#aaronpkMine does but I don't have an interface to attach that to a post
#[Rose]That's what my series does. Posting via MicroPub allows you to upload media as part of it (running that Shortcut), and then formulates it into the right kind of post.
[Lewis_Cowles], [grantcodes], gxt and [xavierroy] joined the channel
#[grantcodes]Yeah so my understanding is published date is supposed to be when something is `published` (or intended to be published) to your site and `created` is when it was started.
#[grantcodes]The main example is usually writing a draft one day and then publishing it another.
#[grantcodes]But I think that follows for ownyourtrakt
#[xavierroy]for context, i can check into a show on trakt when i start watching it... but i've completed watching it only when the episode is over
#Ruxtonconsider trakt as an "offline clent" it wrote the posts 1 year ago, I don't want to published them all to my blog like thye happened a year later
#Ruxton[xavierroy]: and your scrrobble time will be when it completed (checkin_time+show length)
#Ruxtonalso reading the spec, created is not mentioned anywhere
#[xavierroy]so the published time would be completed time and not the checkin time
#Ruxton[xavierroy]: yep and Trakt will treat it lke that the whle way through it's system
#[xavierroy]I checked in an episode last night. and i've got 2 entries for it now
#[grantcodes]At the moment ownyourtrakt will probably duplicate that because of the way the API works. I also never use check-ins so I don't see that in my history
#shokuninDiscord[how would your editor compare with Slate btw?
#[grantcodes]Slate is awesome. It's very different from that. Mine are more for the individual properties that are useful for micropub, and slate generates one rich content value
#[grantcodes]I think I have react-quill built into my editor components, but you could theoretically use slate instead but with slate you have to set up all the ui and html conversion yourself
#[grantcodes]My browser extension micropub client uses slate though so it's definitely possible
wngrDiscord[m], loicm and IWSlackGateway joined the channel
#petermolnar!tell aaronpk something is off: https://telegraph.p3k.io/webmention/13h0oWOqL7iKFRh4DI - target: "https://brid.gy/publish/flickr", summary: "No webmention or pingback endpoint were found at the target" - but it does have <link rel="webmention" href="https://brid.gy/publish/webmention">
#jeremycherfas!tell zegnat If the people who create the stylesheet I want to use via CDN change the content of the stylesheet, then any hash I generate is going to fail, isn't it?
#jeremycherfasAnd that will stop my site from using that stylesheet.
#jeremycherfasThat is why I thought it would have to be the source that generates the hash.
[grantcodes] joined the channel
#[grantcodes]Yes, but a lot of cdns provide specific versions that never change.
#jeremycherfasI guess I should take this up directly with Tailwind. I have seen those integrity hashes with, I think, font-awesome, which may be where I got that idea.
[Rose] joined the channel
#[Rose]From memory most CSS files offered through a CDN do this.
#[eddie]My goal was to remove my server's up/down time from my Micropub workflow
#[Rose]And probably migrate all my current images too, so if I change CMS (again) in the future then the images won't break
#[Rose]Though looking at it, I think it might be smarter to use Aaron's code, and just pay for more storage with my host.
#[grantcodes]If you're on wordpress just now there are a ton of plugins that can use s3 for the media library (which is more useful than a standalone media endpoint imo because you get image resizing and other niceness)
#[Rose]β¬3.48/month for 100GB of storage should keep me going for a while
#[Rose]And having just got my local instance of my site up and running again..!
#[Rose]> Media Cloud works by moving your images, media and other files from your WordPress server to online cloud storage such as Amazon S3, Google Cloud Storage, DigitalOcean Spaces and many others. You can then serve that media through a CDN like Amazon Cloud front, Cloudflare, Fastly and others.
#Loqiaaronpk: petermolnar left you a message 9 hours, 4 minutes ago: something is off: https://telegraph.p3k.io/webmention/13h0oWOqL7iKFRh4DI - target: "https://brid.gy/publish/flickr", summary: "No webmention or pingback endpoint were found at the target" - but it does have <link rel="webmention" href="https://brid.gy/publish/webmention">
#[Rose]Turns out using webFTP is a really bad idea, but there we go all fixed.
[jgmac1106], [grantcodes] and [Lewis_Cowles] joined the channel
#[Lewis_Cowles]When looking at pushing things out [Rose] have you looked at https://www.imgix.com/ they allow you to put media in a S3 or many other types of storage, but they distribute it. A really cool thing is you can upload 1080p images, and contextually serve smaller images, scaled down images which get cached. Your original media is preserved. It's wonderful to have nice tools
#[Lewis_Cowles]Of course budget matters though. I only found out about this service from my job
#[tantek][aaronpk] Bridgy doesn't require u-syndication on the link to Bridgy for the webmention to Publish to work. Does IndieNews require u-syndication on the link to https://news.indieweb.org/en?
#[tantek]I'm not sure that linking to the indienews home page is semantically correct use of u-syndication
#[tantek]e.g. I have code that goes through all my u-syndication links *assuming* each is a link to a permalink where there is a copy of my post, and then presents them in the UI
[davidmead] joined the channel
#aaronpkah yeah using separate publish URLs would work
#aaronpkso I think my reasoning for the u-syndication that way was as follows:
#aaronpk1) it started out _only_ with the long-url method, where that link would in fact be the URL of the syndicated copy
#aaronpk2) I then added the home page method of syndication
#aaronpk3) it returns the URL on indienews in the webmention response, so the assumption is that the software would go replace the syndication link after sending the webmention
#aaronpkI think if I were to design this from scratch I would not do it that way
gxt and [KevinMarks] joined the channel
#[tantek]1) makes sense to me for why it started that way
[benatwork], krt^^, ffl^ and dsrt^ joined the channel; gozalaDiscord[m] left the channel
#[tantek]kevinmarks, re: storing a hash of the page when we linked to it, I'd think we'd want something more forgiving, like maybe hash only the contents of the body element, with whitespace removed?
#[KevinMarks]I was thinking of one that would work for content security but you could also parse for mf2 and hash that? Or maybe the entry.value
ffl^ joined the channel
#shrysrI am having trouble parsing links from my xml feed using webmention.app locally on my vps. It works fine when i supply the actual blog post link / file . I get a type error, and earlier i was getting other errors too while using the xml file
#[KevinMarks]You should be using the html, yes. It's hard to get xml to pass as html
#[KevinMarks]Also, you want the link that the mentioned page adds to go to the html version
#shrysrbut then how can i use webmention.app to scan all the html files of my website?
#shrysrcant I do that with a single command ? because i cannot pass in a directory to the npx webmention command
#shrysri'm also struggling to figure out how to implement a receiving point on my vps.
#shrysrHOLY !@!@#!@ [KevinMarks] i got it working with xml ... i just used the default template from hugo's site and changed it so that the entire content is extracted, instead of a summary.
#[KevinMarks]Ah, right, then it will be in html. Though you may still get the wrong backlink
jjuran, vika_nezrimaya and DisWebDiscord[m] joined the channel
#shrysraaronpk: is my underrstanding correct that the development section of webmention.io's github readme contains the instructions to run webmention.io locally ?