#microformats 2012-10-23

2012-10-23 UTC
robmorrissey, tantek and tantek_ joined the channel
#
tantek edited /admin-to-do (+120) "/* wiki */ more anti-spam measures" (view diff)
robmorrissey joined the channel
#
jrgenujp created /Portable_Restrooms_For_Rent (+2120) "New page: [ ]Preferably, it makes sense to buy a very simple camping bathroom tent to make make use of of with your portable potty. This provides you total privacy and allows you to do all the requi..." (view diff)
#
louisvuitton123456 created /日本より安いです。_『ルイビトンとエルメスはアウトレットには出していませんし、_空港免税店にもありません。_』エルメスは長財布でもかなりの金額になります (+6828) "ル作成ソフã" (view diff)
kennyluck joined the channel
#
qdzfug510 created /User:qdzfug510 (+3066) "" said Texans unpleasant remaining tackle Duane Brown" (view diff)
#
qdzfug510 edited /User:qdzfug510 (+3603) "/* no. 88 group */ new section" (view diff)
#
qdzfug510 edited /User:qdzfug510 (+3209) "/* and far more lucrative */ new section" (view diff)
iSRAELi, mamund, tantek and encolpe joined the channel
#
hcoach97 created /User:hcoach97 (+3077) "コーチ 仙台 Importance of Selecting Good Brands for Pillows" (view diff)
#
hcoach97 edited /User:hcoach97 (+3127) "/* コーチ セカンドバッグ How To Find High Quality Tag Heuer Rep */ new section" (view diff)
#
tantek edited /Special:Log/block () "blocked [[User:JrgenUjp]] with an expiry time of infinite (account creation disabled): Spamming links to external sites" (view diff)
#
tantek edited /Special:Log/block () "blocked [[User:Louisvuitton123456]] with an expiry time of infinite (account creation disabled): Spamming links to external sites" (view diff)
#
tantek edited /Special:Log/block () "blocked [[User:Qdzfug510]] with an expiry time of infinite (account creation disabled): Spamming links to external sites" (view diff)
#
tantek edited /Special:Log/block () "blocked [[User:Hcoach97]] with an expiry time of infinite (account creation disabled): Spamming links to external sites" (view diff)
#
tantek deleted /Portable_Restrooms_For_Rent "content was spam"
#
tantek deleted /日本より安いです。_『ルイビトンとエルメスはアウトレットには出していませんし、_空港免税店にもありません。_』エルメスは長財布でもかなりの金額になります "content was spam"
#
tantek deleted /User:qdzfug510 "content was spam"
#
tantek deleted /User:hcoach97 "content was spam"
zztr joined the channel
#
tantek welcome zztr!
#
zztr judging by the beginning of the wikipedia page, I very much like the OpenDoc approach "create small, reusable components, responsible for a specific task" and "a framework in which these components could run together"
#
zztr hi tantek :)
#
zztr this is what I see as part of the power of the command line
#
zztr i want to bring that composability to the gui somehow
iSRAELi joined the channel
#
tantek interesting
#
tantek well let's at least talk more about citation microformat stuff
#
tantek and see what we can get working with that
#
tantek maybe we can talk command line / ubiquity (that Mozilla experiment) too
nonge_, manu1 and romainneutron joined the channel
#
@javiaurea Contact
#
Loqi e-mail, Skipe, hCard (http://aurea.es/contacto-2/?utm_campaign=WP-to-Twitter&utm_medium=twitter&utm_source=twitter)
tobyink and csarven joined the channel
#
tantek edited /microformats2-faq (+4) "s/fn/name" (view diff)
#
tantek edited /microformats2-faq (-7) "s/s-2/s2" (view diff)
#
tantek created /microformats-2-parsing-faq (+39) "r" (view diff)
#
tantek created /microformats2-parsing-faq (+1096) "drafted" (view diff)
#
tantek edited /microformats2-parsing-faq (+10) "g" (view diff)
tobyink and robmorrissey joined the channel
#
rossanamrocozjsnkstelzer created /My_personal_free_advice_on_how_to_get_your_ex_back_when_your_ex_move_on.2857187 (+2872) "New page: Are you tired of spending your money on get your ex back online courses which gives you no positive result? Then don’t worry in this article I am going to give you how to get your ex bac..." (view diff)
#
rossanamrocozjsnkstelzer created /My_individual_tip_for_you_personally_if_you_are_searching_to_find_out_how_to_get_your_ex_back.4330856 (+2624) "New page: Are you suffering in the painful sensation of breakup and want to understand how to get your ex back inside your lifestyle as soon as once more? Numerous of us encounter this kind of painf..." (view diff)
adb and csarven joined the channel
#
sdhu378wsa created /User:sdhu378wsa (+3115) "toms shoes priced at" (view diff)
#
sdhu378wsa edited /User:sdhu378wsa (+2848) "/* toms eyewear wiser */ new section" (view diff)
encolpe, MacTed, tantek and robmorrissey joined the channel
#
aureliagepfwptnbvkipple created /Tecnicas_Efectivas_de_Posicionamiento_en_Google4274697 (+3361) "New page: Cómo permanecer en la web Alcanzar que la gente accedan a mi sitio web y no al de la competencia es el objetivo de todas las personas que tienen una página web en el internet, de nada m..." (view diff)
#
@MattTurnure Microformats validator: http://hcard.geekhood.net/
robmorrissey joined the channel
#
@FreeWSOdownload Free download: WSO - WP Author hReview (Plugin) http://www.freewso.info/wp-author-hreview-plugin/?utm_source=dlvr.it&utm_medium=twitter
adactio, robmorrissey, tantek, csarven, tobyink and barnabywalters joined the channel
#
Loqi barnabywalters: tantek left you a message on 10/22 at 2:53pm: the normalization step, where it is, is deliberate, as that is following the semantics of those attributes (normalization to absolute URLs), where-as when present in a title attribute or in text, the expectation both in normal usage and in the browser is *no* normalization.
#
tantek and I left it in the wrong channel
#
tantek lol
#
barnabywalters hm, aaronpk has updated Loqi to not use relative dates any more
#
barnabywalters okay, that makes sense
kennyluck joined the channel
#
tantek edited /microformats2-parsing-faq (+88) "sample Q&A to make the TOC show up" (view diff)
#
tantek edited /Special:Log/block () "blocked [[User:RossanamrocozjsnkStelzer]] with an expiry time of infinite (account creation disabled): Spamming links to external sites" (view diff)
#
tantek edited /Special:Log/block () "blocked [[User:Sdhu378wsa]] with an expiry time of infinite (account creation disabled): Spamming links to external sites" (view diff)
#
tantek edited /Special:Log/block () "blocked [[User:AureliagepfwptnbvKipple]] with an expiry time of infinite (account creation disabled): Spamming links to external sites" (view diff)
#
tantek deleted /My_personal_free_advice_on_how_to_get_your_ex_back_when_your_ex_move_on.2857187 "content was spam"
#
tantek deleted /My_individual_tip_for_you_personally_if_you_are_searching_to_find_out_how_to_get_your_ex_back.4330856 "content was spam"
#
tantek deleted /User:sdhu378wsa "content was spam"
#
tantek deleted /Tecnicas_Efectivas_de_Posicionamiento_en_Google4274697 "content was spam"
#
barnabywalters I looked over the implied properties parsing, and have a slight issue
#
barnabywalters this approach is going to cause loads of false positives, giving potentially long and confusing p-name etc properties to microformats which should never have a p-name
#
barnabywalters e.g. h-adr’s p-name will end up being the entire address as well as any other text within that element
#
tantek barnabywalters, we *do* have the use-case of named locations
#
tantek imagine that you had a list of locations, by adr, geo, venue etc. and you needed to put *something* in that list as the *label* of the location item
#
tantek that *something* would have to be the equivalent of a p-name
#
tantek so it's actually a feature to automatically get such a labe/title/*name*
#
barnabywalters tantek: which makes this even more of a problem. If there wasn’t the use case of named locations, we could effectively ignore any implied p-name false positives
#
tantek for almost all software that extracts such data, of any kind, and puts such data in a list
#
barnabywalters but as it is a use case, unnamed locations will get long and potentially confusing names
#
tantek no it makes it just right, because an explicitly named location will get a p-name from the author, and others will get a synthetic one
#
tantek if you have a list of locations, you have to put *something* as the list item labels
#
tantek and rather than something unuser friendly like "unnamed location 1" etc.
#
tantek at least you'll get the long form of the address or geo or whatever that the author *published* as *visible* content
#
tantek and that's what saves us
#
tantek the author *published* as *visible* content -> it's unlikely to be "long and potentially confusing"
#
barnabywalters tantek: unless the author is including the h-adr in prose
#
tantek we're just taking a snippet of text directly from what the author published as prose
#
tantek I think in practice it will be fine because such automatic p-names come *directly* from a snippet of text authored by a human, in the order that the human authored it, to be read anyway on the web page they published.
#
tantek and if the author doesn't want a p-name at all, easy: <span class="p-name"></span>
#
tantek or rather, wants an empty p-name
#
tantek which still, doesn't make sense from the inclusion in a list perspective
#
tantek everything you put into a list needs a short label/title/name for the unexpanded version of the item in the list
#
tantek automatic p-names gives you this nicely
#
tantek in the old days of microformats, the issue you're raising would be referred to as a "theoretical issue". let me know when you come across an actual real world example where you think the implied p-name provides a bad experience in some actual use-case.
#
tantek :)
#
barnabywalters beat me to it ;) I was just writing “true. I suppose if there is an actual issue we will notice it and fix it”
#
tantek now how would you phrase that as a short neutral point of view question so I can capture it in the FAQ? (because it is a good question with non-obvious answer - I'd never thought of the list item label/title/name use-case until you asked your question)
#
barnabywalters tantek: thinking about it, caring for+using implied properties is actually something which we could encourage as part of µf2
encolpe joined the channel
#
tantek I think it will become self-evident without encouragement just by people running parsers / validators to see what their pages produce
#
tantek especially since you can then use all the p-name properties to construct a short *expandable* list of all the microformats in a page by name
#
tantek (in your parser sample output)
#
tantek rather than just "h-card 1", "h-card 2" etc.
#
tantek it allows for a more human friendly/readable expandable list view of all the microformatted data on a page
#
tantek and you can use the implied URL if any to link such a name to the actual item, and precede the name with the implied photo sized down to icon size as well
#
tantek #thisiswhatpopsinmyheadfirstthinginthemorning
#
barnabywalters RE neutral+short version, how’s: “Will using an element’s innerHTML as an implied p-name cause confusing false-positives”
#
barnabywalters heh, it’s almost 6 over here, which is *not* my optimal creative thinking time :/
#
tantek that's not NPOV - "will … confusing false-positives…." is a negative point of view :P
#
barnabywalters ah! missed that :)
#
barnabywalters “Will using an element’s innerHTML as an implied p-name cause false-positives”
#
tantek even false-positive is a negative perspective
#
tantek perhaps phrasing as a question of use would be better
#
barnabywalters well, to a certain degree. arguably if the author did not intend it it’s a false positive
#
tantek especially with the specific microformat you mentioned: h-adr
#
tantek it's actually *not* a false positive since it's 100% by design
#
tantek and predictable, testable etc.
#
tantek the term false-positive only applies to probabilistic things
#
barnabywalters yes, perhaps this is better presented as a “positive+useful side effect” of implied properties as opposed to a FAQ item
#
tantek something more along the lines of: what is the use-case implied p-name properties for microformats like h-adr where authors do not typically specify names for such structure? (but that's too long)
#
tantek a-ha I knew I'd seen something like this before
#
tantek see bottom of http://microformats.org/wiki/microformats2-implied-properties#root_class_only_and_name_property
#
tantek issues: may not work for all microformats, e.g. how would adr or geo work with this?
#
tantek would that be reasonable approximation of the question you asked?
#
tantek how does h-adr or h-geo work with implied p-names?
#
tantek how do h-adr and h-geo work with implied p-names?
#
barnabywalters that seems to fit the bill as well as addressing an existing issue
#
tantek ok, that's good that you encountered the same question as I did before when I logged that issue, definitely means that's worthy of including as an FAQ. perhaps more of a general microformats2-faq rather than just a parsing-faq since it effects authors
#
barnabywalters yesterday I got sidetracked fixing bugs, but php-mf2 is ready to have implied parsing added, so we can actually test this with some real world data
#
barnabywalters and see what happens :)
#
tantek woohoo!
#
tantek we can try it with the growing list of examples in the wild
#
tantek btw now in blog posts whenever I just mentions someone's name or link to them I'm using class="h-card" because it's so much easier to type so little
#
barnabywalters fab!
#
tantek I'm not bothering with class="vcard" and explicit class="fn" etc. too much typing/markup when I just want to keep blogging
#
barnabywalters I think that’s pretty much in line with what others are doing for inline mentions of people
#
tantek e.g. http://tantek.com/2012/296/b1/san-francisco-election-flyer-statistics
#
tantek right
#
barnabywalters last question: when checking for whether or not to look for implied properties, am I looking for p-name or *-name in the list of parsed property names?
#
barnabywalters the parsing guidelines are vague, they don’t specify either way
#
barnabywalters to a certain extent, e-name would be useful
#
barnabywalters u-name or dt-name would be fairly ridiculous
#
tantek in the list of parsed property names, there are no prefixes
#
tantek is that not self-evident from the simple Parsed JSON examples? microformats.org/wiki/microformats2
#
tantek also the order of operations in the parsing algorithm is pretty clear
#
tantek you only parse for implied properties *after* parsing an element for explicit properties
#
tantek per http://microformats.org/wiki/microformats2-parsing#parse_an_element_for_microformats
#
tantek and parsing for implied properties specifically refers to "name", "url" and "photo"
#
tantek http://microformats.org/wiki/microformats2-parsing#parsing_for_implied_properties
#
tantek unprefixed versions of the property names
#
tantek the p- u- dt- e- prefixes are basically parse-time directives that are dropped once parsing is done and you collect the properties/values into your parse tree / JSON.
#
barnabywalters php-mf doesn’t output the canonical JSON yet. Another thing on the todo list
#
tantek odd - it looked like it almost did
#
barnabywalters nope, not quite
kennyluck joined the channel
#
aaronpk barnabywalters: actually loqi switches to absolute dates after some threshold
#
Loqi is done
#
barnabywalters aaronpk: oh, really? cool, haven’t noticed that before
#
berthachan6 created /Ohhh_baby_I_like_It_Paintball_rawwwwww_ODB (+1146) "New page: [[Image:Paintball_4337.jpg|thumb|]] Could Baseball Have a "Bounty-Gate" Equivalent? | FF para otras maravillosas comunidades :D dale amorsha paso algo? Happy International Womens Day! In ..." (view diff)
#
tantek edited /Special:Log/block () "blocked [[User:BerthaChan6]] with an expiry time of infinite (account creation disabled): Spamming links to external sites" (view diff)
#
tantek deleted /Ohhh_baby_I_like_It_Paintball_rawwwwww_ODB "content was spam"
#
barnabywalters argh. XPath syntax for only-node is *hideous*
#
barnabywalters .//img[count(preceding-sibling::*)+count(following-sibling::*)=0]
#
barnabywalters I added support for all the implied properties in http://microformats.org/wiki/microformats2-parsing
#
barnabywalters the only thing which isn’t done right now is resolving relative URLs
#
tantek that's great barnabywalters!
#
tantek does it work in the test page you set up?
#
barnabywalters and now I’m doing real versioning, we’re on v0.1.1 (https://packagist.org/packages/mf2/mf2)
#
barnabywalters yep, I just updated it on my vps
#
barnabywalters give it a test run!
#
tantek remind me the URL?
#
barnabywalters http://waterpigs.co.uk/php-mf2/
#
barnabywalters URL entry field coming soon ;)
#
barnabywalters hm, looks like there’s some odd behaviour happening not covered by my tests
#
tantek hmm, just tried this source and didn't get any properties: <a class="h-card" href="http://tantek.com/">Tantek Çelik</a>
#
tantek edited /microformats2 (+187) "/* Implementations */ packagist and live URLs for php-mf2" (view diff)
#
barnabywalters tantek: yep, turned out my update failed. Should be working in a min
#
tantek barnabywalters perhaps try running the simple examples at the top of http://microformats.org/wiki/microformats2#simple_microformats_2_examples and make sure it outputs the items and properties as shown with in the Parsed JSON sections?
barnabywalters and romainneutron joined the channel
#
barnabywalters tantek: I fixed a couple of bugs, so http://waterpigs.co.uk/php-mf2 is working a lot better with implied properties now
#
barnabywalters still a few odd bits of behaviour, but it’s working a lot better
#
tantek that is much better
#
tantek key thing - remove the prefixes when putting the properties into the JSON
robmorrissey joined the channel
#
tantek e.g. [p-name] => Tantek Çelik - should just be - [name] => Tantek Çelik
BillCriswell joined the channel
#
tantek edited /mediawiki-customization (+64) "/* How to */ link SyntaxHighlight_GeSHi" (view diff)
#
tantek edited /mediawiki-customization (+96) "see also WHATWG MediaWiki customizations" (view diff)
Firass-z0r and tobyink joined the channel
#
tantek goes back to collecting & writing microformats2 FAQ entries.
#
barnabywalters *drumroll*
#
barnabywalters …aaaaand…
#
barnabywalters http://waterpigs.co.uk/php-mf2/ outputs canonical JSON µf representation instead of print_r‘ing an array!
#
tantek woot!
#
Loqi yay!
#
barnabywalters it doesn’t quite handle nested µf right
#
barnabywalters but it is a start
#
tantek it's a huge step forward
#
tantek "type" should be an array
#
barnabywalters good catch. didn’t notice that
#
tantek "type": ["h-card"] rather than just "type": "h-card"
#
tantek try parsing:
#
tantek <a class="h-card h-person" href="http://tantek.com/">Tantek Çelik</a>
#
tantek should give you a type value lik
#
tantek like
#
barnabywalters I’m on v0.1.5 already
#
tantek "type": ["h-card", "h-person"]
#
barnabywalters ah, php-mf2 won’t do that without some alterations
#
barnabywalters is there an example of that on the wiki?
#
tantek yes
#
tantek see http://microformats.org/wiki/microformats2#h-card_org_h-card
#
tantek scroll down to
#
tantek FOR PARSERS ONLY:
#
tantek it's inside a nested microformat
#
tantek but you can extract it and test it standalone as well
#
tantek this:
#
tantek <a class="p-org h-card h-org"
#
tantek href="http://mozilla.org/"
#
tantek >Mozilla Foundation</a>
#
tantek ...
#
tantek "type": ["h-card", "h-org"],
#
tantek "properties": {
#
tantek "name": ["Mozilla Foundation"],
#
tantek "url": ["http://mozilla.org/"]
#
tantek }
#
tantek (fragments obviously)
#
tantek BTW hence the first step of: http://microformats.org/wiki/microformats2-parsing#parse_an_element_for_microformats
#
barnabywalters fab, I’ll add GH issues for this and the other things to do
#
tantek is:
#
tantek "parse element class for root class name(s) "h-x" (and backcompat)"
#
tantek note the "root class name(s)"
#
tantek pluralization
#
tantek coming along quite well barnabywalters - this is great stuff
iSRAELi joined the channel
#
tantek have you tested all the simple examples in the wiki to see if they work as expected?
#
tantek they're all looking good to me (except for the "type" should be an array detail)
#
barnabywalters added at https://github.com/indieweb/php-mf2/issues/3
#
barnabywalters thanks :) Making a generic parser is a fun challenge
#
barnabywalters I went through some of them to make sure. I’m using some of them as test cases, so I don’t have to look at them per say
#
barnabywalters just run `phpunit /tests` ;)
#
tantek great!
#
tantek edited /microformats2-parsing (+95) "see also faq" (view diff)
#
tantek edited /microformats2-parsing-faq (+1123) "checking for explicit properties before implying" (view diff)
#
tantek edited /microformats2-parsing-faq (+24) "/* see also */ see also faq" (view diff)
#
barnabywalters Okay, I added the multiple type thing, and better handling of nested µf
#
barnabywalters at the moment all nested µf are just put in a `children` property, https://github.com/indieweb/php-mf2/issues/4 documents the need for ones attached to a property to be filed under that
#
barnabywalters phew! That’s enough for tonight :) Thanks for the help tantek
#
tantek thank you too barnabywalters!
#
tantek I'm just now writing up the FAQ about how do h-adr and h-geo work with an implied name property
#
tantek certainly worthy of explaining
#
tantek and a nice prod to get me to resolve that already outstanding issue in the spec :)
#
tantek check it out in (your) morning and see if it makes sense - I'll leave a !tell with URL for you
#
barnabywalters oh cool. actually I haven’t tested how effectively that works with h-adr yet
#
barnabywalters loqi is so useful for cross-timezone collaboration
#
barnabywalters ’s brain has seized up due to TOO MANY nested objects and arrays, and so is signing off. Goodnight!
#
Loqi don't let the bed bugs bite
#
tantek edited /microformats2-faq (+2560) "how do h-adr and h-geo work with an implied name property" (view diff)
#
tantek edited /microformats2-implied-properties (+135) "/* root class only and name property */ replace issue with a link to the FAQ and a short inline answer." (view diff)
#
tantek !tell barnabywalters FAQ on the subject as promised: http://microformats.org/wiki/microformats2-faq#how_do_h-adr_and_h-geo_work_with_an_implied_name_property
#
Loqi Ok, I'll tell him that when I see him next
#
apollosan edited /last-modified-examples (+76) "/* See also */" (view diff)
#
tantek edited /last-modified-examples (-76) "Reverted edits by [[Special:Contributions/Apollosan|Apollosan]] ([[User talk:Apollosan|Talk]]) to last version by [[User:Tantek|Tantek]]" (view diff)
#
tantek edited /Special:Log/block () "blocked [[User:Apollosan]] with an expiry time of infinite (account creation disabled): Spamming links to external sites" (view diff)
iSRAELi joined the channel
#
andymcmullen6 created /Becky_Gould_How_do_we_pay_for_this_tax_cut_How_will_we_ href="http://www.1stlevelbacklinks.com/_pay_for_state_priority_issues_like_child_welfare_LB970_neleg" target="_blank">http://www.1stlevelbacklinks.com/_pay_for_state_priority_issues_like_child_welfare_LB970_neleg (+987) "New page: [[Image:http://www.1stlevelbacklinks.com/_3722.jpg|thumb|]] JUST ANNOUNCED: Travis Outlaw will wear jersey no. 25. ReturnOfTheKings You've Gotta Put In Work (*^^*)^^ (^^)* THAT'S MY BEST ..." (view diff)
#
tantek edited /Special:Log/block () "blocked [[User:AndyMcmullen6]] with an expiry time of infinite (account creation disabled): Spamming links to external sites" (view diff)
#
tantek deleted /Becky_Gould_How_do_we_pay_for_this_tax_cut_How_will_we_ href="http://www.1stlevelbacklinks.com/_pay_for_state_priority_issues_like_child_welfare_LB970_neleg" target="_blank">http://www.1stlevelbacklinks.com/_pay_for_state_priority_issues_like_child_welfare_LB970_neleg "content was spam"
robmorrissey and romainneutron joined the channel