#microformats 2015-06-01

2015-06-01 UTC
KevinMarks_, fuzzyhorns, KevinMarks, ivc\zz and netweb joined the channel
#
tantek edited /link-preview-brainstorming (+106) "/* proposals */ expand proposal to include specific other types of body objects" (view diff)
KevinMarks, KevinMarks2, kez, Left_Turn, Erkan_Yilmaz, glennjones, KartikPrabhu, eschnou, krijnhoetmerbot, frontwards, adactio, Mark87_, benborges and netweb joined the channel
#
glennjones edited /microformats2-parsing-brainstorming (+814) "/* feedback on more rel info */" (view diff)
#
glennjones tantek: KevinMarks: Adding a couple comments to rel-urls brainstroming wiki page, about inconsistent structure and a new property for rel=tag which is the last path segment
eschnou and TallTed joined the channel
#
glennjones edited /microformats2-parsing (-1) "/* corrected rel parse examples */" (view diff)
fuzzyhorns, fuzzyhor_, glennjones and benborges joined the channel
#
tantek glennjones: taking a look at the rel-urls stuff now
#
tantek !tell glennjones no need to return last path segment of the URL, because the URL is already there - and that's just a library/framework utility function to get the last path segment of a URL
#
Loqi Ok, I'll tell them that when I see them next
#
tantek the other questions didn't really make sense
#
tantek and thanks for the editorial fix to the parsing spec!
gRegorLove joined the channel
#
glennjones tantek: the logic for rel-urls seems to read that all rels are added so http://testrunner-47055.onmodulus.net/run/microformats-node/microformats-v2/rel/rel-urls/ is correct with 5 items in rel-urls and http://testrunner-47055.onmodulus.net/run/mf2py/microformats-v2/rel/rel-urls/ is wrong with 4 items. Have I understood the logic correctly?
#
Loqi glennjones: tantek left you a message 13 minutes ago: no need to return last path segment of the URL, because the URL is already there - and that's just a library/framework utility function to get the last path segment of a URL
#
tantek glennjones: I'm having trouble seeing the differences in those two long pages
#
tantek oh I see it now - the mf2py parser forgot to put the "alternate" into the rel-urls collection
#
tantek yes - 5 looks correct - looks like a bug in the mf2py parser implementation of rel-urls
#
tantek might want to ping KevinMarks since that's his code I'm pretty sure
#
@jgarber @helenvholmes If you don’t want to mess with RSS, you can use microformats to mark up your pages. http://microformats.org/wiki/microformats2 (twitter.com/_/status/605423462837047296)
#
@jgarber @helenvholmes Using some common class names, you can add semantics to your content. For instance, a blog post: http://microformats.org/wiki/h-entry (twitter.com/_/status/605423610736574464)
#
glennjones Yes, I was just wonder if mf2py was correct, but the way I read the logic on the parsing page it should include the "alternate"
#
@jgarber @helenvholmes Yeah, microformats are great. Low barrier to entry (it’s HTML!) and very powerful. I’m happy to walk you through any of it. (twitter.com/_/status/605424055999705088)
#
tantek jgarber++
#
Loqi jgarber has 2 karma
KevinMarks_ and KartikPrabhu joined the channel
#
@pvestatesca Solano Land Trust has great nature events year-round. Learn more here: http://www.solanolandtrust.org/CalendarList.aspx?hCalendar=1 (twitter.com/_/status/605433702995234816)
eschnou, KartikPrabhu, KevinMarks___ and KevinMarks_ joined the channel
#
KevinMarks I left alternates out of rel-urls because they were left out of rels
#
KevinMarks my assumption was that you'd look up the rel you were interested in, then get more info about each URL represented
#
KevinMarks I can put the redundant alternate in rel-urls, but then we should put it in rels as well
KartikPrabhu joined the channel
#
kevin marks edited /microformats2-parsing-brainstorming (+764) "/* feedback on more rel info */ rel-urls thinking" (view diff)
glennjones joined the channel
#
gRegorLove Catching up on rel-urls...
#
gRegorLove Is the main reason to not break backwards compatibility? The rel-alternates hash has array keys already, but the other rels don't
#
KevinMarks I found that looking up by rel and walking the list of urls, then getting more info if needed worked well
#
KevinMarks the multiple xfn example shows this: http://microformats.org/wiki/microformats2-parsing-brainstorming#xfn
#
gRegorLove And having rels: coworker: url: would break consumers that aren't expecting the 'url' key, but just the actual url
#
gRegorLove Hm
glennjones_ joined the channel
#
KevinMarks I did consider adding another magic thing like alternates for xfn, but then I'd need another one for enclosures and so on
#
KevinMarks rel-urls was the compromise I came up with
#
gRegorLove So rels should still be the first place to look and if you need more info, rel-urls.
#
gRegorLove With the exception of rel-alternates, which would have the extra info in both spots
fuzzyhorns joined the channel
#
KevinMarks yeah, which is why i think alternates shouldn't be in both, or if it is should eb consistent
#
@CairineGreen @AnalyseVic here are tags: Tags: fluid-layout, responsive-layout, accessibility-ready, translation-ready, microformats, rtl-language-support (twitter.com/_/status/605483250253856769)
tantek and voxpelli joined the channel
#
tantek alternates was separate only because it needed more information
#
tantek that is - for any alternate to be useful you need more information, so it didn't make sense in the simple "rels" collection
#
tantek the goal was to keep the rels collection simple
#
tantek whereas with rel-urls, the goals was to include all the information necessary for all kinds of use-cases of various rels
#
tantek thus it makes sense to include alternate in there along with all other rels
#
KevinMarks right, but as yu have written it there is no 'alternate' in rels but there is in rel-urls
#
tantek that's the point
#
tantek 'alternate' doesn't make sense in rels because all use-cases for it require more information
#
tantek whereas 'alternate' *does* make sense in rel-urls because all the information for them is provided there
#
KevinMarks but given we have alternates, why put it in rel-rls too?
#
KevinMarks s/rel-rls/rel-urls/
#
Loqi KevinMarks meant to say: but given we have alternates, why put it in rel-urls too?
#
tantek I'd be ok with dropping 'alternates' collection as I don't know of any current consuming code
#
tantek KevinMarks: because the method of look-up in rel-urls makes more sense (is easier) than the way the current alternates collection works
#
tantek I would expect anyone actually looking for alternates to now use the rel-urls map
#
KevinMarks except that I see rels as provdidng the keys and rel-urls the details
#
KevinMarks no, because they'd have to walk it
#
tantek right, so if we drop 'alternates' then we must add them, even as just urls, to rels
#
tantek point being, for now, it makes sense to have the alternates in rel-urls
#
tantek it's separate issue to consider dropping alternates and along with that change, adding alternate urls to rels
#
KevinMarks if ti does, then we should have alternate in rels too
#
tantek that would make sense if we dropped 'alternates'
#
tantek otherwise it doesn't make sense
#
tantek because anyone actually looking for a specific alternate (which all use-cases do) would use the alternates collection currently
#
KevinMarks I'm saying it doesn't make sense to have a url in rel-urls that isn't in rels
#
tantek but that's more work for everyone, so it doesn't make sense
#
tantek it's easier to just put all the rel urls into rel-urls, than have to explain why they are not all in there
#
tantek less code too
#
tantek principle of least surprise etc.
#
tantek !tell aaronpk does your indie reader use mf2 parsing to look for alternates?
#
Loqi Ok, I'll tell them that when I see them next
#
KevinMarks it doesn't make sense to put the alternate urls in rel-urls without putting them un rels
twisted` joined the channel
#
tantek of course it does. rel-urls is new, and thus it makes sense to get it right.
#
tantek it's less code to put all the rel urls into rel-urls
#
tantek those two are sufficient reasons to do so
#
tantek to put rel alternate urls into rel-urls, regardless of any other conditions
#
KevinMarks I've written the code, you haven't. its the same amount, it just depends where the else goes
#
tantek well I wrote the pseudo-code in the spec for it ;)
#
tantek seemed simpler
#
kevin marks edited /microformats2-parsing (+0) "/* parse a hyperlink element for rel microformats */ alternates not in rel-urls" (view diff)
#
KevinMarks theres the diff
#
tantek what do other folks think? shall we include alternates in the "rels" collection and drop "alternates"?
#
KevinMarks it is actually simpler to make sense of as the alternates case is truly separate
#
tantek KevinMarks: I think you screwed up a list nesting level :P
#
KevinMarks oh yes
#
tantek but as glennjones was pointing out, it makes the parsed output more confusing
#
tantek which is why I'm now advocating dropping "alternates" as the path forward
#
tantek simplifying while we can
andicascadesf joined the channel
#
kevin marks edited /microformats2-parsing (+8) "/* parse a hyperlink element for rel microformats */ fix nesting" (view diff)
#
KevinMarks there's no command-] for wikicode
#
tantek I'd rather go for simplifying KevinMarks - based on glennjones feedback
#
KevinMarks I'm happy with that too
#
tantek alright, I'm going to revert those change then - as I'd rather have the intended rel-urls functionality be in the spec
#
tantek s/change/changes
#
Loqi tantek meant to say: alright, I'm going to revert those changes then - as I'd rather have the intended rel-urls functionality be in the spec
#
tantek edited /microformats2-parsing (-8) "Reverted edits by [[Special:Contributions/Kevin Marks|Kevin Marks]] ([[User talk:Kevin Marks|Talk]]) to last version by [[User:GlennJones|GlennJones]]" (view diff)
#
tantek edited /microformats2-parsing (+229) "add warnings about dropping alternates, use rel-urls instead" (view diff)
#
tantek KevinMarks: ok I've captured the intended changes - take a look: http://microformats.org/wiki/microformats2-parsing#parse_a_hyperlink_element_for_rel_microformats
#
tantek edited /microformats2-parsing-brainstorming (+1020) "/* feedback on more rel info */ follow-up on glennjones questions" (view diff)
#
tantek edited /microformats2-parsing-brainstorming (+98) "for issues, use -issues" (view diff)
#
tantek KevinMarks while I have you here, do you have any opinion on the updated resolution to the nested h-* object value property brainstorm? http://microformats.org/wiki/microformats2-parsing-brainstorming#Nested_h-.2A_objects.27_.22value.22_property
#
tantek see http://microformats.org/wiki/microformats2-parsing-brainstorming##Update%202015-05-29 in particular
KartikPrabhu joined the channel
#
KevinMarks lots to read there
#
tantek KevinMarks: no no - you're familiar with this - just check the fragmention I pasted ^^
#
KevinMarks I meant the previous changes
#
tantek which previous changes?
#
KevinMarks the parsing ones - i'd stepped away
#
tantek for the rels stuff I just added warnings where the algorithm will change
#
tantek no actual change to the algorithm yet
#
KevinMarks you should add the 'alternate' key to rels though
#
Loqi yea
#
KevinMarks as currently you can't look them up with that algorithm
#
KevinMarks you'd want the rels for teh exmaple to be "rels": { "author": [ "http://example.com/a", "http://example.com/b" ], "in-reply-to": [ "http://example.com/1", "http://example.com/2" ],"alternate":["http://example.com/fr"] },
#
KevinMarks that nested h-object issue looks good, and the implied name rule means that the value will match the name rule fro whitespace, whatever the resolution, which is good as that was a point of contention
#
KevinMarks as the value's also get newline cruft at the moment
#
tantek ah right of course add 'alternate' key
#
KevinMarks which makes the code simpler
#
tantek yes
KartikPrabhu joined the channel
#
KevinMarks should I make that edit?
#
tantek nope - adding the whole alternates removal thing as an issue
#
KevinMarks as written the pesudocode is inconsistent between rels and rel-urls
#
tantek no - as written rel-urls is complete, and adding to rels is a separate issue
#
tantek will add a note about the alternates in the list
#
tantek nope, double-checked and existing warnings about pending changes cover the case of including "alternate" in rels
#
tantek because it's not special-cased anymore, it gets added along with everything else
#
tantek http://microformats.org/wiki/microformats2-parsing#parse_a_hyperlink_element_for_rel_microformats is fine with existing warning notes
#
KevinMarks no it isn't
#
KevinMarks you can't use alternate in rel-urls
#
gRegorLove tries to catch up on the rel-urls/alternates stuff
#
kevin marks edited /microformats2-parsing (+89) "/* rel parse examples */ put alternate keys in rels to match rel-urls" (view diff)
#
KevinMarks to use alternate you need the keys I just added to the example
#
KevinMarks which emans we should drop the conditional now
#
tantek you do not need that KevinMarks - you can simply iterate through rel-urls
#
aaronpk can't keep up, wake me when it's over
#
Loqi aaronpk: tantek left you a message 1 hour, 24 minutes ago: does your indie reader use mf2 parsing to look for alternates?
#
tantek KevinMarks: but you're right that the example had that omission - will add that in a warning note
#
kevin marks edited /microformats2-parsing (-96) "/* parse a hyperlink element for rel microformats */ simplify so alternate is consistent" (view diff)
#
tantek kevinmarks please stop - trying to get broader discussion before actually making the changes to the spec
#
tantek that's the point of putting in the warnings first
#
tantek edited /microformats2-parsing (+7) "Reverted edits by [[Special:Contributions/Kevin Marks|Kevin Marks]] ([[User talk:Kevin Marks|Talk]]) to last version by [[User:Tantek|Tantek]]" (view diff)
#
KevinMarks it makes no sense to iterate the rel-urls looking for the alternates
#
KevinMarks that is worse than what we have
#
tantek no, implementations were already doing that
#
KevinMarks no they weren't
#
gRegorLove kylewm said Woodwind looks in both rel + rel-urls
#
KevinMarks so http://microformats.org/wiki/index.php?title=microformats2-parsing&oldid=65021 is my suggestion
#
KevinMarks (I wasn't tryign to have an edit war, but to have something to point to
#
tantek you can point to the warnings in the text
#
tantek and yes that's why I was writing up an issue
#
tantek since it's a *change* to how things are spec and implemented
#
tantek instead of just "changing the spec"
#
KevinMarks the way you specced rel-urls was inconsistent
#
tantek no I spec'd it to be complete
#
tantek it actually *does* work as spec'd. and yes there can be improvements. hence another issue.
#
tantek edited /microformats2-parsing (+271) "note updated rels collection in parsed output assuming we drop "alternates", remove barnabywalter's out of date gist (spec has evolved since)" (view diff)
#
tantek thanks for pointing out that the example did not document the change that would occur
#
tantek I believe I just recaptured that in that edit
#
KevinMarks have rels and rel-urls be parallel is key. Your spec text puts them in an inconsistent stare
#
KevinMarks s/stare/state/
#
Loqi KevinMarks meant to say: have rels and rel-urls be parallel is key. Your spec text puts them in an inconsistent state
#
tantek having rels expanded is better, but it's not "key"
#
KevinMarks which requires the workaround fo walking the rel-urls again checking their rels
#
KevinMarks my implrmentation omits the alternate fromt he rel-uls to presevre this consistency: http://www.unmung.com/?html=%3Ca+rel%3D%22author%22+href%3D%22http%3A%2F%2Fexample.com%2Fa%22%3Eauthor+a%3C%2Fa%3E%0D%0A%3Ca+rel%3D%22author%22+href%3D%22http%3A%2F%2Fexample.com%2Fb%22%3Eauthor+b%3C%2Fa%3E%0D%0A%3Ca+rel%3D%22in-reply-to%22+href%3D%22http%3A%2F%2Fexample.com%2F1%22%3Epost+1%3C%2Fa%3E%0D%0A%3Ca+rel%3D%22in-reply-to%22+href%3D
#
KevinMarks that's this algorithm http://microformats.org/wiki/index.php?title=microformats2-parsing&oldid=65014#parse_a_hyperlink_element_for_rel_microformats
#
tantek KevinMarks: not as documented in the brainstorm which we discussed and agreed, which includes alternates in rel-urls
#
tantek so that's what's spec'd since that's what we agreed on: microformats2-parsing-brainstorming#more_information_for_rel-based_formats
#
KevinMarks hm, my "This generalises to other rel's too, such as rel-feed and rel-alternate that have type, lang etc attributes." was ambiguous there
#
tantek edited /microformats2-parsing-issues (+491) "drop alternates collection and include them in rels" (view diff)
#
tantek KevinMarks_: not ambiguous at all - by explicitly listing rel-alternate in that sentence the intent of inclusion in rel-urls was very clear
#
KevinMarks but I in no way thought that the rels and rel-urls would not match
#
KevinMarks your omitting them from rels is the point of difference
#
tantek anyway, comment / +1 etc. here now: http://microformats.org/wiki/microformats2-parsing-issues#drop_alternates_collection_and_include_them_in_rels
#
tantek that wasnt in your brainstorm so no
#
tantek I was doing quite literal brainstorm application
#
tantek since we discussed and agreed on what was documente
#
tantek d
#
KevinMarks you're saying that your theoretical implementation of an ambiguous brainstorm trumps my actual implmentation of it, that is lined to from it?
#
tantek this is why implementations != specs, but rather implementations help double-check specs
#
kevin marks edited /microformats2-parsing-brainstorming (+156) "/* feedback on more rel info */ link to algorithm" (view diff)
fuzzyhorns joined the channel
#
tantek also why I specifically asked for it to be written up on the brainstorming page, rather than just "code is the spec" in an implementation