#@rem↩️ I'd not heard of backtype, but it sounds like the kind of development that happened when twitter was an open api to play with.
The site I released is "just" a helper on top of webmentions (an evolution of pingbacks), but replies from twitter are also doable via bridgy— (twitter.com/_/status/1141268017113948160)
#petermolnaranyone knows a good guide to html5 "semantic" structure? As in how header, footer, main, article, section, aside, etc should be used, with examples.
#Loqi[Jeremy Keith] How to Section Your HTML | CSS-Tricks
June 18th, 2019
A deep dive with good advice on using—and labelling—sectioning content in HTML: nav, aside, section, and article.
#aaronpk[kevinmarks786] what's it doing? that link is restricted to your account
dougbeal|mb1 and snarfed joined the channel
#jgmac1106anyone have a favorite data visualization tool? Currently all the statisical analysis “done in” ReVIEW is just done by me bespoke to each client and added to reports, statistical needs aren’’t crazy basic agreement stuff, pearson coefficient, interclass correlation, descriptives. Preferably something written for R that spits out pretty graphs.
#jgmac1106data doesn’t have to be correct if you provide customers pretty graphs
#snarfedjgmac1106: if you're using R, Shiny is the obvious baseline
#jgmac1106currently using SPSS offline, but IBM charges so much…and I am not scalable, thx for recommendation
#snarfedi also like http://www.metabase.com/ a lot. it's aimed more at BI than academic research, no deep stats or R support, but it's really good and usable
#jgmac1106both good luck and bad luck I only have enough clients to do the stastical analysis by hand, but it would be really nice to get to a place where I have too many to go at it myself
#jgmac1106amazing how much you can upcharge for math
snarfed, jackjamieson, gRegorLove__ and [kevinmarks786] joined the channel
#[kevinmarks786][aaronpk] the feed is https://svgshare.com/ - on desktop they do load, on android I just get grey boxes.
#aaronpkah that is also because that prevents monocle from loading the images :)
#aaronpkI may revert that, but right now if you have "data save" mode enabled you just don't get the images
#aaronpkmaybe makes more sense for only webmention avatars
#petermolnar[jgmac1106]: feedback: not too long ago I read somewhere that the last html spec that had an RFC was 2.0. I'm genuinely disgusted with the "living standard" thing, so not surprised these things never happen in any way. By the time the it was finished, the lovely google driven whatwg probably moved on already.
#gRegorLoveFound an edge case with my Granary-based Atom feed. Since my notes don't have any p-* properties, the implied name includes the image alt, which granary uses as the <title> (correctly)
#gRegorLoveMy templates do have a p-name, so it's explicit
#aaronpkXRay is supposed to be doing everything documented by h-feed/authorship/etc
#aaronpkso if it's doing stuff that's not yet documented, lets document it!
#gRegorLoveI'm considering declaring template zero, starting with fresh templates. I'd do several things differently if I was adding mf2 to a new site today :)
#aaronpkyep I did that with my latest site rebuild
#aaronpkThankfully I don't hate my new site yet. Longest I've gone so far without a rewrite lol
#gRegorLoveSame! I've had this design since 2011 I think? Used to only last a couple years.
#sknebelaaronpk: good to know - I guess the bits I had at the back of my head actually aren't feed related. good thing I marked it as a joke :P
#sknebel(because aaronpk is right, it is a good starting point to check our documentation. I feel like feed authorship and handling for pages without feed markup might be edge cases not written up well, but well, need to check)
BenLubar, [tantek], [grantcodes], snarfed and jgmac1106 joined the channel
#[fluffy]A thing I’d love to see webmention.io (or some other endpoint) do is have a quarantine for non-whitelisted domains or maybe a rule-based or bayesian filter to try to quarantine obviously-shitty source links
#[fluffy]like based on ngrams in the source URL or something
#sknebelheh, if they do that you could do almost the equivalent of greylisting in e-mail antispam
#[grantcodes]Apologies for getting confused between skenbel and sebsel but I know at least one of you uses graphql a fair bit. I would love if you could have a look over my new together server which bridges micropub and microsub to graphql
#aaronpkI'm not happy enough with the codebase to add features like that
#[fluffy]I’m not sure what use case for a webmention webhook wouldn’t involve storing data permanently though?
#aaronpkMy site accepts the webhook and stores that itself. So I don't actually use the webmention.io API
#aaronpkIt made the code in my site way easier tho
#snarfedoh sure, for your own site, definitely. storing wms in your own site alongside existing storage is very different from running a standalone service that stores all users' data for them, indefinitely
#Loqi[snarfed] #762 resend all past webmentions from newly blocked users
#[fluffy]yeah, and my goal in Publ is to have Publ itself be completely stateless since I want the site to be trivial to migrate between hosting endpoints or make it load-balanceable or whatever. Which is why I rely on webmention.io’s own storage and API for displaying the mentions (client-side although I might eventually add a primitive to let Publ itself do the fetch server-side)
#[fluffy]I figure that client-side display of webmention and comments and so on has a nice side-effect that it’s only visible for Actual Users and won’t generate SEO spam opportunities for others. Kinda bad for accessibility though.
#Loqikevinmarks has 14 karma in this channel over the last year (36 in all channels)
#[grantcodes]But sebsel & skebel the code is here : https://github.com/alltogethernow/server It all works well for me in the together we client but I was learning graphql as I went so I maybe didn't follow best practices as much as I should have
#Loqi[alltogethernow] server: GraphQL server backend. Mixes microsub and micropub into one easy to use package
#[grantcodes]Also microformats does not really map to graphql very well since it is so fluid, but the microsub form of jf2 works well since it is less nested and more strictly defined
NinjaTrappeur joined the channel
#[schmarty]aaronpk: are you open to collecting lists of feeds that aren't parsing in aperture? (and/or monocle preview for that matter)
#[kevinmarks786]So graphql you query via a structure you want? So if they ask for h-card should you dig out all the nested h-cards?
#[grantcodes]Yes more or less KevinMarks but you have to do it down to the individual property. So you would say you want the `name` property of the author h-card
#aaronpk[schmarty]: feel free to file issues like that on xray! that's where the real work is done
#aaronpkoh that reminds me, i should add that cool pre-filled github issue thingy to monocle which opens issues on xray
[jgmac1106], jgmac1106 and chrisaldrich joined the channel
#jgmac1106hey all, I am going to add pagination to my notes feed I am building what's a good number in terms of performance for microsub servers and such? Does this matter
[tantek] joined the channel
#[jgmac1106]fyi schmarty your alt text did not syndicate to Twitter, I think tantek figured out how to do this
#[schmarty][jgmac1106] oh, good call! i posted that to twitter manually via the twitter.com mobile site and i don't recall seeing a prompt to add the alt. 😕
#jgmac1106schmarty I just checked and I couldn’t find it either, granted I am using FF Preview right now so things maybe different
#[tantek]certainly if you believe my site - I blogged the death of XHTML before I added h-feed
#aaronpkso, the way this code worked, the HTTP headers weren't available at the point of making that decision. but i made an update a week ago or so and now they are, so now i can switch off of content-type
#jgmac1106just checked Chrome…can’t add it there…but plus one to FF Preview mobile twitter app way better than Chrome app
#Loqipagination is a UI pattern for navigation across (typically chronologically) sequential pages that show one or more posts such as permalink post pages, archives, search results, and lists of tagged posts https://indieweb.org/pagination
#[jgmac1106]not really a discussion there on when and if one should add pagination. Does is matter for readers? I publish 50-60 notes a day, calendar view?
#Loqitwitter has -1 karma in this channel over the last year (-3 in all channels)
#[jgmac1106]ohh I had one other question about "This previews your feed it does not discover them" I know the url for rss but what is the link for an mf2 feed. I thought that was just your url?
#petermolnarI'm tempted to throw my raw contents dir at new cms systems; due to being lazy one of my stiched panoramas is 40MPixel, with 20k width and it tends to take down any media library that doesn't cache properly ^^
#petermolnarwhen I tried to render my site with grav a long time ago it was a disaster
#petermolnarbut that entries.pub still looks interesting