#microformats 2021-03-01

2021-03-01 UTC
[schmarty], [KevinMarks] and [tantek] joined the channel
#
[tantek] I think [KevinMarks] means do you have an example URL of such directory listings generated by web servers
[KevinMarks], [jgmac1106], KartikPrabhu, jamietanna and [amy] joined the channel
#
jacky I need to get back to that
#
jacky re: opening times
#
jacky agh I feel like I've asked this before
#
jacky regarding microformat parsers and fragments
#
jacky my want for this is to mainly collapse high-frequency posts (or rather, low content posts) to be on a single URL but separated by IDs
#
sknebel jacky: asked what?
#
jacky if microformat parsers considered fragments/anchors in the parsing flow
#
jacky tbh, the more I think about it, this might be an implementation detail that's _outside_ of the scope of the spec
#
aaronpk i remember some discussion of this wrt the PHP parser. one of the tricks is if you parse only the stuff inside the fragment ID then you lose the ability to associate it with other microformats on the page like an author h-card
#
sknebel "To parse an element for class microformats: " [...] "if the element has a non-empty id attribute:
#
sknebel id: string value of element's id attribute"
#
aaronpk it's more of a consumer question than parser question tho
#
jacky but like if I gave a parser the URL "https://jacky.wtf/likes/2020/March#20200301_303", I'd want it to only parse the MF2 found at the element under that bit
#
jacky yeah that's what I'm starting to notice, aaronpk
#
sknebel as aaron says, that is a bad idea
#
sknebel the parser exposes the element id for you in the output
#
sknebel so you can look for it
#
jacky hmm that's probably best then
#
jacky I'd hide that logic in my code then
#
jacky thanks!
[Murray], [snarfed], [Ana_Rodrigues], jeremycherfas, [cleverdevil] and [tantek] joined the channel
#
[tantek] hmm, I'm pretty sure this was an actual issue raised on the mf2 parsing spec and we discussed it during last year's pop-up for mf2 issues
#
[tantek] I can't remember the issue number, but I vaguely remember the conclusions that yes we should include 'id' in the parsed result, and that extracting a specific element by 'id' was better handled in Parser APIs, that is, APIs that call parsers or that parsers themselves expose to be called, as an option to only return that "subtree"
#
[tantek] I think GWG helped push this one forward
#
GWG It depends on perspective
#
GWG trying to parse your site did that
#
[tantek] GWG, what part of what I said depends on perspective? Pretty sure everything I said there was factual
#
[tantek] I wasn't debating the use-case. Merely noted how we solved it
#
GWG I would have given you the credit for the idea
#
[tantek] or credit for being the annoying use-case 🙂
[Rose] joined the channel
#
GWG Annoying?
#
[tantek] or "exception" 😉
#
GWG [tantek]: You are one of a kind
[Jacob_Hall], [scojjac], [jgmac1106], [KevinMarks], [tw2113_Slack_], [fluffy], [snarfed], jamietanna, [Murray], [chrisaldrich], KartikPrabhu and Seirdy joined the channel