#dev 2020-02-12

2020-02-12 UTC
#
[tantek] tempted to file a Bridgy Publish POSSE note to Google Maps as an issue for https://xkcd.com/2260/
#
Loqi [XKCD] Reaction Maps https://imgs.xkcd.com/comics/reaction_maps_2x.png
#
[snarfed] so far i'm not seeing a combined comment + close action in the github API in https://developer.github.com/v3/issues/comments/ , https://developer.github.com/v3/issues/events/ , or https://developer.github.com/v4/mutation/
KartikPrabhu, [chrisaldrich] and chimo joined the channel
#
[tantek] sivy, stripping HTML does not mean stripping link text
#
[tantek] I think that's the mistake you're making
#
[tantek] your plain text note should "just" include the URLs as part of the plain text
#
[tantek] that's how the rest of us do it
#
[tantek] pretty sure this is explained in /Twitter#POSSE
[jgmac1106] joined the channel
#
[jgmac1106] sivy when I use Bridgy the link is still there, but it doesn't look like a link on Twitter
#
sivy yeah, i was thinking about grabing the link urls and making them footnotes or something
#
[jgmac1106] Known is great where you just drop in a url and it converts to link on POSSE no code
#
[tantek] no that's more overthinking
#
[tantek] if you typed links into a note, then just keep them inline like that
#
[tantek] don't do extra work
#
sivy ah..
#
[tantek] no need for footnoes
#
[snarfed] ccccccfnhietgjidergbrlejgtvdhuihlgukrvhbbnhj
#
[tantek] lolwat
#
[jgmac1106] hee hee
#
[snarfed] lol, disregard. gotta love yubikeys
#
sivy in this case it was markown, converted to an <a href>link text</a>
#
[snarfed] (don't worry, not sensitive)
#
[tantek] markdown for notes also sounds overkill but 🤷‍♂️
#
sivy this Note post: http://monkinetic.blog/2020/02/11/#txt-b163e62 had an inline link as well as a pasted link
#
Loqi It seems that there is no good thing that our white capitalist society cannot turn to evil - take for example the independent contractor #work #capitalism https://twitter.com/thotscholar/status/1227266798413897728?s=20
#
sivy i think that's just a behavior i need to avoid
#
sivy save the inline links for ARticles
#
[jgmac1106] disagree, thinkk about it logically, would you write yourself an article to remember one link?
#
sivy no, but i'd write an article if I wante to link to multiple things
#
sivy which is what that Note did
#
sivy (inline link to wikipedia about "independent contractor", and a link to a tweet)
#
[jgmac1106] Yes, I could see that note being a blog post, it has multiple idea units, images, etc
#
sivy right
#
[jgmac1106] the way micro.blog does it as soon as you hit 280 characters you get a title field
#
sivy but lately i've been doing more "small text <url>" Notes which fit netter
#
sivy [jgmac1106] that's... really clever
#
sivy but would require me to not be shit at javascript :)
#
[jgmac1106] The way Known handles it you can have as many characters but it is way drastic in stripping html, only plain text and hashtags and links get converted automatically
#
[jgmac1106] Many of the micropub clients expect you to be able to use html in your notes...I think...
#
sivy i'm a markdown fanatic myself, all my posts are stored as markdown
#
sivy with jekyll-style header bits
#
sivy all my complexity, and still notes are just posts with no title :)
#
sivy if post.Title == "" { // NOTE!
#
sivy lol
#
[tantek] heh
#
[tantek] that part is fine
#
[tantek] perhaps for your HTML stripper, when converting notes, leave the link but drop the linktext
#
[tantek] or pick a plain text way of expressing both that's not as syntaxy as markdown
#
[jgmac1106] sivy here is a note I published on Bridgy, https://twitter.com/jgmac1106/status/1227409276232179713 in the note the link is there but not in bridgy. if I want the link to work I do it like <a href="link.com">link.com</a> Twitter turns the link into working link
#
@jgmac1106 Excited to demo my Poem Pageat IndieWebCamp 2020 (https://jgregorymcverry.com/notes/2020-02-11-2) (twitter.com/_/status/1227409276232179713)
#
sivy [tantek] yeah that's what I'm trying to figure out
#
sivy right now i literally to strip_html(to_markdown(content))
#
sivy i might render it, then find the links and do something VERY loosely structured, depending on length
#
[tantek] I'd suggest considering not using markdown links when authoring notes
#
[tantek] like literally see if you can type plain text like you would a tweet
#
[tantek] instead of typing markdown
#
sivy that is what I keep coming back to
#
[tantek] and the make that work
#
[tantek] markdown is extra work
#
[tantek] or rather, *entering* markdown is extra work, and it's screwing things up for you
#
[jgmac1106] and add blankspaces...as someone who writes a lot of poetry I want note entry to not remove blank spaces and respect line breaks not stick p elements everywhere
#
[tantek] sivy, yes you should preserve blank space in your plain text authoring
#
[tantek] that reminds me I need to move that section
#
[tantek] sivy, read this: https://indieweb.org/note#Whitespace
#
[tantek] Now that I've updated the Bridgy Docs lets see if I can check the Bridgy source to see if it has an allow list for specific reacji that needs to be expanded!
#
[tantek] er, filed a PR to update 🙂
#
[tantek] a-ha it IS an allow list
#
[tantek] oooh time to see if I can improve Bridgy Publish itself!
#
[tantek] and by that I mean granary in this case 😂
#
[tantek] woohoo! first code contribution to Bridgy Publish! https://github.com/snarfed/granary/pull/186 (hoping my copy/paste coding passes all the tests 😄 )
#
Loqi [tantek] #186 add 🚀👀 reacji
#
Loqi 😃
#
[jgmac1106] tantek++
#
Loqi tantek has 24 karma in this channel over the last year (123 in all channels)
sivy joined the channel
#
sivy bah, my old usb apple keyboard jsut died.
#
sivy b0000
[Jeff_Hawkins] joined the channel
#
@swyx ✍️Clientside Webmentions: Joining the IndieWeb with @Sveltejs https://www.swyx.io/writing/clientside-webmentions Finally got around to an initial implementation of @aaronpk's wonderful http://Webmentions.io service. I found @mxstbr's clientside implementation a great balance of build risk and UX! (twitter.com/_/status/1227424568270368768)
KartikPrabhu and [xavierroy] joined the channel
#
GWG Why did rel-urls get added to the Microformats parsers at some point?
nickodd and sivy joined the channel
#
@dance2die Webmentions sounds like a shiny new thing I will work on over the long weekend :) Thankfully Gatsby has a plugin, and the instruction can be found here - https://www.knutmelvaer.no/blog/2019/06/getting-started-with-webmentions-in-gatsby/ (twitter.com/_/status/1227454459015503877)
sivy and [Michael_Beckwit joined the channel; nickodd left the channel
#
aaronpk I think it was to get access to other attributes on the links
swentel joined the channel
#
geoffo am i crazy/stupid to want to make my own indieweb compatible blog rather than just downloading WP with some plugins or using a service?
#
swentel hehe
#
swentel tough question :)
#
swentel geoffo, I you like to experiment, then it's fun to setup all building blocks yourself
#
swentel you learn a lot from it during that process
#
swentel *if
#
swentel but if you're lazy, then just install all stuff :)
#
geoffo i do like to experiment :) i've been dong web dev for about 20 years and i enjoy a challenge
#
swentel there's no crazy/stupid part here I think
#
swentel hehe, same here
#
swentel I wrote the drupal module
#
swentel and it helped learning all the concepts
#
geoffo very nice. i'm having a tricky time figuring out what i should implement first. i think web sign in, webmentions, and microformats first...
#
geoffo i also think i need to do a bit more reading :)
#
swentel I remember I started with webmentions and microformatsfirst
#
swentel I used indieauth/indielogin for web sign in the beginning
#
geoffo sounds like i'm at least on the right track
#
swentel because then you can start receiving/sending and other indieweb people can parse your content
#
geoffo my thoughts exactly
#
swentel and also setup backfeed from bridgy for instance, especially if you use twitter
#
swentel that first ping from that service is a real magical moment :)
#
geoffo yeah... i do. and that's certainly on the list, but seems like a _lot_ more work than those other things :)
#
geoffo not sure about bridgy, but i know from many many many experiences, dealing with twitter/facebook/whatever APIs sucks a big fat one
#
swentel hehe, yeah
#
swentel bridgy does a lot of heavy lifting for you on that part
#
geoffo i'm looking forward to playing with it for sure
#
swentel and snarfed is super responsive for helping out in case you're stuck on something
#
geoffo good to know. i'm really quite excited about getting some of these features implemented. i don't wanna make another stupid silo :)
#
geoffo thanks for your help! i'll definitely be lurking around and asking a lot of dumb questions. ice cream time!!
#
geoffo night
#
swentel good night :)
#
Loqi goodnight!
[tantek] and [LewisCowles] joined the channel
#
Ruxton facebook/instagram API's are pretty much non-existent now
#
Ruxton getting stuff out of them is nightmare
#
wink using them has always been a nightmare. unannounced breaking changes every month.
#
wink but at least you could so something
#
[LewisCowles] as long as they display in a browser there will be ways to get data out of instagram / facebook. They just might be less unattended friendly
#
Ruxton [LewisCowles]: instagram that works, facebook nope, need t be logged n a lot of the time
#
[LewisCowles] being logged in doesn’t mean it’s not possible, it means you need to maintain a presence and cannot scrape others data. It’s probably the most GDPR / CCPA friendly implementation, which I hope now people will see wasn’t about freeing your data, but walling it up now they have it
#
[LewisCowles] you can request an archive if you have access to the account
#
wink it's not an api then
#
wink nobody said impossible
gxt and [jeremycherfas] joined the channel
#
[jeremycherfas] [tantek] You’re right, I too can no longer see the images there. The originals are on Instagram or Twitter, and I still see them there. I also have the originals locally. I imagine the Twitter ones where backfed by Bridgy, and the IG ones by ownyourgram, but I am not sure what Known has done to make them unavailable. [benatwork]
#
[jeremycherfas] First photo posted to this instance of Known was apparently 2017-02-10 (I missed that anniversary) and, ironically, is still available at the silo. https://www.instagram.com/p/BQWLtTgg16X/ although not at the site I own.
#
[jeremycherfas] Image not visible at https://stream.jeremycherfas.net/2017/untitled
#
[jeremycherfas] Looking at the source, I can see that `img="https://stream.jeremycherfas.net/file/e7b28d89fd5bb416defcaf5d63816f39/igSXaDM6.jpg"` but I do not know how to see whether that is in the Known database.
#
[LewisCowles] wink, the benefit of an API is merely consistency. If the API has last minute changes or closes down functionality; that is unfortunate, but not a deal-breaker. I was always very specific that I wasn’t talking about their API, but countering the need for it to be accessible, by pointing out that there is a vector to get information from.
[jgmac1106] joined the channel
#
[jgmac1106] !tell tantek anyone using WP and post kinds publishes a single photo stream and not a grid: https://david.shanske.com/kind/photo/
#
Loqi Ok, I'll tell them that when I see them next
gxt, sscarfe and sivy joined the channel
#
[jgmac1106] PHP question, if I use this <a class="u-url" href="https://jgregorymcverry.com<?php echo htmlspecialchars($_SERVER['PHP_SELF'], ENT_QUOTES, "utf-8"); ?> to get the url of a current note is there a way to n +/-1 The url ends in YYYY-MM-DD-note# but that is all one string? Can I do math?
#
[jgmac1106] I went back to manual but want to figure out how to automate u-url, rel=prev, and rel=next
[grantcodes] joined the channel
#
[grantcodes] You'll need to parse the date. If you could just +/-1 you'd end up on the 32nd or 0st day anyway
#
GWG [jgmac1106]: I am willing to add a grid...if someone sends a PR
#
GWG Anyone know anything about reading Exif 2.31 in PHP?
#
GWG Apparently that's when timezone offset data came in
#
jeremycherfas I think petermolnar is your best bet for that.
sscarfe joined the channel
#
petermolnar I was summoned
#
petermolnar GWG nope, tz in exif is new to me as well
#
petermolnar but I'm using exiftool, the php exif readers are sad and limited
#
petermolnar (so are the python ones)
#
beko[m] 🤣 👍️
TGiske joined the channel
#
jeremycherfas Not an entirely wasted summons, then.
[LewisCowles] joined the channel
#
[LewisCowles] [jgmac1106] prev next can be done using the current primary key (or anything that always increases in value such as a timestamp + primary/unique key) and either prev or next.
#
[LewisCowles] `SELECT ${fields} FROM table WHERE id > ? ORDER BY ${order_field} ASC LIMIT 1` // next
#
[LewisCowles] You could then have one area to get one post order ASC or DEC to get before or after a post.
#
[LewisCowles] `SELECT ${fields} FROM table WHERE id > ? ORDER BY ${order_field} DESC LIMIT 1` // prev
#
[LewisCowles] The example given does predicate your id is travelling upwards for new records. So I’m quite sure UUID4 are not suitable
tsrt^ joined the channel
#
[jgmac1106] gwg, misunderstand, tantek was collecting examples of single stream, though I am sure many in community would like a grid
#
[jgmac1106] this is my photo grid: https://github.com/jgmac1106/homepage/blob/master/css/gallery2.css
#
[jgmac1106] [LewisCowles] tables and rows seem to indicate a database. I don't have one, just php files
[tantek] joined the channel
#
[tantek] [jgmac1106] great! Can you add that to /photo_stream?
#
Loqi [tantek]: [jgmac1106] left you a message 4 hours, 2 minutes ago: anyone using WP and post kinds publishes a single photo stream and not a grid: https://david.shanske.com/kind/photo/
#
[jgmac1106] will do in a bit, I checked micro.blog their photo discovery is a single stream but did not see a way to do a photo stream of my photos
#
[jgmac1106] What is photo stream?
#
Loqi photo stream is (AKA photo feed) a page showing a series of photo posts, sometimes presented as a photo grid, typically for the purpose of showing and catching up with that photos someone has taken and posted recently, without their other posts interspersed https://indieweb.org/photo_stream
[manton], petermolnar, [xavierroy] and [snarfed] joined the channel
#
[snarfed] [LewisCowles] wink you'll probably be interested in https://github.com/snarfed/bridgy/issues/886 . i spent a substantial amount of real world effort trying to scrape FB while logged in. the scraping itself was fine, but their anti-bot detection consistently foiled me. never really got past it.
#
Loqi [snarfed] #886 bring back Facebook by scraping m.facebook.com
#
wink thanks, but actually no. I've left FB behind and I don't care if some comments of mine are stuck in there forever. I knew it was a walled garden and never invested a lot of time there.
#
wink it's nice if people who care have a way to grab their stuff
#
[snarfed] heh i definitely didn't mean to suggest you use it
#
[snarfed] their export works fine if you just want a one time dump
#
[snarfed] my point is, after substantial effort (and experience), i wasn't able to scrape consistently (or really at all) even while logged in
KartikPrabhu joined the channel
#
[snarfed] sorry if i was unclear on the issue! i didn't mean it as asking for comment or feedback, but as a real world example/evidence of how even with logged in and scraping, you can't always consistently get programmatic access to data inside a silo
#
[snarfed] it wasn't CSS classes, it was bot fingerprint detection, ie a combination of lots of factors like user agent, IP, timing, and many others i'm probably not aware of
#
[snarfed] if i banged on it long enough, _maybe_ i could get past it - not at all sure - but then they'd inevitably catch up within days or weeks or months, and i have little to no interest in that arms race
#
[snarfed] huh, ok. sorry if i'm coming off as rude or offensive, i definitely don't mean to be! definitely open to discussion. i've just spent a _lot_ of time and effort actually working on this, for both FB and IG. if i'm unhappy with anything, it's them
#
[snarfed] (hell, jackjamieson wrote an entire academic poster and presentation about my struggles over the years with FB's API and scraping. https://jackjamieson.net/smsociety19/ )
#
Loqi Social Media & Society 2019
#
aaronpk fwiw I've been having similar struggles with Instagram lately too, enough that I closed ownyourgram.com to new signups because it's just not reliable right now
#
[snarfed] :hugops:
#
aaronpk scraping seems to work at very small scales still, but their bot detection starts blocking requests very quickly, and i have too many ownyourgram users to do it reliably at all anymore
#
aaronpk relatedly, i'm hitting Swarm's actual API rate limits too thanks to ownyourswarm, and that's a totally above-board API with OAuth and everything
#
[snarfed] yeah, IG's seems less aggressive than FB's, but similar arms race
#
[snarfed] rate limits are at least transparent
#
aaronpk yeah
#
[snarfed] [LewisCowles] understood! the code is open source, and i mentioned much of what i tried, but i definitely haven't comprehensively documented all of my attempts and the results
#
aaronpk uhoh [LewisCowles] got kicked by Loqi for shouting in all caps from that SQL and now i think i'm missing some of this convo lol
#
[snarfed] sigh
#
[snarfed] i also don't personally think fighting these arms races is a very productive use of the community's time, so i'm definitely not asking or encouraging anyone else to try. but i'm happy to answer any more q's about my attempts if you still want to, and i'd love to hear the results!
#
aaronpk restarting slack bridge
#
sknebel aaronpk: can you fix LewisCowles irc bridge representation?
#
sknebel Jinx
IWSlackGateway and [LewisCowles] joined the channel
#
[LewisCowles] kicked
#
aaronpk i don't have a way to poke at individual bridge connections so ij ust have to restart the whole thing
#
[LewisCowles] Me 😞
#
aaronpk speaking of bot detection
#
aaronpk tweaks Loqi's bot detection to never kick slack users
#
sknebel Loqi doesn't like too many CAPS in messages... Although it probably shouldn't be as agressive for known users (are you on chatnames page?)
#
sknebel Or that :D
#
aaronpk yeah i thought Loqi already bypassed spam kicking for all the registered users
#
[LewisCowles] gives Loqi a brain, a heart and courage so we can travel the yellow brick road
#
Loqi laughs at the brain, a heart and courage so we can travel the yellow brick road
#
[LewisCowles] 😂
[schmarty] joined the channel
#
sivy late to the party, but i MUCH prefer photo stream to grid
[KevinMarks], TGiske, sscarfe, [snarfed], KartikPrabhu and [jgmac1106] joined the channel
#
[jgmac1106] sivy do a blog post explaining why and add it to wiki
#
GWG [jgmac1106]: I understood, I want one... I just haven't created it due talent
[chrisaldrich], gRegorLove, TGiske, NinjaTrappeur and [tantek] joined the channel
#
[tantek] !tell geoffo not crazy/stupid at all. many of us here either have written our own, or heavily customized from existing open source. See https://indieweb.org/content_management_system#Why_write_your_own for more inspiration 🙂
#
Loqi Ok, I'll tell them that when I see them next
#
geoffo thanks @tantek :) that's very encouraging
[Aaron_Klemm] and swentel joined the channel
#
[tantek] geoffo "one of us" 😄
tantek, sscarfe and tantek79 joined the channel
#
tantek !tell aaronpk,ben_thatmustbeme,mattl,rhiaro,strugee,[KevinMarks],[BenAtWork] tomorrow is the 2 year anniversary of the closing of the W3C Social Web Working Group! https://www.w3.org/wiki/Socialwg if you have any particular updates regarding SocialWG/IndieWeb specifications that you worked on or implemented, consider this a good time to post about
#
Loqi Ok, I'll tell them that when I see them next
#
tantek it! If you've implemented a SocialWG spec in the past 2-3 years https://www.w3.org/wiki/Socialwg#Specifications, please fill out its implementation report (linked from each spec)!
#
Loqi [Social Web Working Group] Sixth Social Web F2F meeting
#
tantek heh I should add some proper top level h-entry to that wiki page
KartikPrabhu, nickodd and leg joined the channel
#
GWG [tantek]: Time for a reunion?
leg joined the channel
#
[tantek] yes! at IWC Austin! 😄
#
[tantek] GWG have you made sure all your IndieWeb spec compliant plugins have updated Implementation Reports for those specs?
#
GWG [tantek]: Not recently
#
GWG [tantek]: Make sure the video works. I'll be on
#
GWG Not that I was there for the WG
sscarfe joined the channel
#
@ChrisAldrich ↩️ The real question though, since I did a cursory Twitter search not expecting you to have an account given that quote: vanity Twitter search or ping in your server logs since I don’t see a Webmention endpoint on your site? https://boffosocko.com/2020/02/09/55766839/?replytocom=279064#respond (twitter.com/_/status/1227682694290477056)
[Michael_Beckwit and [Lauro] joined the channel
#
[Lauro] I’m new here 🙂 What’s the best way to help out/participate?
KartikPrabhu and dougbeal|mb1 joined the channel
#
GWG [Lauro]: What would you like to do?
[KevinMarks] and KartikPrabhu joined the channel
#
[KevinMarks] I was catching up with stuff from the weekend, and I saw a discussion about feed ordering
#
[KevinMarks] Apple did promote some RSS extensions for ordering podcast feeds
#
[KevinMarks] I had a very annoying time with letencrypt today
#
sknebel did they ask you to update your client?
#
[KevinMarks] the client updated, then sulked because my host didn't have the version of python they wanted and gave me passive aggressive error messages
#
[KevinMarks] I had to work around it by building python 3.6.9 from source and messing around with symlinks and running in --no-bootstrap mode
#
aaronpk oof
#
Loqi aaronpk: tantek left you a message 4 hours, 8 minutes ago: tomorrow is the 2 year anniversary of the closing of the W3C Social Web Working Group! https://www.w3.org/wiki/Socialwg if you have any particular updates regarding SocialWG/IndieWeb specifications that you worked on or implemented, consider this a good time to post about
#
aaronpk wow two years
#
[KevinMarks] what is order?
#
Loqi It looks like we don't have a page for "order" yet. Would you like to create it? (Or just say "order is ____", a sentence describing the term)
#
[KevinMarks] hm
#
[KevinMarks] what is ordering?
#
Loqi It looks like we don't have a page for "ordering" yet. Would you like to create it? (Or just say "ordering is ____", a sentence describing the term)
#
[KevinMarks] what are collections?
#
Loqi A collection is a type of post that explicitly lists and/or embeds multiple other posts chosen by the author https://indieweb.org/collections
#
@JamieTanna ↩️ Ooh nice! I've been using https://github.com/PlaidWeb/webmention.js for my client-side webmentions and have found it really nice ☺ (https://www.jvt.me/mf2/2020/02/nhz9e/) (twitter.com/_/status/1227732543618920448)
crazed, nickodd, gRegorLove, [Michael_Beckwit, deathrow1 and KartikPrabhu joined the channel; nickodd left the channel
#
sivy finally implemented a tiny character-counter for my notes form
#
aaronpk nice
#
[KevinMarks] I have one at noterlive that was tuned to tweet length
#
sivy and configured the JS to only load whewn I'm logged in
#
sivy wo performance
deathrow1 joined the channel