[chrisaldrich]cal, for your case I think that most in the community here would consider the copy that @threadreaderapp posts to one's website to be the canonical version (and it would likely be marked up on most CMS platforms as such), and the copie(s) that live on Twitter and on the threadreader website to be alternate copies.
[chrisaldrich]cal, it occurs to me that if you could also add syndication URLs to your initial post (rather than as an update) so that if the site supports it, it could show the syndicated URLs for the permalinks for each of the Twitter posts as well as the permalink on threadreaderapp (this would give you some additional inbound links and help to advertise your service and this use case). see: https://www.w3.org/TR/micropub/#adding-a-syndication-url
Loqiu-syndication is a way to discoverably link from your original posts to syndicated copies on other sites like social media silos https://indieweb.org/u-syndication
[jgmac1106]I keep teetering on whether I want to go back to Kirby, the webmention plugins would never work for me, but i think that has to do with my shared host
[jgmac1106]I bought an extra one because I was going to make a course builder app, but decided to stick withKnown and hired [mapkyca] to build a plugin
sebselAnd yes, I have Eddies work open in a tab. It's a project focused more on personal learning about how to do this stuff, tho. I do not expect useful output :P
[Murray]Stab in the dark on this one, but you seem to be using SCSS on CodePen but then not converting it properly on the h-card webpage. When I view source, your SCSS variables are all still there i.e. `$one` instead of `var(--one)`
[eddie]What’s your project, [sebsel]? And yeah all my stuff is MIT (seems like the easiest open source available to someone who doesn’t know much about licenses)
[jgmac1106][Cheuk_/ Cherrie] I may try exporting webmentions to google sheets and then using something to publish them, sending/receiving still way past my skill
[eddie]As far as my projects. IndieWebKit is better at showing *how* things should be done and setting up for a path forward. Indigenous was my learning project so it’s functional so in that way it’s helpful to grab things from but it’s also a mess code wise lol
[eddie]My pleasure. If you run into any questions along the way, feel free to ask. I’m not a Swift pro but I’ve learned some ins and outs while doing Indigenous. And I try to watch every WWDC video
[jgmac1106]I think what I did for adding any url I added a drop down feature in the to select a reply, bookmark, repost, etc but I pretty much just published by using text editor not their UI
[jgmac1106][Cheuk_/ Cherrie] [LewisCowles] I may one day try to install Morris, but I am also thinking about trying to use gSheets to store webmentions using webmention.io (went to dev from main)
[KevinMarks], leg and [jeremycherfas] joined the channel
jmacI have a question about WM storage best-practice: Do folks download, store, and display author photo data, instead of merely storing the u-photo URL from the original mention?
aaronpkjmac: I do download the photo and serve it myself, since otherwise there's way too many different hostnames that the browser has to load and the page slows down, not to mention that some of them disappear later
jmacI'll just switch to storing photo data. (Begs question of whether you key it on author URL, or just separately store the photo for every mention, et cetera...)
jmacaaronpk: Actually I'm curious, do you keep a database of author photos keyed on authors' unique identifiers, to save space and such, or just naively store the image for every mention with an attached u-photo
jmac(An aside: author names are just a mess with Bridgy-fed Twitter mentions, and there's nothing to do about it, because of how some people use Twitter. Replies to October blog posts will be from people's Wacky Halloween Names forever, on my website.)
[jgmac1106]adding <a class="u-url" href="https://jgregorymcverry.com<?php echo htmlspecialchars($_SERVER['PHP_SELF'], ENT_QUOTES, "utf-8"); ?> "rel="canonical"><strong>My Notes</strong> to my notes to make one less manual change
[jgmac1106]really need to think about away to store and call dt-published....that is biggest PITA, then I think about trying to automate rel prev and next
jmacI am hoping for some extended IW productivity this spring & summer, since I am just now entering a planned break from work. (That external circumstances are encouraging me to stay indoors anyway is just an odd coincidence.)
@kevinmarks↩️ #indieweb Lewis Coles: I have managed to get publishing from git to heroku now and started to look at sending webmentions from github, circleCI ran a script to send a post (twitter.com/_/status/1239230426306285571)
[jgmac1106]added this today: <a class="u-url" href="https://jgregorymcverry.com<?php echo htmlspecialchars($_SERVER['PHP_SELF'], ENT_QUOTES, "utf-8"); ?> will try and do it and see if I can remove the extension
aaronpkanyway that's the problem with the webmention for your post, when the u-url of your entry doesn't match the URL of the webmention, that page looks like a feed because it has both an h-entry and h-card
[jgmac1106]for now I will go back to manual till it hurts until I figure out the best approach, canonical, next and prev....those are my automation goals for now
aaronpkthe problem is there's a *lot* of different combinations of things to look for when trying to parse a page when you don't know what the page is already
aaronpkthe webmention URL is the right no-php version, the url on the page is a redirect to the right URL. i don't want to have to follow that URL to see that redirect though.
sknebel"A string is a *valid URL potentially surrounded by spaces* if, after stripping leading and trailing ASCII whitespace from it, it is a valid URL string."
aaronpksknebel: that step of removing leading and trailing spaces to get the value of the href attribute is strictly in HTML right? like the HTML value of the "href" attribute would not contain surrounding spaces?
[snarfed]!tell jmac re "author names are just a mess with Bridgy-fed Twitter mentions" due to people changing their names for holidays etc...true! one workaround if you're interested is, instead of rendering p-name, render p-nickname, which is their username. requires special casing bridgy twitter though. :(
jmac[snarfed]: I am not sure I was aware of p-nickname, thank you! I am not overly concerned about it right now, it was just an interesting side-effect to discover only after building a longer history of bridgy-based webmentions.
Loqijmac: [snarfed] left you a message 7 minutes ago: re "author names are just a mess with Bridgy-fed Twitter mentions" due to people changing their names for holidays etc...true! one workaround if you're interested is, instead of rendering p-name, render p-nickname, which is their username. requires special casing bridgy twitter though. :(
jmacIt reminds me of the newspaper that has to run a correction that a certain tweet mentioned in a printed story was written by Jane Smith and not Jane ScaryBOO!!ghosts, as originally reported
@BryceWrayTX↩️ Had to keep my posts list pages in the .njk templating — just couldn’t figure out the right way to translate the .njk to .11ty.js, even with help. Went back to @talkyard_io because couldn’t translate Max Böck’s webmentions code from .njk, either. Baby steps. :-) (twitter.com/_/status/1239264752838348803)
lowaragoHey, I have a website that I got from GoDaddy web hosting and I want to make a very simple login system. Now, I figured that my only option is to code it in PHP for better security. When I searched how it could be implemented with JavaScript, the results all pointed to server-side coding, like nodeJS. I think that I cannot run a nodejs on my web hosting. So, is PHP my only option?
[fluffy]at some point I should also look into doing paginated queries instead of just a one-and-done “oops, there were more than 30 webmentions, OH WELL” thing
[fluffy]either with a “show more…” button if there’s more, or just automatically fetching them all. But the amount of load that’d put on things would be worse than just increasing the query limit, I imagine.
IWSlackGateway, [aaronpk] and [fluffy] joined the channel
jmacGood evening friends. This is me thinking out loud about a utility for the manual review and moderation of one's webmention inbox: https://jmac.org/misc/wm-mockup.html
jmacYes that's me cheating a bit by which value is pre-selected. But note that there's a dropdown there, which lets you select the base domain, or domain + path parts
aaronpkother general problems with full moderation like this is that it quickly gets overwhelming maintaining the list and people will often just giveu p
jmacI want to ask "would it, though?" An unusual feature of IndieWeb is that the domain is the basic unit of currency, as opposed to account@domain. I may very well be naive but I think that making it normal to block/allow whole domains or top-level groups) at a time might make moderation easier than it seems
jmacWould you feel more forgiving if this tool had a configurable global default action: either [Hold all unlisted posts for moderation] or [Accept all unlisted posts]?
aaronpklike the rest of of indieweb, there's a place for a wide variety of tools that work quite differently, and everybody doesn't have to make the same decisions. If you think this is useful to you, then that's great! there will likely be others who agree, and also others who wouldn't use it for various reasons, and that's fine!
[snarfed]jmac: definitely look at prior art! eg most existing blog engines/CMS have sophisticated moderation queues and tools. allow/block by domain is a bit less common but there are a lot of existing parallels in those tools with mature UXes, etc.
jmacYeah, domain-based allow/deny is the specific-to-webmention feature you generally won't find on other moderation queues, and I am interested in exploring whether that grants extra leverage somehow
jmacI am politely skeptical of Vouch and not 100% allergic to the idea of manual moderation, but I grok that "this will become unusably unscalable if it actually gets popular" is a bad path