/irc-logs / freenode / #microformats / 2011-08-10 / end

Options:

# Session Start: Wed Aug 10 00:00:00 2011
# Session Ident: #microformats
# [00:14] * Quits: ericduran (~ericduran@173-203-243-241.static.cloud-ips.com) (Quit: ericduran)
# [00:22] * Joins: andr3 (~andr3@bl14-221-101.dsl.telepac.pt)
# [00:44] * Quits: mwunsch (~mwunsch@64.70.114.88) (Remote host closed the connection)
# [03:16] * Joins: jrodgers (~jrodgers@CPE000d93221ae3-CM0012c99f023e.cpe.net.cable.rogers.com)
# [03:18] * Quits: @tantek (~tantek@70-36-139-219.dsl.dynamic.sonic.net) (Quit: tantek)
# [03:48] * Quits: jrodgers (~jrodgers@CPE000d93221ae3-CM0012c99f023e.cpe.net.cable.rogers.com) (Quit: jrodgers)
# [04:07] * Joins: tantek (~tantek@173-164-152-242-SFBA.hfc.comcastbusiness.net)
# [04:07] * ChanServ sets mode: +o tantek
# [04:34] * Quits: @tantek (~tantek@173-164-152-242-SFBA.hfc.comcastbusiness.net) (Ping timeout: 260 seconds)
# [04:45] * Joins: tantek (~tantek@173-164-152-242-SFBA.hfc.comcastbusiness.net)
# [04:45] * ChanServ sets mode: +o tantek
# [04:54] * Quits: @tantek (~tantek@173-164-152-242-SFBA.hfc.comcastbusiness.net) (Quit: tantek)
# [05:24] * Joins: nonge_ (~nonge@p5082A57A.dip.t-dialin.net)
# [05:28] * Quits: nonge (~nonge@p50829E1C.dip.t-dialin.net) (Ping timeout: 276 seconds)
# [09:07] * Joins: tantek (~tantek@70-36-139-219.dsl.dynamic.sonic.net)
# [09:07] * ChanServ sets mode: +o tantek
# [09:13] <Loqi> [[existing-rel-values]] http://microformats.org/wiki/index.php?title=existing-rel-values&diff=44361&oldid=44353&rcid=57505 * Tantek * (-332) removed rel-lytebox|demo value. Neither of the referenced GRDDL nor WCLR defines "demo" or "lytebox".
# [09:23] * Joins: voxpelli (~anonymous@2001:67c:2354:babd:c082:1a0e:9a90:e597)
# [10:44] * Quits: patcito (~123@190.42.68.69) (Remote host closed the connection)
# [10:45] * Quits: andr3 (~andr3@bl14-221-101.dsl.telepac.pt) (Quit: andr3)
# [10:51] * Quits: Atamido (~atamido@76-205-170-42.lightspeed.austtx.sbcglobal.net) (Read error: Connection reset by peer)
# [11:01] * Joins: csarven (~csarven@wlan-nat.fwgal01.deri.ie)
# [11:08] * Quits: nonge_ (~nonge@p5082A57A.dip.t-dialin.net) (Quit: Verlassend)
# [11:29] * Joins: andr3 (~andr3@pickles.tp.telepac.pt)
# [11:46] * Quits: Amorphous (jan@unaffiliated/amorphous) (Ping timeout: 250 seconds)
# [11:52] * Quits: manu` (~chatzilla@pool-96-240-190-251.ronkva.east.verizon.net) (Ping timeout: 255 seconds)
# [11:53] * Joins: manu` (~chatzilla@pool-96-240-190-251.ronkva.east.verizon.net)
# [12:02] * Joins: Amorphous (jan@unaffiliated/amorphous)
# [12:17] * Quits: andr3 (~andr3@pickles.tp.telepac.pt) (Quit: andr3)
# [12:27] * Joins: Atamido (~atamido@76-205-170-42.lightspeed.austtx.sbcglobal.net)
# [12:53] * Joins: andr3 (~andr3@pickles.tp.telepac.pt)
# [14:59] * Quits: ChiefRA (~RA@unaffiliated/chiefra) (Read error: Connection reset by peer)
# [14:59] * Joins: ChiefRA (~RA@86.125.108.21)
# [14:59] * Quits: ChiefRA (~RA@86.125.108.21) (Changing host)
# [14:59] * Joins: ChiefRA (~RA@unaffiliated/chiefra)
# [15:13] * Quits: andr3 (~andr3@pickles.tp.telepac.pt) (Quit: andr3)
# [15:56] * Joins: Soopaman (~soopaman@66.49.228.203)
# [16:02] * Joins: mwunsch (~mwunsch@64.70.114.88)
# [16:30] * Joins: voxpelli-laptop (~anonymous@2001:67c:2354:babd:fdee:36e1:352c:4cf2)
# [16:32] * Quits: voxpelli (~anonymous@2001:67c:2354:babd:c082:1a0e:9a90:e597) (Ping timeout: 260 seconds)
# [16:32] * voxpelli-laptop is now known as voxpelli
# [17:44] * Quits: mwunsch (~mwunsch@64.70.114.88) (Remote host closed the connection)
# [17:45] * Quits: voxpelli (~anonymous@2001:67c:2354:babd:fdee:36e1:352c:4cf2) (Quit: voxpelli)
# [17:46] * Joins: mwunsch (~mwunsch@64.70.114.88)
# [18:04] * Quits: mwunsch (~mwunsch@64.70.114.88) (Remote host closed the connection)
# [18:05] * Joins: mwunsch (~mwunsch@64.70.114.88)
# [18:19] * Quits: mwunsch (~mwunsch@64.70.114.88) (Remote host closed the connection)
# [18:20] * Joins: mwunsch (~mwunsch@64.70.114.88)
# [19:01] * Joins: kennyluck (~kennyluck@114-43-125-134.dynamic.hinet.net)
# [19:06] * Quits: csarven (~csarven@wlan-nat.fwgal01.deri.ie) (Quit: Leaving.)
# [19:14] * Quits: ivc (ivc3@elitus.x-pec.com) (Remote host closed the connection)
# [19:57] <Loqi> [[search-engines]] http://microformats.org/wiki/index.php?title=search-engines&diff=44362&oldid=44263&rcid=57506 * Tantek * (+368) move defunct search engines to new section
# [20:22] * Quits: mwunsch (~mwunsch@64.70.114.88) (Remote host closed the connection)
# [20:24] * Joins: mwunsch (~mwunsch@64.70.114.88)
# [20:46] * Quits: mwunsch (~mwunsch@64.70.114.88) (Remote host closed the connection)
# [20:47] <Loqi> [[rel-home]] http://microformats.org/wiki/index.php?title=rel-home&diff=44363&oldid=41180&rcid=57507 * Tantek * (+495) added example in wild: use rel-home as part of feed discovery from post permalink pages
# [20:48] * Joins: mwunsch (~mwunsch@64.70.114.88)
# [20:48] <@tantek> so I think I finally found a good use for rel="home"
# [20:49] <@tantek> when a post permalink page links to the feed of the *site* for auto-discovery purposes, there's something not quite right about using just rel="alternate"
# [20:49] <@tantek> that is
# [20:49] <@tantek> alternate implies that the destination URL is an alternate version of the *current* page
# [20:49] <@tantek> which it isn't in the case of a post permalink page linking to the *site* feed
# [20:49] <@tantek> thus one solution would be to combine 'alternate' with 'home'
# [20:50] <@tantek> on such links
# [20:50] <@tantek> cc: Kevinmarks
# [20:51] <@tantek> I've started doing this on my site - per suggestion to include feed auto discovery on my permalink pages.
# [20:52] * Quits: mwunsch (~mwunsch@64.70.114.88) (Remote host closed the connection)
# [20:55] * Joins: KevinMarks (~KevinMark@c-71-204-145-244.hsd1.ca.comcast.net)
# [20:56] * Joins: mwunsch (~mwunsch@64.70.114.88)
# [20:57] * Joins: dglazkov (~dglazkov@nat/google/x-ujnqpqfkvhxpxvut)
# [20:57] * ChanServ sets mode: +o dglazkov
# [20:57] * Quits: danbri (~danbri@ip176-48-210-87.adsl2.static.versatel.nl) (Remote host closed the connection)
# [20:57] <@dglazkov> tantek: got a sec?
# [20:58] <@tantek> dglazkov - got an opinion on rel="home alternate" as noted above http://krijnhoetmer.nl/irc-logs/microformats/20110810#l-61 ?
# [20:59] * Joins: patcito (~123@190.42.92.29)
# [20:59] * Quits: mwunsch (~mwunsch@64.70.114.88) (Remote host closed the connection)
# [21:00] <@dglazkov> tantek: sure, "home alternate" sounds reasonably intuitive
# [21:00] <@tantek> ok, given that actual practical use case for rel-home, I'm going to add it to the registry for HTML5
# [21:04] * Joins: mwunsch (~mwunsch@64.70.114.88)
# [21:05] <Loqi> [[existing-rel-values]] http://microformats.org/wiki/index.php?title=existing-rel-values&diff=44364&oldid=44361&rcid=57508 * Tantek * (+307) added rel-home to HTML5 registry given the actual use case of machine discovery of a feed for a site from a permalink page
# [21:06] <Loqi> [[existing-rel-values]] M http://microformats.org/wiki/index.php?title=existing-rel-values&diff=44365&oldid=44364&rcid=57509 * Tantek * (-2) /* HTML5 link type extensions */ typo
# [21:09] <singpoly1a> @tantek: that rel values (to me) reads: "this other page is *both* an alternate of this page *and* the homepage"
# [21:09] <singpoly1a> I'd rather leave the rel off altogether. discovery should be more based on type="" anyway, for most of the feed discovery use cases
# [21:12] <KevinMarks> singpoly1a: you men an alternate of the homepage?
# [21:13] <KevinMarks> *mean
# [21:15] <Loqi> [[rel-home]] http://microformats.org/wiki/index.php?title=rel-home&diff=44366&oldid=44363&rcid=57510 * Tantek * (+1253) added suggested use of home+alternate for linking to site feeds from permalink pages to a new brainstorming section
# [21:15] <singpoly1a> KevinMarks: no. tantek is using it to mean alternate of the homepage. It seems more like alternate of current page and is the homepage
# [21:15] <singpoly1a> rel values combine with "and", no?
# [21:16] <@tantek> singpoly1a - no
# [21:16] <@tantek> not quite
# [21:16] <@tantek> e.g. rel="
# [21:16] <@tantek> sorry
# [21:16] <@tantek> rel="alternate stylesheet"
# [21:16] <@tantek> they also combine with media
# [21:16] <singpoly1a> people use alternate and stylesheet on the same thing? what does that mean?
# [21:16] <@tantek> e.g. rel="alternate" media="handheld" is how you do auto discovery to the mobile version of a site
# [21:17] <@tantek> singpolyma: http://www.w3.org/TR/html4/present/styles.html#h-14.3.1
# [21:17] <singpoly1a> wow, I had never seen that before
# [21:17] <@tantek> semantics of rel values combine
# [21:17] <singpoly1a> alright then. I thought rel was more like class
# [21:18] <@tantek> so in this case, "alternate home" is the meaning, and the type makes it clear it's a feed
# [21:18] <singpoly1a> right. in that case, cool
# [21:18] <@tantek> rel is like class in that it is a space separated set of values
# [21:19] <@tantek> but class values are undefined in the HTML spec
# [21:19] <@tantek> however rel values are both defined in the HTML spec, and now have an official registry: http://microformats.org/wiki/existing-rel-values
# [21:20] <singpoly1a> so, how does xfn work with that? rel="friend met" mean both friend and met, no? I guess in this case "the kind of friend that I have met" is basically the same semantic
# [21:20] <@tantek> right
# [21:20] <singpoly1a> does order matter in rel values, then?
# [21:20] <@tantek> no
# [21:21] <singpoly1a> ok
# [21:21] <@tantek> per http://www.w3.org/TR/html5/semantics.html#attr-link-rel
# [21:22] <@tantek> dglazkov - I've added the brainstorm proposal for use of rel home+alternate here: http://microformats.org/wiki/rel-home#use_with_rel-alternate - please feel free to add your opinion :)
# [21:23] <@dglazkov> tantek: :)
# [21:23] <singpoly1a> interesting that the URI for that page is .../semantics.html
# [21:23] <@dglazkov> tantek: are you ever in MTV these days?
# [21:24] <Hixie> tantek: dude, don't use the TR/ page, that's always out of date
# [21:24] <@tantek> no I'm unfortunately too old to be on MTV which I think has an age cap of ~25?
# [21:25] <@dglazkov> tantek: heh. How about the age cap for Mountain View?
# [21:25] <singpoly1a> Hixie: I expect your version doesn't use the word semantics in any URIs? ;)
# [21:25] <@tantek> Hixie, noted. :)
# [21:25] <@tantek> http://dev.w3.org/html5/spec/semantics.html#attr-link-rel
# [21:25] <Hixie> that's better :-P
# [21:26] <Hixie> (the whatwg.org copy has a few more fixes still, but for most purposes the dev.w3.org and whatwg.org copies are equivalent)
# [21:27] <@tantek> cool
# [21:28] <@tantek> Hixie, in other news, I found an actual machine-readable use case for rel="home" when combined with rel="alternate" (see above http://krijnhoetmer.nl/irc-logs/microformats/20110810#l-61 )
# [21:28] <singpoly1a> So, that section of the spec sort of implies my initial intuition: " if a single link element has a rel attribute with the value next stylesheet, it creates both a hyperlink (for the next keyword) and an external resource link (for the stylesheet keyword)"
# [21:28] <Hixie> i saw
# [21:28] <Hixie> anyone planning on implementing it?
# [21:28] <Hixie> that's the thing that killed rel=feed :-(
# [21:28] <@tantek> not sure, I just started publishing it :)
# [21:28] <Hixie> (lack of implementations)
# [21:29] <@tantek> well rel=feed was unnecessary - all of its use-cases were already covered by existing solutions
# [21:29] <@tantek> in this case I'm pretty sure I've identified a problem that existing solutions don't solve
# [21:29] <@tantek> or I'm welcome to being shown the existing solution :)
# [21:30] <@tantek> Hixie - feed discovery implementations are plenty buggy and problematic btw
# [21:30] <@tantek> even Google Reader fails to parse link rel correctly (treats it as a literal rather than space-separated set of tokens)
# [21:30] <Hixie> not sure "even" is the word i would there :-P
# [21:30] <@tantek> would appreciate if you could ping the Google Reader folks to fix that - I've reported the bug years ago but to no avail
# [21:30] <Loqi> pong
# [21:31] <singpoly1a> most autodiscovery implementations tend to ignore <a> as well
# [21:32] <Hixie> tantek: i think in practice the use case is "subscribe to a feed", and rel=alternate type=rss/atom seems to handle it fine
# [21:33] <Hixie> tantek: the semantics of this are rather dubious of course
# [21:33] <Hixie> tantek: but since when has that ever won the day :-(
# [21:33] <@tantek> Hixie, the problem is that rel=alternate type=rss/atom on a permalink page would/should literally mean a feed for *that* page, e.g. a feed of the comments on the post
# [21:33] <@tantek> \
# [21:34] <@tantek> and some CMS's put out multiple rel alternates for a post permalink now
# [21:34] <singpoly1a> and there is in practise often *also* such a feed
# [21:34] <@tantek> and thus it's ambiguous which to use
# [21:34] <Hixie> sure
# [21:34] <Hixie> like i said, the semantics are quite dubious
# [21:34] <@tantek> so for the feed readers / indexers *that care*, rel="home alternate" provides a way to distinguish the feed for the *site*
# [21:34] <Hixie> but in practice people seem to be happy with just disambiguiting by title=""
# [21:34] <Hixie> so...
# [21:34] <Hixie> *shrug*
# [21:35] <singpoly1a> A machine can't really diambig by title
# [21:35] <Hixie> anyway, what matters at the end of the day is whether implementations decide this is a good idea
# [21:35] <@tantek> yeah, I tend to agree. however, I'm also fine with trying new semantic markup to see if anyone ends up caring years later :)
# [21:35] <Hixie> that's pretty much my philosophy too :-)
# [21:35] <singpoly1a> Hixie: what actually matters is if publishers decide it's a good idea
# [21:35] <Hixie> no
# [21:35] <@tantek> oh great Hixie, now you're just feeding the conspiracy folks. ;)
# [21:36] <Hixie> singpoly1a: what matters is whether users end up using it. for that you need publishers and implementors -- and implementors tend to be the harder nut to crack, as there's always far more publishers than implementors.
# [21:36] <@tantek> see, they have the same philosophy, they *must* be colluding!
# [21:36] * Joins: Askarii (~Askarii@216.113.83.169)
# [21:37] <singpoly1a> Hixie: if enough publishers implement it, then I can use it. If implementors only have it, that does nothing for me or my users
# [21:37] <singpoly1a> where by "implementors" I assume you mean your usual "the big N browsers"
# [21:37] <@tantek> singpolyma - when you say "I can use it" - are you implementing something?
# [21:37] <singpoly1a> tantek: yes
# [21:38] <Hixie> singpoly1a: by "implementors" i mean anyone writing consuming code
# [21:38] <singpoly1a> Hixie: well, if that's all you mean then having publishers is sufficient for small-time implementors to use it in useful code
# [21:38] <@tantek> singpolyma feel free to try it on my site, e.g. try it on this post permalink: http://tantek.com/2011/220/b1/web-actions-a-new-building-block
# [21:38] <Hixie> singpoly1a: i.e. anyone writing code to implement the feature in a user agent (i.e. any software working on behalf of a user)
# [21:39] <@tantek> (I just added this today because of a bug report that someone couldn't automatically find my feed from one of my posts)
# [21:39] <singpoly1a> tantek: I'll certainly roll support into my various things that do autodiscovery next time I'm working on them
# [21:39] <Hixie> singpoly1a: what matters is whether any small-time implementors or big-time implementors do use it
# [21:39] <Hixie> singpoly1a: there are plenty of features that have been used by billions of pages but never used to any significant extent by implementors
# [21:40] <singpoly1a> Hixie: like what?
# [21:40] <Loqi> [[rel-home]] http://microformats.org/wiki/index.php?title=rel-home&diff=44367&oldid=44366&rcid=57511 * Kevin Marks * (+111) /* use with rel-alternate */ check home in supported browsers
# [21:40] <Hixie> singpoly1a: rel=feed for example
# [21:41] <Hixie> actually that one wasn't adopted by authors, bad example
# [21:41] <Hixie> um
# [21:41] <@tantek> singpolyma - in practice yes, if you implemented it, it would make a difference for rel-home in terms of helping it advance from 'draft' to 'specification', as we now require at least 1 implementation (consuming code) not-from-the-author to advance to spec: http://microformats.org/wiki/process#Specifications
# [21:41] <Hixie> singpoly1a: rel=up would be a better example
# [21:42] <@tantek> Hixie, did authors ever use rel=up?
# [21:42] <Hixie> yeah, there's plenty of pages using it
# [21:42] <@tantek> I wonder what for
# [21:42] <@tantek> breadcrumbs?
# [21:42] <singpoly1a> Hixie: I've used userscripts that supported rel=up. also, I think Opera does/did support it
# [21:42] <singpoly1a> though I haven't seen many pages actually supporting it
# [21:43] <@tantek> singpolyma - feel free to add any documentation of specific support you know of (e.g. Opera documentation) to http://microformats.org/wiki/rel-up
# [21:43] <Hixie> rel=index is another example
# [21:43] <Hixie> rel=external was the 5th most-used value a few years ago
# [21:43] <Hixie> is rel=bookmark used by any consuming software?
# [21:43] <singpoly1a> so, up/index/next/prev/home are all things I've wanted to use / have used in the past, but the big problem for me has always been lack of publishers using it
# [21:43] <Hixie> that was the 4th most-used
# [21:44] <@tantek> Hixie, rel=bookmark was indexed by Technorati looking for blog post permalinks
# [21:44] <singpoly1a> bookmark is used by tons of microformats consuming software (as part of hAtom ot very least)
# [21:44] <Hixie> k
# [21:44] <Hixie> anyway my point is just that just because it's published doesn't mean it's widely used
# [21:44] * @tantek agrees.
# [21:44] <Hixie> er
# [21:45] <Hixie> (used as in implemented)
# [21:45] <@tantek> as in consumed/parsed
# [21:45] <singpoly1a> and by "widely" you mean "by the big 4/5" ?
# [21:45] <@tantek> singpolyma - if you're asking me, I'd say by *anyone*
# [21:45] <Hixie> singpoly1a: not necessarily the major browsers, no
# [21:46] <@tantek> and if you do know of implementations of any specific rel value, then please document/cite them on the wiki
# [21:46] <@tantek> in an "Implementations" section on the page for each rel value that you know that is implemented/parsed
# [21:46] <@tantek> without documentation of specific implementations, there's no evidence that there is any use/implementation.
# [21:46] <Hixie> singpoly1a: one tool used by 100 people isn't "widely". A few tools each used by a few hundred thousand might be.
# [21:47] <singpoly1a> use/implementation.
# [21:47] <singpoly1a> oops
# [21:47] <@tantek> to some extent, even a tool used by a few people would be worth documenting, as that might be a sign/harbinger of further implementation (not always, but sometimes is)
# [21:47] <singpoly1a> Hixie: right. so, if that's the criteria than I agree. good things tend not to get deployed by major corps, and thus not to groups of that size
# [21:48] <singpoly1a> tantek: I'm poking around a bit right now to see if I can find some impls I've used
# [21:49] <@tantek> singpolyma awesome
# [21:50] <@tantek> and if there's no page yet for the rel values you're documenting, feel free to create it
# [21:50] <Hixie> tantek: certainly the trend is relevant, yes.
# [21:50] <@tantek> even if you just start with
# [21:50] <@tantek> {{stub}}
# [21:50] <KevinMarks> so rel-home and rel-up are the first two links in a breadcrumb chain
# [21:50] <@tantek> == implementations ==
# [21:50] <Hixie> gotta go, bbiab
# [21:50] <@tantek> * [URL name] - description, what it does with the rel value
# [21:50] <@tantek> etc.
# [21:50] <KevinMarks> hm, or rathe the first and last links in the breadcrumb chain
# [21:51] <KevinMarks> what would you label intermediate ones?
# [21:51] <singpoly1a> up/down
# [21:53] <KevinMarks> <nav><a rel="home" href=''>home</a>:<a rel="???" href=''>2. formats</a>:<a rel="up" href=''>2.1 microformats</a></nav>
# [21:53] <KevinMarks> what's the ???
# [21:53] <singpoly1a> up up ;)
# [21:53] <@tantek> KevinMarks perhaps it would be a good time to start documenting a few *real world* examples of breadcrumb link structures so we can analyze them, e.g. on http://microformats.org/wiki/breadcrumb-examples :)
# [21:54] <@tantek> singpolyma - no that was rejected
# [21:54] <KevinMarks> down works if I'm on 2.1 and am pointing at 2.1.1
# [21:54] <@tantek> rel is a set
# [21:54] <@tantek> # of occurrences does not affect the semantic
# [21:54] <singpoly1a> that should go in the spec, then :P
# [21:54] <KevinMarks> strictly should have <ol> around the breadcrumbs
# [21:54] <@tantek> it is
# [21:54] <@tantek> that's what "set" means :P
# [21:55] * singpoly1a looks at spec again
# [21:55] <singpoly1a> oh... it uses "set" very colloquially. I didn't even notice
# [21:56] <@tantek> https://secure.wikimedia.org/wikipedia/en/wiki/Set_%28computer_science%29
# [21:56] <singpoly1a> Sure, I am aware of what "set" mean formally
# [21:58] <@dglazkov> like a tea set
# [22:01] <@dglazkov> tantek: have you been following the component model discussion at all?
# [22:01] <Loqi> [[rel-home]] http://microformats.org/wiki/index.php?title=rel-home&diff=44368&oldid=44367&rcid=57512 * Tantek * (+73) /* Discussions */ link to IRC discussion
# [22:02] <@tantek> dglazkov - no idea what you mean - would need a URL for specifics.
# [22:02] <@dglazkov> tantek: http://wiki.whatwg.org/index.php?title=Component_Model_Use_Cases
# [22:02] <@tantek> ah, the XBL2 conversation
# [22:02] <@tantek> no I'm not really following it as it's a huge problem to attempt to solve
# [22:03] <@tantek> I'm focusing on simpler solutions for the simple cases in CSS3-UI
# [22:03] <@tantek> I'm not even sure that any "component model" solution would be any simpler than just using Javascript libraries for such UI widgets
# [22:03] <@tantek> and if it's not simpler for authors, then why bother.
# [22:04] * Quits: Askarii (~Askarii@216.113.83.169) (Read error: Connection reset by peer)
# [22:05] * Joins: Askarii (~Askarii@216.113.83.169)
# [22:06] <@dglazkov> tantek: using JS libraries takes you a long way, certainly
# [22:06] <@dglazkov> tantek: but the trip always ends in the abstraction layer leaking and the mirage breaking down
# [22:08] <Loqi> [[rel-home]] http://microformats.org/wiki/index.php?title=rel-home&diff=44369&oldid=44368&rcid=57513 * Tantek * (+357) /* use with rel-alternate */ link to an example in the wild for those that want to see/try it perhaps with an implementation, noted possible implementation support
# [22:09] <@dglazkov> as described by http://wiki.whatwg.org/index.php?title=Component_Model_Use_Cases#Widget_Mix-and-Match
# [22:10] <Loqi> [[rel-next]] N http://microformats.org/wiki/index.php?title=rel-next&rcid=57514 * Singpolyma * (+614) list some implementations of rel-next
# [22:12] <singpoly1a> http://help.opera.com/Linux/9.50/en/toolbars.html mentions rel="next" explicitly, and then says it supports some other kinds of navigation links, but doesn't list which ones :(
# [22:17] <Loqi> [[rel-up]] N http://microformats.org/wiki/index.php?title=rel-up&rcid=57515 * Singpolyma * (+204) list some implementations of rel-up
# [22:28] <Loqi> [http://twitter.com/andyrutledge] RT @t: #aea @stubbornella CSS tips (@lukew http://j.mp/stbcss). one error: classes/#microformats *are* used by search: http://ttk.me/t4DG1
# [22:28] <Loqi> [http://twitter.com/smiller] RT @t: #aea @stubbornella CSS tips (@lukew http://j.mp/stbcss). one error: classes/#microformats *are* used by search: http://ttk.me/t4DG1
# [22:32] * Quits: mwunsch (~mwunsch@64.70.114.88) (Remote host closed the connection)
# [22:34] <Loqi> [http://twitter.com/justtobetrendy] RT @t: #aea @stubbornella CSS tips (@lukew http://j.mp/stbcss). one error: classes/#microformats *are* used by search: http://ttk.me/t4DG1
# [22:36] <Loqi> [http://twitter.com/valerierose] Food Bloggers: What are your thoughts on recipe formats e.g. microformats & hRecipe? cc @wubbahed @foodtechconnect
# [22:36] <Loqi> [http://twitter.com/feastie_dot_com] Food Bloggers: What are your thoughts on recipe formats e.g. microformats & hRecipe? cc @wubbahed @foodtechconnect
# [22:36] <@tantek> Thanks Loqi for the tweet mentions of microformats. And while I appreciate the RTs, I think you can leave them out.
# [22:36] <Loqi> you're welcome
# [22:41] * Joins: mwunsch (~mwunsch@64.70.114.88)
# [22:44] * Quits: mwunsch (~mwunsch@64.70.114.88) (Remote host closed the connection)
# [22:48] <Loqi> [http://twitter.com/t] @tabatkins that's not how rel works. already covered in IRC http://krijnhoetmer.nl/irc-logs/microformats/20110810#l-84 (ttk.me/t4DG3)
# [23:00] <Loqi> [[rel-home]] http://microformats.org/wiki/index.php?title=rel-home&diff=44372&oldid=44369&rcid=57516 * Tantek * (+774) /* use with rel-alternate */ added FAQ regarding misconception that rel values act independently (since now two smart people have asked the question)
# [23:02] <singpoly1a> Tantek: good to know I count as "smart people" ;)
# [23:14] <@tantek> now only if we could folks to discuss things on IRC instead of Twitter (ahem, tabatkins)
# [23:15] <hober> you could awlays /invite TabAtkins #microformats
# [23:16] <@tantek> hober, I tried
# [23:16] <hober> ahh
# [23:16] <@tantek> then I pm'd him
# [23:25] * Joins: TabAtkins (~tabatkins@nat/google/x-ufxxcnbvmvchxazp)
# [23:27] <TabAtkins> tantek: The sole example of interacting rel values is "alternate" and "stylesheet". In every other case, rel values are independent. Thus, "home" and "alternate" shouldn't combine either.
# [23:28] <@tantek> well I'm going for a run. but for now, in short, in reply to https://twitter.com/#!/tabatkins/status/101399365033271296 - there are of course non-sensical rel value combinations - that's what happens when you combine predefined values. and this is also a key outcome of how rel is different from class, since the keywords are defined and registered, deliberate can be taken to use them in combination and define special meanings when u
# [23:28] <@tantek> to do so.
# [23:28] <@tantek> deliberate *care
# [23:29] <TabAtkins> tantek: I'm not talking about nonsensical combinations. I can imagine situations in which "next stylesheet" makes sense.
# [23:29] * Joins: voxpelli (~anonymous@83.176.255.215)
# [23:30] <TabAtkins> Similarly, I can imagine situations in which "home stylesheet" make sense. If you throw an "alternate" in there, does it apply to one or the other, or perhaps both?
# [23:30] <TabAtkins> tantek: Combined processing without expliciting grouping is a bad thing. "home-alternate" gives a combination rel value that's completely unambiguous.
# [23:31] <@tantek> it has to be defined, but that's the point, because rel values have to be defined anyway, combinations are open to definition, which is what we're doing with "home alternate" because it's useful (has a practical machine-readable use case)
# [23:32] <@tantek> disagreed - explicit hyphenation of potential combinations of values leads to too many values.
# [23:32] <TabAtkins> I understand the use-case of defining a feed for the homepage (separate from the feed for the current page). I disagree with your attempt to assign meaning to rel combinations, which is unprecedented aside from a single example.
# [23:32] <@tantek> unprecedented aside from a single example = there's a precedent
# [23:32] <@tantek> and a well established one
# [23:32] <@tantek> thus it's reasonable to follow it
# [23:33] <TabAtkins> tantek: "Single precedent" and "legacy mistake" sound awfully similar.
# [23:33] <@tantek> instead of your general "dislike of"
# [23:33] <@tantek> if you think "alternate stylesheet" is/was a mistaken then make that point
# [23:33] <@tantek> rather than appealing to "you don't like it"
# [23:33] <TabAtkins> I pointed out a technical weakness of combined-rel semantics.
# [23:34] <@tantek> theoretical weakness that doesn't seem to have affected "alternate stylesheet"
# [23:34] <@tantek> so we're fine there
# [23:34] <@tantek> as KevinMarks noted, it would be interesting to see what existing rel="home" implementations do
# [23:34] <voxpelli> TabAtkins: eg. http://tools.ietf.org/html/rfc5988.html#section-4 is on your side - but not sure if I personally agree
# [23:35] <TabAtkins> tantek: Yes, because there's only one single example of it. This is the sort of problem that gets worse with scale. If every combination is potentially meaningful, it becomes more difficult to avoid meaning conflicts as you add more meaningful combinations.
# [23:35] <voxpelli> tantek: Did you see my tweet about the semantics of alternate in Activity Streams auto discovery?
# [23:35] <@tantek> KevinMarks' comment: http://microformats.org/wiki/rel-home#use_with_rel-alternate
# [23:36] <@tantek> voxpelli - did not - URL?
# [23:36] <@tantek> TabAtkins - no one ever said design was easy
# [23:36] <@tantek> ;)
# [23:36] <voxpelli> tantek: here's the thread on Google Groups: https://groups.google.com/forum/#!topic/activity-streams/XJm-AwQJigo/discussion
# [23:37] <TabAtkins> tantek: Glibness aside, do you accept that the problems caused by rel combinations don't show themselves strongly when there are few valid combinations, but do when there are many?
# [23:37] <@tantek> I think you can make that statement about combinations and complexity and problems about *any* space.
# [23:37] <@tantek> e.g. CSS border related properties ;)
# [23:38] <@tantek> doesn't mean we shouldn't try and do something useful with the simple combinations
# [23:38] <TabAtkins> I don't think the analogy is relevant. Combining orthogonal things is good - the combinatorial explosion works well for you. Combining *non*-orthogonal things may not be.
# [23:39] <@tantek> I'd rather try simple things and find out than not try for fear that the complex would fail.
# [23:39] <@tantek> in fact that's usually how things work
# [23:39] <@tantek> folks say, that will never work because of x complexity
# [23:39] <TabAtkins> I would rather go for a scheme that doesn't have the same technical problems but has equivalent semantics. ^_^
# [23:39] <@tantek> and then other folks say (or are too ignorant to know better) let's try this simple thing
# [23:39] <@tantek> and then it works
# [23:40] <@tantek> combinatorial complexity exists whether you use hyphens or not
# [23:40] <TabAtkins> Frex, defining that "alternate" within a comma-separated rel name has a particular meaning.
# [23:40] <@tantek> voxpelli - not sure what the point of that thread is
# [23:40] <@tantek> rel="feed" is unnecessary
# [23:40] <@tantek> that's been shown already
# [23:40] <TabAtkins> tantek: No, the problem is distinct. rel="home stylesheet alternate" is ambiguous. rel="home-alternate stylesheet" (or one of the other two possibilities) is not.
# [23:41] <TabAtkins> s/comma-separated/dash-separated/
# [23:41] <voxpelli> tantek: The point I was trying to make there was that the "alternate" relation might not be a proper way to link to a stream of related actvities
# [23:42] <@tantek> depends on from what page!
# [23:42] <voxpelli> tantek: which was kind of the same issue that you brought up with "home alternate" - "alternate" means an alternate representation of a specific page - not a link to something related
# [23:42] <@tantek> from a page that shows HTML updates, then using rel="alternate" to link to the feed or activity stream is correct
# [23:43] <@tantek> e.g. from tantek.com to /updates.atom - rel="alternate" is correct
# [23:43] <@tantek> but yes, alternate means an alternate representation of *the* page
# [23:43] <@tantek> not the site
# [23:44] <voxpelli> tantek: wouldn't rel="related" or similar be more correct for eg. post sites?
# [23:44] <@tantek> rel="related" is redundant
# [23:44] <@tantek> every link is somehow related
# [23:44] <@tantek> what's a post site?
# [23:45] <voxpelli> post page*
# [23:45] <@tantek> no
# [23:45] <@tantek> because those can have feeds too
# [23:45] <@tantek> e.g. of their comments
# [23:45] <@tantek> that's note in the brainstorm: http://microformats.org/wiki/rel-home#use_with_rel-alternate
# [23:46] <@tantek> *noted
# [23:46] <voxpelli> tantek: of course, but for eg. a twitter stream on a tweet page
# [23:46] <@tantek> every page that shows a stream in HTML could have an alternate in AS-JSON
# [23:46] <@tantek> regardless of the scope
# [23:46] <voxpelli> of course
# [23:47] <@tantek> even a tweet page
# [23:47] <@tantek> could have an AS-JSON of the folks that RT'd it
# [23:47] <voxpelli> and that should be a del="alternate"
# [23:47] <@tantek> because it's shown on the page
# [23:47] <Loqi> [[rel-home]] http://microformats.org/wiki/index.php?title=rel-home&diff=44373&oldid=44372&rcid=57517 * Singpolyma * (+342) /* use with rel-alternate */ quote the spec
# [23:47] <voxpelli> The common use case though is: 1. I visit a tweet. 2. I think the tweeter is brilliant and wants to follow him.
# [23:48] <voxpelli> So it was the case of the activity stream related to that specific use case that I was referring to
# [23:48] <Loqi> [http://twitter.com/getwired] @GrumpusNation I wrote up an XML microformat spec a year ago for restaurants. I wish Google would help small restaurants succeed.
# [23:49] <voxpelli> tantek: would you in such a use case use rel="home alternate" type="application/stream+json"?
# [23:50] <voxpelli> or perhaps even rel="author alternate"?
# [23:51] <@tantek> there's no such thing as an XML microformat spec
# [23:53] * Quits: abki|home (~amirouche@2a01:e35:2ef3:d930:21e:64ff:fe8b:aab2) (Ping timeout: 255 seconds)
# [23:58] <Loqi> [[rel-home]] http://microformats.org/wiki/index.php?title=rel-home&diff=44374&oldid=44373&rcid=57518 * Tantek * (+665) /* use with rel-alternate */ noting difference between non-normative examples (what you can't imply from) and what the spec actually normatively says.
# [23:58] <@tantek> voxpelli what's the use case specifically?
# [23:58] <@tantek> i.e. linking from *what* URL to what *other* URL?
# [23:59] <singpoly1a> tantek: thanks for the clarification on the wiki :)
# Session Close: Thu Aug 11 00:00:01 2011

The end :)