[tantek][KevinMarks] I’m thinking a “rel” repo and then we can have individual files in there for the sections of the rel registry. That way consuming applications can choose the subsets they want, and as rel values either get adopted or rejected we can move them accordingly (and have issues to track them)
[tantek]Longer term we can also consider moving rel specifications there too like /rel/tag.md (Markdown is “good enough” for specs right? Or we can stick with HTML too but that can be decided later)
aaronpk[tantek]: re video poster image, there would need to be either an implied parsing rule for video tags, or some other prefix because you need two different attributes of the same tag to parse into different values
[tantek]for a multivideo post (if anyone actually does that, I mean it's possible on Instagram), we would likely need to do something like treat poster on a video element with src and class u-* similar to how we treat alt on an img element with src and class u-*
[tantek]This is how you markup a video with a poster so that you can have both URLs show up meaningfully in the parsed mf2 result: <video class="u-poster" poster="poster.jpg"><source class="u-video" src="video.mp4" /> </video>
[tantek]alt like automatic parsing would be better for many reasons, including simplifying the markup even more than that. currently to use the method I showed you have to move the video src into a nested source element in order for it to get picked up separately from the poster
[tantek]once you have more than one video, unless they *all* have poster images marked up like that, you're going to end up with poster vs video URL mismatches
[tantek]of course anyone putting 2 + 2 together here will quickly notice that there is another problem - there is no way to specify the alt text for the poster image!
[tantek]e.g. something like a posteralt attribute would be the "simple" way to do it predictably and easily explainable to web developers. similar to how there's an hreflang attribute on a/link elements that is used to provide a hint of the language of the document at the href attribute
[tantek]though if we do decide to file an issue / change the mf2 parsing spec to do automatic parsing of the poster attribute in a video tag with src like that, I would want to add an optimistic suggestion to also look for a "posteralt" attribute and parse for that too, treating it like the alt on img src
[tantek]I don't think we ever had an issue for fancier video tag parsing because the last time we discussed it we got stuck in a similar way to how we got stuck talking about img alt
@voxpelli What’s the best framework for creating mobile apps? React? Flutter?
Me and @Okkido: The POSH-PWA
You heard it here first.
(Plain Old Semantic HTML Progressive Web App)
https://www.456bereastreet.com/archive/200711/posh_plain_old_semantic_html/
http://microformats.org/wiki/posh
You agree @adactio @ThatEmil @t, right? (twitter.com/_/status/1164795829070073857)
[tantek] [KevinMarks] I’m thinking a “rel” repo and then we can have individual files in there for the sections of the rel registry. That way consuming applications can choose the subsets they want, and as rel values either get adopted or rejected we can move them accordingly (and have issues to track them)
[KevinMarks] MD can contain html, and github pages can serve raw html too
[Lewis_Cowles] [tantek] ++
KartikPrabhu vendan: not sure which test you are using, but they might not be up-to-date or accurate
aaronpk hm it's not here... https://github.com/tantek/post-type-discovery/issues/29