#microformats 2014-07-23
2014-07-23 UTC
lmjabreu_, tantek and KartikPrabhu joined the channel
pfefferle, tobyink, krendil, RCheesley, KartikPrabhu and barnabywalters joined the channel
# Loqi barnabywalters: tantek left you a message 10 hours, 55 minutes ago: I found a page with hProduct but phpmf2 doesn't seem to find it: http://pin13.net/mf2/?url=http://www.staples.com/Swingline-747-Rio-Red-Stapler-20-Sheet-Capacity/product_562485 - however mf2py picked it up: https://gist.github.com/tommorris/5d9111af10b606c3f3bd (via tommorris)
# barnabywalters good grief, that’s disgusting
# barnabywalters not that, the XML inside HTML with duplicate elements for microformats and whatever the XML vocabulary is
# barnabywalters all of which is invisible data
# barnabywalters no, I’m talking about the bit of markup which actually contains hproduct: https://gist.github.com/barnabywalters/66bc200190c4ae52844e
# barnabywalters exactly
# barnabywalters aggressively ignores schema.org
# KartikPrabhu schema is a good example in overengineering
# KartikPrabhu i actively deleted old schema markup from my site when I upgraded to mf2
# barnabywalters horrible markup aside, it’s my fault the backcompat isn’t parsing as I somehow forgot to add hproduct backcompat. fixing now
# barnabywalters tommorris: uh, aren’t some of those mixing properties and types a little?
# barnabywalters or is this hypothetical? :)
# barnabywalters e.g. type: yago:SoftwareDeveloper, job-title: Head of Forestry
# barnabywalters obviously such contradictions are possible with saner types, but less likely
# barnabywalters I suppose that’s true
pfefferle joined the channel
# barnabywalters !tell tantek thanks for the heads-up about hproduct, for some reason the backcompat mappings were missing. Pushed to master, will add to the next tagged release
# barnabywalters tommorris: thanks for the ready-built class mapping :) what’s the rationale behind adding e-description to the classes for review?
# barnabywalters I removed it for maximum spec-compatibility
# barnabywalters assuming that if you had a solid reason with real-world examples it would be documented on the wiki
# ChiefRA barnabywalters your parser highlights the errors too ?
# ChiefRA I'm interested in parsing a mf1 hlisting and to spot the implementation errors
# barnabywalters ChiefRA: what do you mean?
# barnabywalters php-mf2 doesn’t do any particular error reporting
# barnabywalters as it’s quite difficult to figure out what an “error” is
# ChiefRA :-? but it should be no?
# barnabywalters it sounds like the sort of errors your talking about are the job of a validator, not a parser, to detect
# ChiefRA in deed, validator :)
# ChiefRA do you know any?
# barnabywalters e.g. indiewebify.me/validate-h-entry/ for h-entry
# barnabywalters not for h-listing, unless google’s rich snippet testing tool handles them
# barnabywalters I don’t even know what hlisting is tbh :)
# ChiefRA yeah, the problem with Google is that it doesn't say anything about it, after parsing.
# ChiefRA It would be tremendously useful such a validator.
# ChiefRA it's for real estate
# ChiefRA listing all the properties one after another, like hproduct
# ChiefRA but specifically designer for real estate.
# barnabywalters apparently it wasn’t widely enough used to make the transition to microformats2 (yet, at least)
adactio joined the channel
# ChiefRA well, that's in part Schema's fault.
# barnabywalters it sound like there’s a lot of overlap with h-product
# ChiefRA yes, to some point, but being specifically designed for Real Estate, it should help.
# ChiefRA the thing is, I need a validator :) do we have a validator for h-product at least?
# ChiefRA I can use it like that by replacing the main name.
# barnabywalters nope — building useful validators is a lot of work and I personally only have time and expertise to maintain them for the most commonly used vocabularies
# barnabywalters as can be seen by the number of issues on indiewebify.me :) https://github.com/indieweb/indiewebify-me/issues
# ChiefRA ok, so do we have ANY kind of mf2 validator?
# ChiefRA do you*
# barnabywalters but feel free to adapt any code from indiewebify.me which is useful into a h-product/hlisting validator
# barnabywalters “mf2” validator — validators generally are per-vocabulary
# barnabywalters so indiewebify.me has validators for h-entry and h-card
# ChiefRA ok, it seems like we need to add the h-product as well.
# barnabywalters http://shrewdness.waterpigs.co.uk/test/ isn’t exactly a validtor yet but a testing tool for h-feed
# barnabywalters tommorris: indiewebify.me doesn’t even do that (u- vs p-) level of validation yet as it just looks at the parsed output
# ChiefRA Yeah, I would love to have a validator which can point you in the right direction, like: hey, your u-photo is outside of the corresponding DIV or smth like this.
# barnabywalters “see whether it parses right” — is very easy for humans but a lot more difficult for computers :)
# ChiefRA barnabywalters yes, totally agree :) that's why we have to do it if possible :)
# ChiefRA to keep up with the Schema spreading...
# ChiefRA Google's own tool
# barnabywalters a logical validator for each of the hundreds of types? or more of a syntatic validator?
# ChiefRA barnabywalters I would say a syntactic validator, to look for the classes and to tell you that you have missplaced the elements in-out of the divs etc.
# ChiefRA as ussually that's the problem.
# barnabywalters tommorris: lol
# ChiefRA :)))))
# ChiefRA seems that title is still a ranking factor, although browsers desn't display it anymore since TABs are taking over
# ChiefRA tabbed browsing
# tobyink If somebody supplemented http://schema.rdfs.org/ with some useful disjunctions (you can't be a Dentist and an OceanLiner simultaneously) then any OWL processor fronted by an RDFa/Microdata parser could act as a very nice schema.org validator.
# tobyink http://rdf.data-vocabulary.org/rdf.xml is old. schema.org replaces it. schema.org's schemas are only published in HTML, so schema.rdfs.org is a third-party RDF-ification of them.
# barnabywalters tommorris: what if I told you… that tweets were all title
# tobyink http://rdf.data-vocabulary.org/ChangeLog - no changes since 2009.
# ChiefRA there is breadcrumb on schema.org
# tobyink rdf.d-v.org was Google's solo effort at creating an RDF vocab, before they threw their lot in with MS and Y!
# ChiefRA and it has even a well defined implementation syntax :)
# ChiefRA Schema.org is winning ground in front of mf...
# ChiefRA the thing is, I still work with A LOT of HTML4* websites,
# ChiefRA so I can';t implement schema....
# barnabywalters unless of course you actually care about validation
# ChiefRA well, indeed, but if I change it, I get A LOT of w3c errors.
# ChiefRA Oh, I do care about it :)
# ChiefRA me me , pick me!!
# ChiefRA :))
# barnabywalters ChiefRA: out of interest, why?
# ChiefRA it's an important step in keeping your code clean and "understandable" by browsers etc.
# ChiefRA Google does not take into account - is not a ranking factor - validation.
# ChiefRA but can get you down as a side effect, if Googlebot can't fully understand your HTML code.
# ChiefRA if there is harsh competition on certain niches, Googlebot will pick your competition instead of your site if it can better comprehend their HTML code.
# barnabywalters it can be useful for spotting syntax errors, but google has it’s own tools (as you posted earlier) for making sure it’s crawlers can understand your code
salanto joined the channel
# ChiefRA John Mueller said is not mandatory to validate your website, but it certanly helps if you do. :)
# barnabywalters so if google’s tools tell you they can understand the rich snippets, why care whether or not the markup is offically allowed in whatever fake version number of HTML you’re forced to publish?
# ChiefRA tommorris I do agree that they (Google) are not perfect.
# ChiefRA they do have a lot of room for improvements.
# ChiefRA but that depends on us, to place bug reports etc.
# ChiefRA they can't be everywhere anytime to spot all the bugs unless our community helps them with it.
# ChiefRA I always submit bug reports and alike when I find them, and not necessarely HTML related.
# ChiefRA I can give you the most recent example: because I submitted an Android bug related to how Android is displaying numbers when you save them in the phone book and how it's displaying them when you dial, now they fixed the bug and my contry, Romania, has the numbers sorted out the right way, as we are used to see them on all other devices.
# ChiefRA and that's worldwide implemented on all Adroid devices.
# ChiefRA so I do believe spotting the bugs and taking the time to report them helps.
# barnabywalters ChiefRA: cool, nice that you got to see the whole feedback loop!
# ChiefRA barnabywalters :) it took me 7 months, but I've lived to see the result on my Galaxy S5 :D
# ChiefRA that's why I proposed that if you can, you should implement other vocabularies into your validator barnabywalters :)
# barnabywalters interesting that validator.nu is funded my Mozilla: http://about.validator.nu/##I+would+like+to+thank
# barnabywalters ChiefRA: you’re right, I think there’s some value in a more generic syntax validator as well as vocab logical validators. I might try putting one together
# ChiefRA barnabywalters please by all means, please tell me about it when you do, as I will be testing it to report porblems :D
# ChiefRA Error: Almost standards mode doctype. Expected e.g. <!DOCTYPE html>.
# ChiefRA From line 1, column 1
# ChiefRA to line 1, column 121
# ChiefRA <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">↩<html
# ChiefRA =)) validator.nu has a LOONG way to go to be near perfect.
# ChiefRA I believe you, but if I change it to HTML5 doctype, I get a LOT of errors.
# ChiefRA which have to be fixed, and the coders are way too busy to do that.
# ChiefRA and it's a whole platform.
# ChiefRA I can't fix it manually for each website.
# ChiefRA it has to go through a long process.
# ChiefRA so I skip it for now and try to patch things up, like using mf instead of schema etc.
# ChiefRA Don't get me wrong, i love mf.
# ChiefRA the only thing i cry about is that hlisting is not YET supported by Google.
# ChiefRA I mean it sees it, but that's about it. It doesn't treat it in any way in search resulta.
# ChiefRA results*
# ChiefRA I know. but that involves changing the DOCTYPE :)
# ChiefRA so here we go again into the same problem :)
# tommorris It's cargo cult programming. https://en.wikipedia.org/wiki/Cargo_cult_programming ;)
# ChiefRA damn, it's the first time I heaf about this term.
# ChiefRA thanks :D
# ChiefRA I'm smarter now.
netweb joined the channel
# ChiefRA yeah, i believe so...
# ChiefRA did you guys knew about this? https://code.google.com/p/aump/
# ChiefRA it's kinda old if I look at the archive date of creation.
# barnabywalters ChiefRA: I hadn’t come across it, do you use it?
# tommorris I could say <!DOCTYPE html PUBLIC "Tom's Magical HTML variant - http://tommorris.org/magic/pixies.dtd"> but that doesn't mean WebKit or Gecko is going to treat the page any differently. All it cares about is the "DOCTYPE html" bit to not chuck it into quirks mode
# barnabywalters the demo page is broken http://code.davidjanes.com/aumfp/demo/
# ChiefRA nope, I just stumble upon it now, searching for hlisting into the Google code, for some real examples.
# ChiefRA tommorris totally agree with you :)
# ChiefRA mentality it the FIRST thing that has to change
# ChiefRA it's*
# ChiefRA I do want it if possible, but I would love to have it as a validator of somewhat
# ChiefRA if not, I can manually debug it with the help of the parser output.
# ChiefRA ba da.
# ChiefRA tommorris and barnabywalters if you will create the Hlisting parser please leave me a PM :) I'll happily use it :)
# ChiefRA I have to go now, see you all later. and thanks :)
# barnabywalters ChiefRA: bye!
# ChiefRA ok, thanks :)
# barnabywalters well the first thing which needs to happen is for a h-listing mf2 vocab to be created
# ChiefRA I can do that.
# ChiefRA I can submit it on the proper page.
# ChiefRA thanks, I submitted a ticket to get my Github password back.
# ChiefRA :)))))))))
# barnabywalters okay I vastly improved the php-mf2 testing interface as the first steps towards making it into a syntatic validator: http://waterpigs.co.uk/php-mf2/
# barnabywalters at the very least, Codemirror does basic HTML syntax checking
pfefferle and pfefferle_ joined the channel
# ChiefRA back
# tobyink front
# ChiefRA aside :)
pfefferle_, adactio, philipashlock, gRegor`, TallTed, trendynick_ and pfefferle joined the channel
tantek joined the channel
jschweinsberg, chiui and uf-wiki-visitor joined the channel
# barnabywalters morning tantek — no problem
# barnabywalters tommorris: Re: GDATA that’s hilarious
# barnabywalters wow, I had never heard of “ROR”: http://www.rorweb.com/
# tantek tommorris: perhaps just make a /wiki/MediaRSS page and then link all the references on other pages to that?
# tantek and then move all the archive / dead links / references to the /wiki/MediaRSS page?
jschweinsberg and ivc joined the channel
# tommorris anyone remember *this*? http://microformats.org/wiki/TweepML
# tommorris edited /twitter-syntax (-33) "all the old tr.im shortlinks are dead. url shorteners really are a solution looking for a problem..." (view diff)
philipashlock_ joined the channel
# tantek tommorris: did you see /wiki/tr.im ?
KartikPrabhu joined the channel
# tommorris edited /wiki-better-than-email (+759) "/* what if someone starts an edit war on an issue */ adding a discussion about Wikipedia's BRD" (view diff)
robmorrissey joined the channel
# tantek tommorris: agreed. please add to /wiki/rel-tag-issues - feel free to cite this in the logs too
jschwein` joined the channel
# tommorris edited /rel-tag-issues (+1388) "added issue: Formal deprecation of object-level rel-tag usage" (view diff)
KartikPrabhu joined the channel
# KartikPrabhu tommorris: agreed with the rel-tag scoping issue. Can't decide if that means rel-tag is completely useless
# KartikPrabhu yes but when would anyone do that?
# KartikPrabhu specially with tag...
# KartikPrabhu why use both rel-tag and u-category?
# KartikPrabhu true
# KartikPrabhu will have to check my usage
# KartikPrabhu it is more work for author if you say use rel-tag on permalink but not in feed page
hadleybeeman and jschweinsberg joined the channel
# tommorris hadleybeeman: http://microformats.org/wiki/PDF mentions gov.uk's standardisation of PDF. ;)
# hadleybeeman If you want to be pedantic, tommorris, the wording you want isn't "recognised by the UK government's policy paper;" take it from "PDF/A must be used for static versions of documents produced for download and archiving that are not intended for editing, in particular PDF/A-1 and PDF/A-2."
# hadleybeeman But, overall — glad we could help. :)
gRegor`, KevinMarks and krendil joined the channel
# tobyink ANSI INCITS 30-1997 (R2008)?
philipashlock and tantek joined the channel
# @prettvodd RT @cishetboy: THANKPETE IM LAUHGING OS HCARD CHARLOTTe (twitter.com/_/status/492082338629943296)
KartikPrabhu and tantek joined the channel