#microformats 2014-02-12

2014-02-12 UTC
#
tantek thanks bret - sometimes the questions can be a bit odd.
#
bret also, did you hear about browsers removing microdata stuff?
#
bret a few weeks ago
#
hober of course, microdata *wasn't* dropped in the whatwg spec.
#
tantek hober - as we both know, the whatwg spec has plenty of experimental / hopeful things too.
#
tantek bret - not a surprise about browsers removing microdata stuff - it's unnecessary for any real world web use cases.
#
tantek it was made up to satisfy hypothetical use-cases from RDF(a) folks
#
tantek it did however provide additional data points about desire for general models of HTML annotation
#
tantek and a canonical JSON representation of HTML annotations
#
tantek which we've re-used in microformats2
#
tantek so to that extent, microdata provided some good R&D
#
Hixie tantek: microdata is hardly "hopeful" at this point... (and it's hardly "just a Google SEO thing", there's many search engines that use it)
#
Hixie also, it wasn't made to satisfy RDFa use cases, it was made to satisfy a whole range of use cases, few of which came from RDF people since I couldn't get them to tell me any use cases.
#
tantek Hixie, it's a "Google SEO thing" because that's the motivation behind developers paying any attention at all to it. It's irrelevant to them that other search engines do anything with it.
#
Hixie well, much of HTML is "a Google SEO thing" then :-P
#
tantek and also obvious from the articles written about microdata predominantly by SEO spammers/scammers.
#
tantek nah, HTML is mostly a "make it look good in the browser" thing
#
tantek that "whole range of use cases [including] … from RDF people" is also far more than is ever practically needed
#
Hixie practically needed by whom?
#
Hixie i mean, i have zero use for microdata personally
#
Hixie (or microformats, for that matter)
#
tantek by people who actually ship things on the open web
#
tantek Hixie, yeah, you stopped shipping on the open web (hixie.ch) and now just post on G+
#
Hixie G+ isn't on the web? :-)
#
tantek so that doesn't surprise me
#
Hixie (you realise i spend literally every day writing a web page on the open web, right?)
#
tantek no G+ is not "on the web" - it's a web app that builds stuff out of JSON
#
tantek never seen G+ posts show up as search results on any other search engine
#
tantek you can't curl G+ posts - ergo they're not on the web
#
Hixie o_O
#
Hixie [Ian Hickson DRM] on bing.com brings up my G+ post
#
Hixie as the first hit
#
tantek well look at that - thanks - first I've seen :)
#
Hixie uh huh
#
tantek why not at least mirror your G+ posts to hixie.ch?
#
tantek PESOS style
#
Hixie what's the advantage, other than fragmenting my audience?
#
tantek URLs that have a better chance of working on 5+ years
#
Hixie (btw, 90% of my g+ posts are ACL'ed, which hixie.ch doesn't support, and which is the main reason i use G+)
#
tantek Buzz URLs are already busted
#
tantek yes. ACL posts are a challenge for the #indiewebcamp crowd, though we've got a few folks who have gotten it working.
#
Hixie (btw, curl does seem to work on G+ URLs, though the resulting files don't seem to render their contents due to JS errors since they're now running on the wrong domain. but hte content is indeed there.)
#
Hixie (so it's not just JSON, dunno where you got that from.)
#
tantek right - they shouldn't be dependent on JS to "render their contents"
#
tantek that's the problem
#
Hixie it works fine if you disabled JS
#
tantek also, G+ pages are slow as crap, why I don't bother to even read it anymore
#
Hixie e.g. go to http://damowmow.com/temp/drm-g+-post with JS disabled and the post renders fine (that's the curl'ed file)
#
tantek it's like a 10-100x slower version of FB with more space wasteful design
#
Hixie pats tantek on the head
#
Hixie i'm happy to give you that it's slow, since that doesn't seem to argue that it's not on the open web.
#
tantek anyway, I understand why you use it for authoring and ACL purposes. Just saying you should mirror at least your public posts to your own domain.
#
tantek yeah - the bing result is good start.
#
Hixie i see no advantage to splitting the audience like that.
#
Hixie but my original point still stands, now that we've established i do post on the web, which is that i still don't need microdata or microformats :-)
#
Hixie but that doesn't mean others don't need them
#
Hixie microdata is used quite a lot
#
tantek sure, see above about SEO spammers
#
tantek very popular in that crowd who will do anything they perceive will game google
#
tantek whereas (nearly) everyone else would rather do the least amount of work possible
#
tantek which for search / rich snippets is microformats, and for site-to-site use cases (e.g. indieweb), is microformats2
#
tantek both for publishing and consuming - less code, less work
#
tantek maybe you don't have use for them directly, but certainly the specs you write use class names (your own set) which then you post process into spec HTML
#
Hixie microdata isn't useful for SEO spammers, dunno where you're getting this from.
#
Hixie what class names do i post-process?
#
tantek Hixie - try a search for microdata SEO and see all the SEO spammer articles about it ;)
#
Hixie there's twice as many google hits for [microformats SEO] as [microdata SEO], so this may not be proving the point you want
#
tantek number of hits don't prove anything - look at the individual articles, how recent they are, number of different sources, who the sources actually are etc.
#
tantek and yes, before microdata, the SEO crowd had started to advocate microformats as well
#
Hixie "i'm going to insult microdata but not do any of the work to actually show this insult is valid" is FUD, dude.
#
Hixie there's two pages of results for [microformats seo] in the past hour, only six posts for [microdata seo] in the past hour.
#
tantek not at all - just saying citing number of results is irrelevant (such numbers have been shown to be unreliable anyway)
#
tantek and instead providing methodology to do so qualititatively
#
Hixie yes. i have performed your research and found it does not support your point.
#
tantek since when does google actually provide time ordered results of any quality? blog search is dead.
#
Hixie since many moons? it's under "search tools" on the home page
#
tantek for anything time ordered, Twitter search (a silo search, sadly) does a better job
#
Hixie (i don't have a horse in this race, just so we're clear. i don't care about either of those. i'm just saying what you're saying is bogus.)
#
Hixie twitter: microdata seo, most recent post, 8 hours. microformats seo, most recent post, 12 minutes.
#
Hixie not sure how else to get useful data out of that.
#
tantek wow - so those results are quite different than what I was seeing perhaps 6-12 months ago last time I checked
#
tantek now checking google results for "past hour"
#
Hixie gotta go, bbl
#
tantek nearly all the microdata mentions are keyword stuffing - not actual mentions
#
tantek first six google results for microformats SEO in the past hour are the same thing - duplicates
#
tantek similarly with the last four results on the first page "Stat My Web"
#
tantek and the 2 results on the second page also
#
tantek so yeah, none of the results for "microformats SEO" in the past hour from Google have anything to actually do with microformats
#
tantek and the only result about "microdata SEO" from the google in the past hour is itself a search result for training videos
#
tantek Hixie, so your analysis/assertion of "two pages of results for [microformats seo] in the past hour, only six posts for [microdata seo] in the past hour" when actually analyzed, i.e. each result, is shown to demonstrate nothing.
#
tantek this is what I mean by you have to actually look at the results, who they are from etc.
#
tantek citing search result numbers is (nearly) useless - you should know better than that.
#
tantek Hixie, if you "don't have a horse in this race", then why not drop microdata from the HTML spec? Since it's clearly not "essential" for HTML.
#
Hixie it's used on millions of pages
#
Hixie it'd be pretty dumb to drop something that's actually widely used and not harmful.
#
Hixie (just because i've no interest in structured data doesn't mean HTML doesn't need a mechanism for it)
#
Hixie (i mean, the use cases for which microdata was made are real, and nothing else is addressing them)
#
tantek microformats and microformats2 address the real world use cases just fine
#
tantek nah, you used various non-real RDF use-cases for microdata just because they were the squeaky emailers
#
tantek thus making it more complicated than it needed to be in practice
#
tantek RSS and XHTML were also used on millions of pages, and they've been effectively dropped too
#
tantek as is meta keywords
#
Hixie what RDF use cases? the RDF people literally gave me no use cases after months and months of me asking them for use cases.
#
tantek BTW how are your stats on non-schema-org uses of microdata?
#
Hixie all the use cases for microdata had nothing to do with RDF, as far as I recall
#
Hixie and microformats doesn't handle those use cases either, they were mostly about people making up proprietary vocabs for private use
#
tantek oh all those emails they sent, with "use-cases" in an attempt to justify merging RDFa into the HTML spec
#
tantek microdata was essentially a foil to keep RDFa out of the HTML spec (which is fine)
#
Hixie it really wasn't
#
tantek it's just that it's clear that's not necessary any more
#
Hixie i don't think any e-mails regarding use cases that led to microdata came from RDF people
#
Hixie i don't need a foil to keep RDF out of HTML
#
Hixie i can just not put RDF in
#
Hixie :-)
#
tantek this was back when you cared about W3C HTML :)
#
tantek I realize it's no longer true
#
Hixie no, microdata predates that
#
tantek and besides even RDFa people don't really care much about it any more, they've moved onto JSON-LD ;)
#
Hixie (XHTML is a good example, btw. I wouldn't remove XHTML from the spec either.)
#
Hixie (and XHTML is used far less than microdata.)
#
tantek XHTML doctype? or application/xhtml+xml?
#
tantek I'd say doctype used more, mimetype less
#
Hixie yeah, that's amazing to me. JSON-LD is the kind of thing i was telling them todo for years, and they kept saying i was wrong, and then finally... -_-
#
Hixie XHTML doesn't have a doctype. it's just XML.
#
Hixie i mean the HTML namespace, really.
#
tantek sorry, I should say, XHTML 1.0 ;)
#
Hixie XHTML 1.0 is obsolete.
#
tantek right, but used on millions of pages
#
Hixie for some definition of "used" that i don't agree with :-)
#
tantek heh
#
Hixie HTML = stuff sent as text/html, XHTML = stuff sent as XML with an HTML namespace
#
Hixie i'm saying i can't remove that XHTML from the spec
#
Hixie XHTML 1.0 sent as text/html = HTML, and isn't even mentioned in the HTML spec as far as i recall
#
Hixie it's just handled by the default error handling logic.
#
tantek sure you can define "XHTML" that way
#
tantek XHTML 1.0 had a spec that defined it, which also allowed for text/html.
#
tantek what's the use of xmlns? I couldn't find any in practice.
differnet joined the channel
#
Hixie tantek: like, how much is it used? it's pretty minimal.
caseorganic and tantek joined the channel
#
tantek whoa - 10 year anniversary of when KevinMarks and I introduced microformats at Etech
#
tantek edited /Template:DraftSpecification (+5) "using IRC instead of email list" (view diff)
#
tantek edited /hreview (+20) "latest h-review" (view diff)
#
@hepatuwu SchoolEmpathy, Inc. Save North Carolina hCard® Members Hundreds Time They Advanta: .eLu (twitter.com/_/status/433503818878185472)
#
tantek !spammer hepatuwu
#
Loqi Got it! There are now 169 spammers blacklisted
#
@t mentioned, since learned/made * IRC+wiki works * syndication: hentry * recommendation: hreview * markup for time: http://tantek.com/2014/042/t3/learned-made-irc-wiki-works-hentry-hreview (twitter.com/_/status/433505376060067840)
#
@iBrandan_ Never seen such a pussy nigga. Hadda pull a #Hcard (twitter.com/_/status/433508944313475072)
#
@matro RT @t: mentioned, since learned/made * IRC+wiki works * syndication: hentry * recommendation: hreview * markup for time: http://t.co/3DUzme… (twitter.com/_/status/433509136299331584)
dvirsky, eschnou, dvirsky_, waterbaby999, barnabywalters and pfefferle joined the channel
#
@SaudKarrysta Karrysta: Membuat info hcard dibawah judul postingan http://karristaent.blogspot.com/2014/02/membuat-info-hcard-dibawah-judul.html?spref=tw (twitter.com/_/status/433578705659494400)
#
@SaudKarrysta Posted by : Saud Karrysta Posted on - Wednesday, February 12, 2014 with 26 comments Salam buat sobat blogging... http://karristaent.blogspot.com/2014/02/membuat-info-hcard-dibawah-judul.html?spref=fb (twitter.com/_/status/433578468828545024)
eschnou, matthias_pfeffer, waterbaby999, TallTed, elux, netweb, caseorganic, dvirsky and tantek joined the channel
#
john5 edited /recipe-examples (+0) "fixed broken link" (view diff)
caseorganic, eschnou, hober, TallTed, caseorga_, caseorg__ and neuro` joined the channel
#
neuro` Happy microbirthday
#
tantek thanks neuro`!
#
neuro` You're welcome.
caseorganic and tantek joined the channel