#dev 2021-08-26

2021-08-26 UTC
rockorager, KartikPrabhu, hendursaga, hendursa1 and tetov-irc joined the channel; astrojuanlu[m] left the channel
#
@MsCryptomom1 @gentrymediapro https://webmentions-next.netlify.app/ Trade (twitter.com/_/status/1430871257050980352)
nertzy_ joined the channel
#
@gentrymediapro https://webmentions-next.netlify.app/ (twitter.com/_/status/1430874528125706240)
hendursaga joined the channel
#
@gentrymediapro Learning web-mentions: https://webmentions-next.netlify.app/post/demo-post-1 (twitter.com/_/status/1430276392193863680)
chenghiz_, nertzy_, [snarfed] and shoesNsocks1 joined the channel
#
[tantek] lol that's a different kind of webmentions spam
jamietanna1 joined the channel
#
[tantek] Murray[d] simple formats are hard is the short answer to "why isn't there a simple CV format"
#
[tantek] the default technical / engineering approach to formats is to over-engineer (e.g. be completionist), and to make stuff up, both of which tend towards complexity, not simplicity
#
[tantek] there was some sort of made-up XML attempt back in the day
#
[tantek] which of course required completely restructuring any existing resume content you might have
Seirdy, alex11, jeremy, wagle_, eco_, P[m]1, diegov and adi1 joined the channel
#
[tantek] aaronpk, and other IndieAuth implementers, I note that the IndieAuth spec only talks about sign-in in (non-normative) examples, and in particular only once mentions the client "transforms that to http:" in one example
#
[tantek] two things:
#
[tantek] 1 (non-normatively) reference the microformats.org Web Sign-in spec which actually does have text explaining the plain domain to URL expansion as something for developers to implement
#
[tantek] 2 is it time to change the implied scheme/protocol for web sign-in (and IndieAuth/Login in particular) explicitly to "https" ? I was thinking of "just" going ahead and doing this in the Web Sign-in spec (and then updating implementations accordingly), but figured I'd run it by this crowd first
#
[tantek] citation for that "transforms into http:" https://www.w3.org/TR/indieauth/#redirect-examples
#
[tantek] oops out of date section
#
[tantek] https://indieauth.spec.indieweb.org/#url-canonicalization is the latest
#
[tantek] interesting that that makes it actual spec text instead of an example (I missed that change) and gives a choice! "by prepending either an `http` or `https`"
#
[tantek] I'm thinking of making it a MUST to https in Web Sign-in
#
[tantek] for now I've updated Web Sign-in to say "or" (like IndieAuth does), and added "preferably" for https to start nudging in that direction
nertzy_, jeremycherfas, KartikPrabhu and tetov-irc joined the channel
#
Murray[d] tantek++ that makes sense, and interesting to know something has been attempted (even with XML). From an "inside the industry" thing, I think a lot of the smaller boards would be happy to see a format that could be exported/imported easily, but the larger ones likely don't want to push for it (don't want it to be easy to leave)
#
[tantek] yes your summary analysis sounds right to me Murray[d]. in addition if/when "larger ones" decide to push for something, they more often than not make something up which fits their internal schema, and then fiat propose it to the world instead of actually developing something openly
#
[tantek] so the only chance that a truly "open" format has is to start small, with an ecosystem of smaller players developing / iterating it (when they can make breaking changes and not worry about folks updating things). Only then when it's fairly well interoperably implemented, and growing in adoption is it worth the effort to attract the attention of the "bigger" players to support it
#
[tantek] it's too tempting for the "larger ones" to abuse their power in the industry and "just ship" something to make a de facto standard, subverting any open community efforrts
#
Murray[d] and cementing their own position as custodians of the ecosystem, no doubt
#
Murray[d] oh well 😅
#
[tantek] yup
rockorager, KartikPrabhu and nertzy_ joined the channel