#dev 2023-05-18

2023-05-18 UTC
eitilt, bterry, geoffo, [davidmead], epoch, gRegor and [jeremycherfas] joined the channel
#
[jeremycherfas] That fontstack thing looks very interesting. Thanks [kevinmarks]
KMF joined the channel
#
bkil [tantek]: You have mistaken it for Friendica, a second generation server. That one has S2S interoperability with ActivityPub, Diaspora, Friendica, pumpio, StatusNet, Tumblr, Wordpress, email, RSS (formerly Facebook, Twitter and ~XMPP) and C2S interoperability with ActivityPub, Mastodon, Friendica, GNU Social, Twitter.
#
bkil It would be fun to see new code in identi.ca as well, though. I wouldn't mind if they even fixed up their redirection code to fix historical dead links. It's so rare to see projects fixing links that have long since gone bad - evolution usually goes the other way.
#
bkil By the way, a relevant question. Do you guys monitor your access log for 404 to find any unintentionally broken links and possibly fix it? Does your 404 page have a button that automatically files a report about it where they can fill in the source of the link? There is a stub for this: https://indieweb.org/404 Would be nice to cross link: https://indieweb.org/longevity https://indieweb.org/zombie https://indieweb.org/lost_sites
#
bkil https://indieweb.org/post_migration https://indieweb.org/301 https://indieweb.org/redirect https://indieweb.org/own_your_links https://indieweb.org/archival_copy https://indieweb.org/URL_design https://indieweb.org/site-changes https://indieweb.org/site-deaths
#
rubenwardy I monitor Google Search Console for that
#
rubenwardy my error logs contain a lot of bots trying random asp/php stuff I don't have
Xe and gRegor joined the channel
#
bkil You could clean the log a bit via fail2ban. We could also create a collaborative list of regexp filters so we could keep this monitoring activity laser focused.
#
rubenwardy Here's an interesting line in my logs
#
rubenwardy "GET / href="https://forum.minetest.net/viewtopic.php?t=22871" target="_blank">https://forum.minetest.net/viewtopic.php?t=22871 HTTP/1.1" 404 1030 "-" "Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:x.x.x) Gecko/20041107 Firefox/x.x" 0.636
#
rubenwardy Looks like a bot is trying to follow <a href="https://forum.minetest.net/viewtopic.php?t=22871"> but assumes it's a relative URL... bit broken
#
bkil Yes, that's a classic one. It would be fun to create a wiki page where we could share log #fun with each other.
#
bkil They are looking for poorly implemented dynamic web backends that use techniques such as allow_url_fopen in PHP that they could use from the outside for an anonymous proxy chain.
#
rubenwardy that is an actual absolute link on my page though
#
rubenwardy you'd assume they'd just do GET / href="https://example.com" target="_blank">https://example.com
#
bkil It's a poorly implemented crawler then. I see hack attempts much more often in my logs, though.
cambridgeport90, IWSlackGateway, holiday_medley, jjuran, bterry and [snarfed] joined the channel
#
vladimyr capjamesg: [KevinMarks]: it would be really cool if bsky.link supported generic way of resolving bluesky urls sort of like elk (mastodon) client does
#
vladimyr You can prefix any masto/pleroma link with elk.zone and it will automatically recognize is it a profile or a post url
gxt__ joined the channel
#
vladimyr For instance https://elk.zone/https://mastodon.social/@Gargron redirects to https://elk.zone/mastodon.social/@Gargron and shows profile page
#
vladimyr While https://elk.zone/https://mastodon.social/@Gargron/110385295014756787 redirects to post/discussion page https://elk.zone/mastodon.social/@Gargron/110385295014756787
[tantek] joined the channel
#
[tantek] ^
#
[tantek] literally what I'd like to see for mf2.link as well, except support resolving any bluesky or mastodon URLs without having to tell mf2.link which it is
[KevinMarks] joined the channel
#
[KevinMarks] hm. we could make bsky.link respond to http://bsky.app URLs but static and open, the way http://psky.app did
#
vladimyr [tantek]: i can take a shot with fedi/actpub urls by swapping masto specific request with generic activitiystream id resolving like masto/feedifeed does
#
[KevinMarks] though there are a bunch we need to be careful of as we don't want to fetch your settings link for example. We could make it work for posts and profiles
#
vladimyr [KevinMarks]: that sounds cool but keep in mind it kinda makes it more difficult for bookmarklet/shortcut author
#
vladimyr Because it is bit more challenging to swap only a domain part instead of just concatenating urls
#
[KevinMarks] we can support both, it's a route regex
#
vladimyr Btw what's vxtwitter? 🤔
#
vladimyr Oh it is /a thing/ https://github.com/dylanpdx/BetterTwitFix
#
Loqi [preview] [dylanpdx] BetterTwitFix: Fix Twitter video embeds in Discord (and Telegram!)
gRegor joined the channel
#
vladimyr I'll just leave this here https://github.com/ianklatzco/psky.app/blob/master/main.py#L302-L310 :D
geoffo joined the channel
#
IWDiscordRelay <c​apjamesg#4492> I love how people are implementing the same ideas I had 🙂
#
IWDiscordRelay <c​apjamesg#4492> I started my Python wrapper in March; ianklatzco started in April. I started bsky.link in early April; ianklatzco started three weeks ago.
#
IWDiscordRelay <c​apjamesg#4492> My ideas aren't particularly novel though. I need to do a better job at promoting them I think.
#
IWDiscordRelay <c​apjamesg#4492> Oh wait, psky doesn't do embeds. Nevertheless.
#
[tantek] bkil, those are lots of good suggestions in lengthy posts for chat! would you consider blogging them on your site as independent suggestions / points that we could cite?
IWDiscordRelay, mouse[d], petermolnar, [benatwork], [schmarty], [manton] and cambridgeport90 joined the channel
#
[snarfed] cambridgeport90 as far as we determined (background on https://twittercommunity.com/ ), despite the email and UI wording, Twitter suspended a huge swath of API apps (maybe the bigger or recently active ones, but not consistently) not due to any policy violations but just as part of the free => paid move. they also haven't actually disabled the rest of the API apps, almost 3w after the original 4/29 deadline, so who knows
gRegor, pharonix71, epoch, [kimberlyhirsh] and bterry joined the channel
#
gRegor FYI, Dreamhost is going to try to upgrade PHP7.4 sites to PHP8.1 on May 25. If their tests fail after update, they'll revert and notify you.
#
gRegor You can contact them if you want to opt out of the auto update
#
gRegor They also have a PHP Extended Support fee $5/month per account if you're running unsupported PHP versions (which 7.4 is currently unsupported). https://help.dreamhost.com/hc/en-us/articles/360004672631-PHP-extended-support
#
[schmarty] ooo wow
#
[tantek] yeah I'm still on 7.4 because I have to debug why is 8.1 breaking
#
gRegor thankfully that fee is per account, not per domain running old PHP.
#
[tantek] true
#
[tantek] though I think I only have two domains needing 7.4 right now
#
aaronpk I need someone to pause the calendar while I update some of my sites
[jacky] joined the channel
#
[jacky] if Known uses that blog archive export format that http://Micro.blog uses, could that eventually become an implicit standard for established IndieWeb services? 🤔
#
[jacky] what is BAR
#
Loqi blog archive format is a data format proposed by Manton Reece for the export of a blog, based on a zip file and top level HTML h-feed inside, that is supported by micro.blog https://indieweb.org/.bar
#
gRegor Was thinking the same thing
#
[tantek] yes, two "big" interoperating implementations like that, including one open source, would be enough critical mass to move BAR along in terms of standardization
[chrisbergr], jjuran and bterry joined the channel
#
capjamesg This may be a weird question but does DNS take longer to propagate depending on the TLD?
bterry joined the channel
#
vladimyr To extend on [tantek]'s point if someone is willing to broaden the support for .bar bearblog.dev and mataroa.blog seem like good and welcoming opensource candidates
#
gRegor capjamesg, I don't think I've experienced that. Probably depends more on the provider's TTL
#
capjamesg https://airportpianos.org/
#
capjamesg We're live.
#
gRegor love the niche internet :) capjamesg++
#
Loqi capjamesg has 33 karma in this channel over the last year (98 in all channels)
#
vladimyr capjamesg: love the idea! Minor suggestion, put some padding around body text cause on mobile it goes edge to edge
#
[tantek] capjamesg++
#
Loqi capjamesg has 34 karma in this channel over the last year (99 in all channels)
#
capjamesg Why would 11ty return a HTML page as plain text? https://airportpianos.org/airports/EDI
geoffo joined the channel
#
Zegnat capjamesg: it should not. Do you run Eleventy as a server or only to build the static site?
#
capjamesg I'm using Netlify to run it.
#
capjamesg I use `eleventy` as the command to build the site and then serve from the `_site` directory.
#
Zegnat If you are using it only to generate a static site, the problem is not with Eleventy but with whatever you use to server the HTML to the browser (nginx? apache?). That server does not understand that a file without extension ("EDI") is HTML and instead sends it as text/plain.
#
Zegnat Mmm, wonder if Netlify has a setting for that
#
Zegnat Can’t find anything about how Netlify decides what content type to send :(
#
Zegnat But this sounds like the problem. Did Eleventy generate an "EDI" file or an "EDI.html" file? If the first, you actually want it to generate the latter, and then you can hide the .html using Netlify’s own Pretty URLs setting: https://answers.netlify.com/t/html-files-without-a-file-name-suffix-are-served-as-content-type-text-plain/38234
#
vladimyr Or use their custom headers settings
#
capjamesg I have just made that change.
#
capjamesg I'll see what happens.
#
capjamesg Still not working :(
#
capjamesg I think I have fixed it.
#
capjamesg Now the titles aren't showing up from the title: front matter,
#
capjamesg https://airportpianos.org/pianos/ord/
#
Loqi [preview] Chicago O'Hare International Airport (ORD) Location: Chicago, United States Pianos at Chicago O'Hare International Airport A grand piano is ...
[KevinMarks]1 joined the channel