2023-04-11 UTC
mouse[d], IWDiscordRelay, [timothy_chambe], btrem and corlaez joined the channel
# IWDiscordRelay <capjamesg#4492> 🤦
# sknebel oof, seems indeed endless and not just stack explosion
IWSlackGateway joined the channel
# sknebel dont have time to dig into it right now though :/
IWSlackGateway and [KevinMarks] joined the channel
# sknebel of course it doesnt make sense as mf2, but you only can know that after parsing
# sknebel but even if it were an artificial example, endless loop is not the right thing to do :D
# capjamesg Has there been a h-section-fluid-container all this time that I didn't know about?
gRegor and [tantek] joined the channel
# [tantek] sounds like a good "evil" test case to add once it's been fixed
# aaronpk i'm really curious what is the minimum markup that can cause this bug
[schmarty] joined the channel
[chrisbergr] joined the channel
btrem and [jacky] joined the channel
# [jacky] this is one of those kind of scenarios where it _might_ help to have a limit on what classes to expect for `h-` objects, perhaps?
[marksuth] and ur5us joined the channel
# sknebel I *think* one could do limiting after parsing. would be interesting to experiment with what "works"
# sknebel although we dont have a good collection of real examples sadly
# aaronpk pretty sure there are some h-feeds with no properties in the wild
ur5us and [snarfed] joined the channel
# [tantek] a page which is a list or collection can be represented as an h-feed, no properties, and a bunch of h-entry children
ur5us and gRegor joined the channel
# [jacky] aaronpk: do you know if the PHP parser does DOM node transversal or is it doing some selector logic?
# [jacky] The Rust parser takes the DOM walking approach (it's harder IMO but I think it prevents this looping case)
gRegor joined the channel