#dev 2022-10-03

2022-10-03 UTC
jacky and Ocean joined the channel
#
jacky what is yax
#
Loqi It looks like we don't have a page for "yax" yet. Would you like to create it? (Or just say "yax is ____", a sentence describing the term)
#
jacky hmm
#
jacky came across https://yax.com/
geoffo joined the channel
#
[tantek]4 Not clicking on it now but does it involve shaving? (A common #indieweb-dev pastime)
#
[tw2113_Slack_] "Yax is an open source project that helps people build websites. Yax is the _do-it-yourself alternative_ to commercial site builders"
#
[tantek]4 Interesting. I wonder what's the quality like of its output and if there's an opportunity there to add mf2 (like upstream in their OSS repo) where it would be useful to the folks building sites
jacky joined the channel
#
[KevinMarks] https://twitter.com/molly0xFFF/status/1576748078652162049?s=20&t=Pferf_H-LmzHHQyNwcYIGA
#
@molly0xFFF I documented and published my annotation site code for anyone who wants to reuse it: https://github.com/molly/annotate (twitter.com/_/status/1576748078652162049)
#
jacky heh, I just coming to link that to you [KevinMarks]
geoffo and jacky joined the channel
#
[tantek]4 annotation << [https://www.mollywhite.net/ Molly White]’s annotation site code: https://github.com/molly/annotate
#
Loqi ok, I added "[https://www.mollywhite.net/ Molly White]’s annotation site code: https://github.com/molly/annotate" to the "See Also" section of /annotation https://indieweb.org/wiki/index.php?diff=83677&oldid=82110
#
[tantek]4 ^ how to add < < info from a tweet without the tweet, another WWMD [schmarty]++
#
Loqi [schmarty] has 16 karma in this channel over the last year (33 in all channels)
#
[schmarty] tantek++ ty for the diligence in unwrapping!! twitterlinks--
#
Loqi twitterlinks has -2 karma in this channel over the last year (-3 in all channels)
#
Loqi tantek has 21 karma in this channel over the last year (72 in all channels)
#
[tantek]4 🙏
#
[tantek]4 the UX of that library is quite nice, and clever use of the <mark> element as well. the annotations are only clickable with JS turned on.
jacky and gRegorLove_ joined the channel
#
epoch so webmentions would kind of lose some of the appeal if the source was a data URI? :]
#
epoch I guess you'd still be able to post comments with all the markup
#
epoch it just wouldn't be hosted anywhere except however the webmention endpoint person wanted to
mro, gRegorLove_, pharalia, tetov-irc, petermolnar, barnaby, corlaez, geoffo, gxt, Seirdy, [benatwork], [hollie] and gRegorLove__ joined the channel
#
[hollie] I'd like to make a landing page for myself where I can set my browser to, that will have links to everything I use daily and all my many classes and various things I've subscribed to. This isn't super private or revealing info or anything, but it is personal, and I'm curious how findable it is? If I don't link to it from my website, how easy is it to accidentally run into? I guess implicit in this is also the question, is this a good
#
[hollie] idea? I like having stuff organized and I often go between browsers, so I dig the idea of a page with all my stuff in one place. Is that problematic in any way?
#
barnaby if the URL is sufficiently random/unguessable, and it isn’t linked from anywhere, then it should be extremely hard to find
#
barnaby if you only want to use it from one computer, you could just make a local HTML file and set its file:// URL as your homepage though
#
h4kor[m] Also add a robots txt to avoid accident indexing
#
sknebel but *don't* make the robots.txt entry specific to that url, otherwise you just published a big pointer to "here is something I dont want indexed!"
#
barnaby yeah, you could put at a location like /private/longrandomstringofcharacters_hsdbfjshrb.html, and then use robots.txt to say “don’t index anything in /private”
#
barnaby although it sounds like there are other, better methods for preventing indexing https://developers.google.com/search/docs/crawling-indexing/block-indexing
#
h4kor[m] Yes host it on /stuff/random-string and disallow/stuff/*
#
barnaby “If your web page is blocked with a robots.txt file, its URL can still appear in search results, but the search result will not have a description”
#
barnaby but really provided it’s hard to guess and not linked from anywhere, you should be safe
jacky joined the channel
#
jacky epoch: it depends imo
#
jacky like I could see it being a way to encode some json or html of content
#
jacky but consider it anon
#
jacky that's purely up to the endpoint/service to handle
#
jacky epoch: def consider it for /Webmention-brainstorming :)
#
gRegor Wouldn't be able to do webmention verification so I think it would be an easier way to spam if they supported it
mro and jacky joined the channel
#
jacky that's true
#
jacky that actually leads me to something else, the consideration of signing Webmentions (completely optional but a tandem to a vouch)
#
jacky using HTTP Signatures
[snarfed] joined the channel
#
[snarfed] we already effectively get that with TLS + serving from a given domain, right?
#
[snarfed] what would signatures add?
#
jacky I could see that helping a bit if the conventional verification couldn't be done
#
[snarfed] got an example use case?
#
jacky it could give things like data: URIs a way to be verified (through a sig)
#
jacky sending a webmention from a local device that doesn't have a Web presence
#
[snarfed] the benefit of TLS is that the whole PKI ecosystem behind it is widely accepted and adopted
#
jacky that's true and not denying or wanting to supplant that
#
[snarfed] all other protocols that add their own signatures get a bunch more complexity because you then need to PKI, which is generally ugly and painful to adopt
#
gRegor so non-web-mention? :)
#
jacky gRegor: lol essentially (or to-the-Web-mention)
#
sknebel sounds a tiny bit like you are reinventing activitypub ;)
#
jacky sknebel: only slightly heh
#
[snarfed] (ActivityPub is a good example of signatures adding complexity. the standard says, "signatures are required, but they're hard, so we won't try, yolo figure it out!" and the result has been kinda awful)
#
jacky I guess my want/idea/experiment here has another shortcoming, despite this an optional check of sorts, if someone attempted to send a salmention in relation to this, without a expected signed message from the sender, it'd fail what sort of validation this provides
#
jacky handwaves the whole thing
#
[snarfed] lol
#
[snarfed] as usual a specific concrete use case could be a good place to start
[fluffy] joined the channel
#
jacky yeah the most immediate thing would be allowing non-HTTP resources to interact with the IndieWeb
#
jacky but that's more simply resolved by bridging them (though creating a potential bottleneck - depends on one's model)
#
jacky a bridge would be easier for people
#
[snarfed] understood! and that's also still pretty abstract, not an end user use case
#
[snarfed] right
#
[snarfed] source and target URLs that are fetchable over HTTP are pretty core to webmention
#
jacky that is, yeah - I think _specifically_ I'm thinking about things like /SSB, there's a growing hybrid Web presence of the community and I'm wondering if some of those can be retrofitted to interact with the conventional Web
#
[snarfed] ah ok! got it
#
jacky yeah, lol, I was being a bit vague because the ssb-side is _still_ early so I was working with a primitive that I understood here
#
[snarfed] wm should be usable on local networks, local hostnames and HTTP can still work, but SSB and alternative transport protocols are a bigger jump
#
jacky yeah, I think bridging (or using a relay? I think that's the term AP uses too) might be the most optimal solution
#
jacky especially with maybe a private webmention
[campegg] joined the channel
#
sknebel one could expand it maybe to "fetchable over a URL", but since at least the webmention delivery itself is specced as being HTTP...
#
jacky yeah which makes since for the time (and tbh, implementation burden)
#
[campegg] capjamesg, I meant to mention this earlier, but it slipped my mind; I got a weird response when I sent that webmention earlier… not sure if it’s my sender or your endpoint, though 🤷‍♂️ https://campegg.com/media/images/jamesg_webmention.png
mro and jacky joined the channel
#
capjamesg That's strange [campegg].
#
capjamesg The link headers look fine in a curl -I request.
#
sknebel you could check your logs if you got a good request?
#
[campegg] Hmmm. Must be something on my end, then. I tried sending some internally and they worked out OK. I did a ‘view source’ on your homepage, can couldn’t see any reference to your microsub endpoint, so was wondering where that was coming from
#
[campegg] *and couldn’t see
jacky joined the channel
#
jacky live example of tweet editing: https://twitter.com/i/web/status/1577013642490699776
#
@TurtlingB Edited Tweet https://twitter.com/i/web/status/1577013642490699776 (twitter.com/_/status/1577013642490699776)
jacky joined the channel
#
[campegg] [jamesg483] I think I tracked it down… is possibly something to do with this function in the package I’m using for webmentions: https://github.com/beatonma/django-wm/blob/master/mentions/tasks/outgoing/remote.py#L157
jacky joined the channel
#
@danleatherman Added webmentions to my site this weekend with @CodeFoodPixels @eleven_ty plugin. So easy! https://github.com/CodeFoodPixels/eleventy-plugin-webmentions (twitter.com/_/status/1577024152795770880)
#
[campegg] My regex isn’t the best, but it looks like it’s returning *all* the `link` values in the header, rather than just the `webmention` one. Does that seem like a reasonable read there?
#
jacky oh that's interesting
#
jacky I put it into https://regexr.com/6vavv (love this tool)
#
epoch the .*s are grabbing a lot more than you really want?
#
epoch I usually swap them out .* for something like [^ ]*
#
epoch like for the url capture group, since urls shouldn't have spaces anyway.
#
[campegg] jacky thanks! I tried it using a curl of capjamesg’s site and it seem to match correctly https://regexr.com/6vb05
#
epoch it it matching on the first URL, then everything between up until it finally finds the webmention rel
jacky joined the channel
#
epoch checks new regex
#
[campegg] epoch, that’s what I thought, but maybe not
#
epoch try this as a test: <https://websub.io>; rel="websub", <https://webmention.io>; rel="webmention", <https://websub.io>; rel="websub"
#
jacky wonders if such a regex is on the wiki for people to use
#
epoch it was matching from the first url, until the rel in the middle
#
epoch maybe instead of .* do [^>] since > is where you want the group to end anyway
#
epoch https://regexr.com/6vb0h
#
epoch hrm. the . before the rel in the regex isn't needed maybe.
#
[campegg] Thanks!
#
[campegg] jacky++
#
Loqi jacky has 29 karma in this channel over the last year (68 in all channels)
#
[campegg] epoch++
#
Loqi epoch has 1 karma over the last year
jacky joined the channel
#
capjamesg [campegg] indieweb-utils has a webmention sending implementation :) https://github.com/capjamesg/indieweb-utils/blob/main/src/indieweb_utils/webmentions/send.py
#
[campegg] capjamesg 🙂
#
[snarfed] capjamesg++ awesome! also https://github.com/snarfed/bridgy/blob/main/webmention.py , https://oauth-dropins.readthedocs.io/en/latest/source/oauth_dropins.webutil.html#module-oauth_dropins.webutil.webmention
#
Loqi capjamesg has 29 karma in this channel over the last year (90 in all channels)
#
[snarfed] oops sorry, the github link should be https://github.com/snarfed/webutil/blob/main/webmention.py
#
[campegg] capjamesg++
#
Loqi capjamesg has 30 karma in this channel over the last year (91 in all channels)
#
[campegg] [snarfed]++
#
Loqi [snarfed] has 19 karma in this channel over the last year (46 in all channels)
jacky joined the channel
#
jacky has anyone tried including "expanded bios" on their website?
#
jacky something like https://cattsmall.com/about/#bios ?
#
jacky I'm thinking about doing it for my site (come into a case where I need one but I don't want to repeat this, I'd rather link to it)
#
jacky indieweb-relevancy is that I'd want it to be part of my rep h-card
#
jacky but I'm a bit stuck if I should put it in my e-note, p-summary or a mix thereof
#
[hollie] This is a long shot but does anyone use Sublime Text and the BrowserSync plug-in?
jacky joined the channel
#
[hollie] Oh heck nevermind. Looks like that plug-in is dead on GitHub. I'll have to figure something else out.
tetov-irc, barnaby and jacky joined the channel
#
gRegor I use Sublime but not familiar with that plugin
lanodan left the channel
#
gRegor jacky, interesting UX on the different sizes. I've been meaning to update my bio since it's quite old. I use a p-summary for a one-line bio, p-note for the full bio. I should update that to e-note.
gxt and pharalia joined the channel