#dev 2019-03-25

2019-03-25 UTC
#
jacky def in a rabbit hole just b/c I want to play with the rest of the web
#
jacky stumbled onto https://help.fivefilters.org/full-text-rss/site-patterns.html
#
jacky it's like a hammer to read content from other sites
#
jacky debating using something like this if mf2 isn't found and use it to render a mf2 friendly page
#
jacky also found https://en.wikipedia.org/wiki/Really_Simple_Discovery
#
jacky looks like a way to 'advertise' editors for a site
[jgmac1106] and KartikPrabhu joined the channel
#
@jgmac1106 I was looking for a way to do threaded webmentions on Known. Wanted, want, a plugin but think I can use the notification panel https://quickthoughts.jgregorymcverry.com (https://quickthoughts.jgregorymcverry.com/s/2aAyel) (twitter.com/_/status/1109990549992456194)
#
@jgmac1106 ↩️ But you aren't brought back to the notification panel after publishing. Threaded webmentions still will take 5-7 clicks. No workflow improvement over my current system. Back to thinking up a plugin I could never build. (https://quickthoughts.jgregorymcverry.com/s/1HDraa) (twitter.com/_/status/1109992208076279809)
KartikPrabhu joined the channel
#
[jgmac1106] Jacky can you check and see if the post have Open Graph stuff for Facebook and Twitter?
#
jacky [jgmac1106]: I could but that still doesn't give me useful authorship info
#
jacky I also don't want to rely on FB/Twitter to tell me what a page is (esp when it's a condensed form of what's in the page)
KartikPrabhu, snarfed, cweiske, barpthewire, [sebsel], iasai and [kevinmarks] joined the channel
#
[kevinmarks] there is author info in OGP/twitter cards, but only in silo form, you'd have to expand the accounts into urls
iasai joined the channel
#
jgmac1106[m] is there away to parse pdfs? Most of my bookmarks are pdfs I have to manually add the title
[sebsel], [frank], [Rose], [jgmac1106] and [kevinmarks] joined the channel
#
[kevinmarks] it's not straightforward
#
Zegnat There are a number of PDF parsers, I think, could be interesting.
#
Zegnat Then you could also accepts webmentions from PDFs. Hmmm
#
cweiske $ exiftool -json ~/Dokumente/NSAsecurityPosters_1950s-60s.pdf | jq .[].Title
#
cweiske "National Security Agency (NSA) security/motivational posters from the 1950s and 1960s"
#
cweiske most pdfs have a metadata title attribute
#
cweiske that exiftool is able to read
[keithjgrant] joined the channel
#
[keithjgrant] The problem with working fully open source now (I just started at Red Hat!): I now miss Github notifications for my own projects amid the flood
#
cweiske oh, congratulations!
#
Zegnat cweiske: nice, I didn’t know exiftool worked with PDFs!
#
cweiske exiftool works with everything :)
#
Zegnat Haha
#
[Rose] Assuming the metadata attached to the PDF is correct 😛
iasai joined the channel
#
cweiske that's why I wrote "most"
jeremych_ joined the channel
#
Jeremycherfas I guess that might be how Bookends gets metadata out of PDFs. Normally does a great job.
#
Zegnat Once upon a time I was going to look into PDF/A parsing, thankfully it was dropped
[kevinmarks] joined the channel
#
[kevinmarks] Pdf is basically a graphics format, so you either write parsers that know how certain tools write them, or you render and ocr.
iasai and [jgmac1106] joined the channel
#
[jgmac1106] But now most will have a DOI
#
[jgmac1106] Older docs don't and the database records are most accurate but usually pay walled...
#
[jgmac1106] I mean currently I just copy and paste the title it isn’t too much trouble
#
[jgmac1106] I wonder how hypothes.is does it. They are pretty accurate... Get around most garbage in, garbage out
#
[jgmac1106] Zegnat webmentions to/from pdf.... interesting
#
Zegnat Probably something like what cweiske said. Or possibly even specific scrapers. I forget the biblio tool’s name, but sknebel once linked me to a repo full of context extractors for websites without specific mf2/ogp/json-ld markup
#
[jgmac1106] Jacky see above
iasai joined the channel
#
[jgmac1106] Not for Pdfs but that is what Jacky was working on all weekend... Reliably extracting author
#
Zegnat Oh, yeah, there was this pretty big project for it. Trying to find it again... may have been part of Zotero..? I am going to spent another 10 minutes looking, else sknebel will probably just know when he turns up after work
eli_oat joined the channel
#
@jgmac1106 ↩️ well if you mark up your pages with microformats you can do all kinds of cool stuff like send webmentions, over a million to date; RSVP to events; and use new awesome social readers. Just means you keep your metadata and your HTML in… https://quickthoughts.jgregorymcverry.com/2019/03/25/alephnaught2tog-well-if-you-mark-up-your (twitter.com/_/status/1110185362796802053)
#
Zegnat !tell jacky for extracting author (and more) from all sorts of pages you may want to look at some prior art here: https://github.com/zotero/translators - basically extractors for different websites!
#
Loqi Ok, I'll tell them that when I see them next
#
Zegnat [jgmac1106] ^^^
#
Loqi [zotero] translators: Zotero Translators
#
Zegnat Pretty interesting, has specific extractors for several newspaper websites. I thought about looking into porting some of those to XRay (and thus Loqi) but didn’t feel like XRay was extendible enough
eli_oat, [kenbauer] and iasai joined the channel
#
@_am1t That's very similar to the idea behind Microsub - "it decouples managing subscriptions and delivering content from the presentation of that content." I have enjoyed using the clients that support it. https://indieweb.org/Microsub #scriptingnews http://scripting.com/2019/03/25.html#a131235 (twitter.com/_/status/1110200244820758528)
iasai, KartikPrabhu, jackjamieson, [jgmac1106], [cleverdevil] and [manton] joined the channel
#
[manton] Hi everyone! I've been adding drafts and q=source support to Micro.blog's Micropub endpoint and the native client apps. One thing I'm realizing... If you save a draft in the Mac app using a home-grown Micropub endpoint that doesn't support drafts, it will likely just get published. Is this something we should worry about, or are folks in this situation just kind of on their own to know what features their server supports?
#
[manton] I mentioned this in a GitHub thread, but I think it would be helpful to roll together several of the newer Micropub extensions into the official spec, almost like a "version 2" (except backwards compatible and not actually versioned) so that it's easier to talk about what features a client/server supports.
#
[cleverdevil] IMO, its probably fine for now, but you may want to pop up a warning for users that are publishing to custom endpoints.
#
[cleverdevil] (At least, the first time they publish).
#
[cleverdevil] Longer term, it'd be nice to be able to interrogate the Micropub endpoint to ask what features it supports.
#
[cleverdevil] After all, micropub does have q=config.
gRegorLove_ and [Rose] joined the channel
#
[Rose] I agree with [cleverdevil]. It would also be great to have a documentation page of "these endpoints support this" for people to check, so the WordPress plugin, naturally micro.blog, etc., are all listed there.
#
[cleverdevil] So, there is this -> https://indieweb.org/Micropub/Servers#Implementation_status
#
[cleverdevil] Not sure how up-to-date it is.
#
[manton] [Rose] Good point. The wiki has some information about implementations, but I think everything is a little too scattered right now... To understand Micropub, you need to read 3 things: the spec, the wiki, and GitHub. https://indieweb.org/Micropub-extensions#Query_for_Post_List
snarfed joined the channel
#
[cleverdevil] That said, I think that `q=config` should be updated to include more.
#
[Rose] I would think this should be microblog first documentation as it's features in your apps?
#
[Rose] Not saying we don't need to improve documentation elsewhere, but for your primary target audience (I suspect), that's going to be key. Lots of them don't really know about IW or understand this stuff. Of course, if they're all hosted with you or on WordPress, or savvy enough to know what works then that's ok, but documenting it on your end would be good too.
#
[manton] I'm definitely going to update my documentation. But yeah, this is a non-issue for anyone hosting their blog on Micro.blog. It's only an issue with Micropub servers that I haven't tested against (which is most).
#
[Rose] I'm sure we can help test against those ;)
#
GWG I can help with documentation
#
GWG I support some properties
#
GWG [manton]: Any chance I can make a case for webmention syndication and permalinks again?
#
[manton] @GWG Sure! Remind me what you mean by permalinks?
#
GWG I want to link from my original post on my website to the Syndicated version on micro.blog
#
GWG So I need a way to identify the copy
#
GWG And the other one was to trigger syndication to Micro.blog by something other than a feed
[tantek] joined the channel
#
[manton] Got it, thanks for the reminder.
#
GWG No, thank you
#
GWG I appreciate you
KartikPrabhu and [asteres] joined the channel
#
Zegnat Couldn’t you syndicate to micro.blog by copying the micropub request to the mb endpoint, GWG? If I do not misremember, a micropub request also as the ability to give you back the final URL (location header?) so you can then automatically link to it
#
sknebel Zegnat: I don't think blogs not hosted on mb have an endpoint there?
#
Zegnat Ooh, is that the thing. So syndication isn’t really the action of syndicating here, but the action of requesting mb to check for new content to mirror?
#
Zegnat Apparently my mind has different mental mappings for those two flows :P
iasai, jackjamieson and [manton] joined the channel
#
[manton] Yeah, "syndicating" to Micro.blog is a little different because you don't have to do anything. It just reads your feeds. But that means there's no obvious place to discover a Micro.blog permalink for your post, which also includes the conversation. I think the plan was for M.b to also send a Webmention to your blog, passing along that permalink.
#
KartikPrabhu [manton]: sounds more like a micropub "add" https://www.w3.org/TR/micropub/#add
#
KartikPrabhu lot of the use-cases which overload webmention are actually micorpub use-cases
[jgmac1106] joined the channel
#
Zegnat If the mb entry includes a link to the original, sending a webmention is totally legit though, because it does include a link
[asuh] joined the channel
#
Zegnat If the mb entry includes <a href="" rel="canonical">Original</a>, and then sends a webmention to the canonical entry, through /original-post-discovery the webmention receiver should be able to extract the canonical URL. And if that matches a post the receiver controls it could automatically update it with a u-syndication link to the mb entry.
#
Zegnat (Just a thought.)
#
Zegnat More thoughts: https://indieweb.org/original-post-discovery (not sure anyone is actively using it)
iasai joined the channel
#
[manton] [Zegnat] It does include a link to the original, but I don't think it includes rel=canonical. That's a good idea.
#
Zegnat That wouldn’t solve the triggering GWG was discussing. But it could enable mb sending a webmention once a post is live and the original source automatically listing it as a syndication copy :)
#
Zegnat Although thinking about it, while mb could support some sort of webhook, isn’t that what websub is? (A webhook for the original publisher to post to.)
#
Zegnat So I feel like it could be: WordPress publishes to feed, triggers websub, mb is subscribed through websub and gets the post, mb processes the post and publishes it with a rel-canonical, mb sends a webmention to the origin post notifying the author of the succesful syndication.
#
Zegnat Note that basically none of those steps introduce new tech. Only /original-post-discovery is newish in there for the origin to recognise the syndication.
#
Zegnat Building blocks <3
#
sknebel or add MP and the site can make an MP request and learn the URL in the response
#
Zegnat That would make sense, but I am not sure the syndication-type micro.blog feeds are supposed to be used for standalone blogging? So maybe mp posts to it do not make sense?
[tantek] joined the channel
#
[manton] Micro.blog does support WebSub as you described too, although I don't think it's used often or well-tested.
#
[manton] For feeds, the idea is absolutely to use regular, existing blog feeds. Some people use specialized feeds but that isn't required.
#
[tantek] Has anyone actually had an original post change meaning after they've replied to it? How often? Once a year? A month? Would be quite surprised if it's any more than that.
#
[tantek] (re: all the what's the use-case for signatures discussions over the weekend)
dougbeal|mb1, [jgmac1106] and [cleverdevil] joined the channel
#
GWG Zegnat, to your earlier question re micro.blog, yes
doubleloop joined the channel
#
[tantek] Yeah, didn't think so
[eddie] joined the channel
#
[eddie] Zegnat: We actually have brainstormed pretty much exactly what you talked about here: https://github.com/microdotblog/issues/issues/79
#
aaronpk I've definitely had original posts disappear, most often due to domain name lapses
#
GWG I just want to be able to easily link to micro.blog posts from their original posts and finer tune what goes there
#
[eddie] Also [manton] and [cleverdevil] it would be interesting to pair up a bunch of extensions as a "Micropub update package", but also a query for supported extensions also exists as a proposed extension here: https://github.com/indieweb/micropub-extensions/issues/7
[borismus] joined the channel
#
[manton] [eddie] Thanks, I missed that.
[asuh] joined the channel
#
[tantek] just saw this, feels devy: https://twitter.com/andybelldesign/status/1105440749687066624
#
@andybelldesign Happy 30th Birthday to The Web 🎉 https://pbs.twimg.com/media/D1dQf4uWkAAAwkD.jpg (twitter.com/_/status/1105440749687066624)
#
[eddie] tantek: Correct, I don't think anyone really felt like there was a valid use-case for the signatures discussion except the person investigating the signatures
#
[eddie] manton++ for drafts support!
#
Loqi manton has 27 karma in this channel over the last year (65 in all channels)
#
[jgmac1106] eddie....right after that conversation I met with one the artist where I am building an IndieWeb emcommerce store, we are working on a webmention "certificate of authority" he wanted to sign it....I am just going to use an img of signature..the url is the signature
#
[eddie] [jgmac1106] are you talking about a physical signature or a digital certification signature? I'm confused
iasai joined the channel
#
[tantek] [jgmac1106] related tweet to art and certificate of authority: https://twitter.com/edent/status/1006248586395508737
#
@edent I don't understand the blockchain hype. A startup has certified my artwork & placed their verification on the bitcoin blockchain. Now art dealers & auctioneers can feel secure that I am the original artist. One small problem… I am not Leonardo da Vinci! https://www.verisart.com/works/23f2c64a-08c6-4a42-8013-84ac8422dffb https://pbs.twimg.com/media/Dfbpv0lX4AAf88z.jpg (twitter.com/_/status/1006248586395508737)
#
[jgmac1106] Just a physical... I thought digital signature to be legit... But resized noone would care... Ideally would be an image of a signature carrying digital certification but that felt like overkill
#
[jgmac1106] A webmention from an artist saying you own X piece since Y date is enough IMO... Putting in the pic of signature bc customers want ir
#
[jgmac1106] That is awesome @tantek
#
sknebel heh, (non)-permanence of authors "certificates" or "approval" is a hot topic
#
sknebel s/authors/artists
#
[jgmac1106] Realized not resized... Have no idea if neurons resize when you realize
#
sknebel and given the number of dead links you encounter when reading about art that's less than 5 years old, yeah I'd want something more permanent (like paper) if there's actual value attached to it
#
[tantek] what is phparch
#
Loqi It looks like we don't have a page for "phparch" yet. Would you like to create it?_ (Or just say "phparch is ____", a sentence describing the term)
#
[tantek] what is php[architect]
#
Loqi It looks like we don't have a page for "php[architect]" yet. Would you like to create it? (Or just say "php[architect] is ____", a sentence describing the term)
#
[tantek] haha oops Slack to IRC conversion
#
[tantek] well Wikipedia calls it https://en.wikipedia.org/wiki/Php_architect
#
[tantek] what is php architect
#
Loqi It looks like we don't have a page for "php architect" yet. Would you like to create it? (Or just say "php architect is ____", a sentence describing the term)
#
[tantek] php architect is a magazine about [[PHP]] that featured the IndieWeb on the cover of its March 2019 issue as well as in an article on page 17.
#
Loqi ok
#
[tantek] phparch is /php_architect
#
Loqi ok
#
[tantek] phparch << https://twitter.com/DavidWolfpaw/status/1108916061607522305
#
Loqi ok, I added "https://twitter.com/DavidWolfpaw/status/1108916061607522305" to a brand new "See Also" section of /php_architect https://indieweb.org/wiki/index.php?diff=58566&oldid=58564
#
sknebel doesn't that more belong on /Articles_about ?
#
[tantek] phparch << https://pbs.twimg.com/media/D2OpRUHX4AMIAeu.jpg
#
Loqi ok, I added "https://pbs.twimg.com/media/D2OpRUHX4AMIAeu.jpg" to the "See Also" section of /php_architect https://indieweb.org/wiki/index.php?diff=58567&oldid=58566
#
[tantek] sknebel yes a summary belongs on /Articles_about
#
[tantek] though we have other pages on conferences and such when it has included an indieweb talk
#
[tantek] what is Personal Democracy Forum
#
Loqi Personal Democracy Forum is (often abbreviated PDF) an annual conference in New York City, NY since 2004, where Tantek Çelik gave an IndieWeb talk in 2014, and has a goal of nurturing "a world-wide conversation about technology’s impact on government, politics, media, and democratic societies" https://indieweb.org/Personal_Democracy_Forum
iasai, [Rose], swentel, chrisaldrich, MorningMoon, tbbrown and KartikPrabhu joined the channel