pretty slick too. it launches very quickly (e.g. compared to Gtalk, Skype), is fairly responsive, and lets you make audio/video *calls*