#indiewebcamp 2013-01-25

2013-01-25 UTC
tilgovi, jaquecoustaeau, scor, mxuribe, spinnerin, mxuribe1, eschnou, morrocco_mole, Stevef, zztr, lmorchard, tantek, andreypopp, danbri_, karlcow_, karlcow, friedcell and Stevef_ joined the channel
#
Stevef_ tantek: a couple of ideas on comment markup http://www.html5accessibility.com/tests/comments.html
#
Stevef_ 1 conservative the other a bit out there
mxuribe, andreypopp, donpdonp, scor, Stevef_ and adactio joined the channel
#
singpolyma Stevef_: <<Allow ul or ol elements to contain elements other than the li element>> ??
#
Stevef_ singpolyma: I didn't say it was a good idea, just an idea :-)
#
Stevef_ the discussion on the html wg list appears to be converging on use of lists as a pattern for comments
#
singpolyma cool
#
tantek Stevef_ do any of those actually participating in the discussion have a blog themselves, with comments? If they don't, I don't really think much of their opinion of what markup pattern blog comments should have.
#
tantek Either people should self-dog-food their proposals, or they're just hand-waving, armchair theorists at best.
#
tantek regardless of it being at "W3C" or in a "WG"
#
Stevef_ tantek: I think all of them do
#
tantek (also why I tend to ignore most emails ( / lists ) and treat them as more like support forums than anything substantial)
#
tantek I hope they provided citations of real world examples then.
#
Stevef_ I provided real world examples
#
Stevef_ and input form leoinie watson (for example) is based on her real world experiences with how commants are marked up
#
Stevef_ as a user
#
tommorris wonders whether the semantics of <section> match up with that of a blog comment.
#
tommorris Or perhaps not.
#
tommorris People do refer to them as "comment sections".
#
tommorris is tempted to email Hixie asking if <nuclearwastegroundofhell> should be added to mark up comment sections.
#
tommorris It'd allow user agents the option to filter them and automatically dose users with iodine supplements to prevent radiation poisoning.
#
Stevef_ tantek: i am trying to work out how to improve the experienece for real world users who actually consume the semantics
#
Stevef_ and such semantics are burned in by browsers
#
tantek Stevef - an ordered list makes sense since they are time ordered, and a list.
#
Stevef_ yeah thats the thinking of those involved in the discussion
#
tantek it's clear that comments are worthy of their own consideration, for all the use cases documented
#
tantek barnabywalters, aaronpk, tommorris, since you guys have thought about (and hacked a bunch?) about coding comments on your own site, and are familiar with microformats2 markup, here's a straw proposal to consider
#
tantek what if we introduced an "h-comment" which is essentially the same as an "h-entry", except that it has a new "u-in-reply-to" property (which does the obvious thing that you guys are already tracking in your blog storage).
#
tantek ?
#
Stevef_ tantek: btw yes, get your comments about venue etc, when investigating stuff i try to involve people regardless of venue, and encourage particpation
#
tantek of course, like any microformats property, you could use "u-in-reply-to" on your full fledged posts that you have as well, e.g. your self-hosted @-replies to tweets which you then syndicate to twitter
#
tantek and on such posts, you could of course combine and have them be both an h-entry and an h-comment - class="h-entry h-comment" --> this is an entry and it's a comment on something else.
#
tommorris tantek: u-in-reply-to as a URL, presumably. and if it's on the same page, presumably an anchor.
#
tantek barnabywalters, aaronpk, tommorris - if the real world gut-check of that straw proposal sounds reasonable to you from a self-implementation perspective, I'll take the discussion to #microformats and we can go into details from there
#
tantek tommorris - yes, as implied by the microformats2 "u-*" prefix, the "u-in-reply-to" property would parse looking for URLs per the microformats2 parsing spec.
#
tommorris tantek: not just for comments btw, also for main posts.
#
tommorris so for instance: http://tommorris.org/posts/8054 - post I made in response to Aral.
#
tommorris the first link should probably have rel="u-in-reply-to"
#
tommorris (it now has that.) ;-)
andreypopp, tantek, brennannovak, barnabywalters and eschnou joined the channel
#
tantek tommorris - I think rel="in-reply-to" would be sufficient, but if we really want to name it properly, the rel value should be a noun naming the thing it's pointing to, relative to the thing it's pointing from
#
barnabywalters tantek: the rev attribute was deprecated, right?
#
tantek correct. too hard for people to get it right when authoring = more noise than signal = bad markup feature.
#
tantek otherwise, yes, we could theoretically say rev="reply"
#
tantek tommorris, looks like the POSSEd copy of your posts/8054 : https://twitter.com/tommorris/status/294497028766056448 didn't set the "in-reply-to" field when calling the Twitter API and thus your tweet doesn't show what tweet it is in response to (on Twitter's site).
#
tantek Just FYI as I thought you said you had this working (I think barnabywalters and aaronpk have their POSSEing of replies working with maintaining such threading on Twitter itself).
#
@tommorris @aral @StevieBuckley You have to do some source jumping, but it’s ultimately data US Bureau of Labor Statistics d... http://t.co/7BTkCBTg
#
tantek Stevef_ - also I found this: http://microformats.org/wiki/comment-brainstorming
#
barnabywalters tantek: tommorris: yep, you set the in_reply_to_status_id param when POSTing, but you have to get the URL right
#
barnabywalters I can’t remember if it’s /status/id or /statuses/id
#
tantek there's been a lot of brainstorming / thinking about comment microformats over the years, and hopefully we've done a decent job of documenting the research
#
tantek barnabywalters - when you figure that out (if it’s /status/id or /statuses/id) that would be worth documenting on the wiki for Twitter
#
tantek http://indiewebcamp.com/Twitter
#
Stevef_ tantek: thnks will look
#
barnabywalters it’s in some of the code I open sourced
#
barnabywalters yep, it’s status, singular
#
tommorris tantek: cheers, I'd been meaning to fix that
#
waterpigs.co.uk edited /Twitter (+1462) "/* How to */ Added note => tweet POSSE algorithm" (view diff)
#
waterpigs.co.uk edited /Twitter (+1) "/* = POSSE Notes out to Twitter */ Corrected title markup" (view diff)
#
tantek oh dear, time to get ready for a wiki conflict merg ;)
#
tantek.com edited /Twitter (+3237) "braindump write new section, How to POSSE to Twitter (needs some specific example citations), merge with / make new subsection from Barnaby's on POSSEing notes pseudocode. no longer a stub." (view diff)
#
tantek and merge done
#
tantek :)
#
barnabywalters wow, nice! And I thought my one went into too much detail ;)
#
tommorris I'll have a look in a sec
#
tantek has been thinking about details of POSSEing for far too long.
#
tantek (and has at least coded up most of it live on his site :) )
spinnerin_ joined the channel
#
tantek.com edited /Twitter (+609) "/* POSSE Notes to Twitter */ add specific permalinks to posts and POSSE tweets to demonstrate examples, note specific CASSIS function to use" (view diff)
#
tantek.com created /CASSIS (+706) "draft a stub" (view diff)
#
tantek.com edited /Twitter (+296) "/* POSSE Articles to Twitter */ examples of article/post names/titles followed by blog post permalinks" (view diff)
andreypopp, tantek and spinnerin_ joined the channel