[tantek]connectivity (even mifi / cell / tethering) on Caltrain is quite poor (like hard to even send Slack messages, lots of failures and retries) and a good reminder of why minimum page weight, minimum connections, progressive enhancement still matter
[jgmac1106]the CT shoreline...because some genius thought let's use this coastline for train tracks gets spotty but yeah DC>Boston good coverage..does get overwhelmed in peak travel hours
@t#Caltrain commute #SF-MV-SF still had bad connectivity even on #mifi #hotspot cell #tethering. Good reminder of #webdev needs for min page sizes & requests, progressive enhancement that works if/when CSS/JS doesn't load. Not #js;dr QT [@t] js;dr = JavaScript required; Didn’t Read. Pages that are empty without JS: dead to history (archive-org), ... http://tantek.com/t4a41 (twitter.com/_/status/1090791596348588032)
eli_oat, nloadholtes_, oodani, KartikPrabhu, [tantek], [asuh], cweiske, snarfed, leg and swentel joined the channel
LoqiZegnat: petermolnar left you a message 2 minutes ago: I read my name, I can't bother to decrypt the RDF conversation, what do I need to be involved in?
ZegnatMight give it a start today. Some of the ideas this LD stuff is based on are just a bit far removed from the IndieWeb I am used to that it didn’t all make sense
ZegnatReading rhiaro’s PhD thesis actually helped with the PersonalProfileDocument thing. It describes a spectrum of Representation for profiles: http://dr.amy.gy/chapter3#profiles-taxonomy. Once I pictured the spectrum it was easy to see the specific choices I made on my IndieWeb site versus the choices WebID made.
jgmac1106[kevinmakrs] this is what Medium said to write.as, We recently experienced an interruption with API, and the ability to generate new oAuth-based applications has been restricted. I have reenabled that feature.
jgmac1106but nobody knows if ability to generate new oAuth-based applications has been restricted is still in effect, noone can make an app, and there is no more WordPress plugin or IFTT recipe
[jgmac1106]no worries on my part, just a weird way for them to handle the issue....so many devs asking or complaining and no real updates...time to redo the house odds on the site-death pools
[schmarty]petermolnar: if you have "base" defined in your pages, readers should be able to figure out the relative URLs? i forget how that works until i break something.
[jgmac1106]why is this a p-author and not a u-author is it because the author's name is in plain text? <p>Published by <a class="p-author h-card" href="http://example.com">W. Developer</a>
[schmarty]iirc, whenever there's a nested mf2 element (like an h-card) there is also a "value". what's in "value" is determined by the prefix you used. so p-* would make it plain text, u-* would be a url, etc.
[schmarty]it's a hint to whatever is consuming it. consumers that understand the nested h-card will use that, so you're helping consumers that don't expect a nested element.
[schmarty]w.r.t. u-url for each badge entry, if this page is the permalink for all of these badges, you can put an id="some-unique-id" on each <div class="h-entry">, then use that for the u-url for the badge.
[schmarty]... and your markup around the issue date would change to something like <a class="u-url" href="http://edu407.jgregorymcverry.com/badges/helloworld.html#some-student-domain.example.com"> ...
[schmarty]actually just <a class="u-url" href="#some-student-domain.example.com"> ... might be enough, since a consumer should understand that you're referring to the current page.
[schmarty]if you remove the "http://edu407.jgregorymcverry.com/badges/" from the link, it means "look at the current page and parse it for an element with id 'Emily'", which i think is what you want.
[jgmac1106]I think we did...I think this is how I we issue bulk badges...granted now need to think about having students building their portfolio for displaying but I probably won't do that entry level
@brianjessei guess it's 11 years now that i've been micro-blogging on this one web site that doesn't connect to other micro-blog web sites for some reason! #MyTwitterAnniversary #indieweb #openmicroblogging #ostatus #pubsubhubbub #webmention (twitter.com/_/status/1091025304703778816)
[grantcodes][schmarty] on the meta micropub client sounds kinda similar to how some "personal dashboards" are set up, there's one I use that I think is called organizr, basically just a page with a bunch of tabs and iframes
[schmarty][grantcodes] i think i should start a wiki page under my user page about it. with an intermediate micropub endpoint you can munge requests in ways that a given client might not support.
[schmarty]e.g. automatically adding tags/categories, processing content for links or people tags. or even wrapping a micropub client that doesn't support edits!
[jgmac1106]yeah, these won't matter really I wanted a test template, once the washer machine stops I am going to do a little video about the criteria and evidence mismatch in the badge and the task and criteria mismatch here: http://edu407.jgregorymcverry.com/moduleone.html, will then show them how I edit the page and badge in the video
@pauljacobsonIt’s such a relief when I manage to send a webmention from my blog as a reply to a comment. This stuff is probably so much easier to do than I think it is, and I’m missing something pretty obvious. (twitter.com/_/status/1091039763908431872)
[eddie]hmm I can't find an article on aaronpk's website talking about his nickname cache. I'm pretty sure he completed it, must just not have blogged about it
[eddie]But jacky essentially, you use the person's URL on your website then when your website is syndicating, it should check to see if your nickname cache contains a twitter handle for the person and if so use https://brid.gy/about#silo-content to use the correct handle on twitter but nowhere else
@pauljacobson↩️ I'm using Independent Publisher 2, along with a variety of IndieWeb plugins such as MF2 Feed, Webmention (incl for comments), Post Kinds (I lost the custom fields at some point). Webmentions generally work, although I forget to @ mention people in replies. (twitter.com/_/status/1091047559718752256)
[eddie]for example one day when we get a Bridgy Micropub input built, you could just put the bridgy-twitter-content as a property in the Micropub request
@danicotillas↩️ All right, thinking about what @jgmac1106 says about webmention that catch my attention. At the beginning I couldn't see much difference thinking about RSS or federated systems but this is another story, far more interesting (twitter.com/_/status/1091049996252790786)
[cleverdevil]FWIW, my vision of how it would work is that you use a special mention format based upon domain name, and then the nickname cache as a service translates that into appropriate mentions across all known platforms for that h-card.
[jgmac1106]but [snarfed] there couldn't be one grand directory so CMS didn't have to store all the data worry about img cache as people changed photos?
[jgmac1106]I am also thinking I am in a Wᴀʀʀᴇɴ Eʟʟɪs fan club, we have an IndieWeb ring, this includes the directory with APIs that other member of the rings can query, so thus the opt-in
snarfed[cleverdevil]: most domain registrars etc have email forwarding features, i expect most of us already have that set up (or can) and can verify me@
[eddie]someone still have to write all the parts to look for a domain in a post, fetch data from gravatar (if available) and make custom post data to twitter or other syndication
aaronpki haven't caught up on this conversation, but I can say that serving cached avatars has ended up taking more storage and bandwidth than i imagined
[eddie]I think there is another HUGE missing component in this conversation in regards to the purpose of a nickname cache, which is making it easier to @mention people than having to memorize domains. To say gravatar could be used as a nickname cache I think really undersells the potential value to people. I think the IndieWeb directories already have potentially more info and utility than gravatar would have for us, it uses the info FROM people's webs
[eddie]I will say I have no idea how old my gravatar info is, i'm sure it doesn't use my domain email AND even if I log in to update my info on gravatar I promise it won't stay as up to date as my h-card on my website
[schmarty]anyway that was the reason for the excitement. there's a bunch o' data here and that existing plumbing could be used to test some UX experiments.
aaronpkbut then you can't rely on people using any sort of convention for their gravatar email address in order to use that, otherwise you're asking people to do new work
[schmarty]the win of finding this out about gravatar is that you could right now build a tool that will try to lookup an @name and get back meaningful info to use when tagging them in a post and updating that content for a POSSE version of that post on a silo where they have an account.
[eddie]that's my point, [cleverdevil] the time spent on gravatar could turn something like grantcodes' directory when it was based on indiemap into a searchable api
[grantcodes]Problem with hcards and silo urls is potential different urls / domains, eg if someone links to fb.me you have to figure out that is Facebook etc
[schmarty]but even as we "learn" people's @names, it's not always gonna stay consistent. e.g. as cleverdevil said: he missed out on his usual username for gravatar. also, snarfed is schnarfed on twitter. inconsistentcies are gonna happen.
[schmarty]i guess my point here is that building out the UX of actually using a nicknames cache in a CMS or other posting UI is the more interesting part of this
[grantcodes]Has anyone else tried the @ mentions in the postrchild extension? Not perfect by any means but, think it's the only public micropub client with mentions
[eddie]Okay, so another question. If you make a nickname cache service, how would a micropub client (which is what most of us use) search it? Do you hard-wire it in or have a user's website/micropub server define it?
[eddie]absolutely perfect shouldn't be the enemy of good. But for a prototype, imho, Gravatar seems like the worst potential plumbing considering all the other tools we have available
[cleverdevil]Basically, fetches the provided URL, uses representative-h-card.php to identify the h-card, parses it, and then does some regex to identify different identities and how to mention them.
[cleverdevil]I eventually wanted to roll this into a Known plugin that would automatically do all of this rewriting of mentions depending on syndication targets, but I sort of ran out of steam.
jackylike not even to drag it out but if they just exposed what bits could be a username in the markup; that translation layer wouldn't be needed (with the regex)