#indiewebcamp 2016-05-27

2016-05-27 UTC
#
aaronpk I thought the tests repo was just the data
#
willnorris I'm also not wild about the fact that there are now tests for at least some proposed (but not yet accepted) changes to mf2. The recent one I ran across was using video.poster as an implied u- property
#
gRegorLove I'm pretty sure the tests repo is the output of glennjones parser (node? mf-shiv? same thing? unclear to me)
#
willnorris a library adding support for these proposed changes is one thing, but I don't think they should be captured in microformats/tests until they're formally accepted (whatever that process looks like)
#
gRegorLove What is microformats-shiv?
#
Loqi It looks like we don't have a page for "microformats-shiv" yet. Would you like to create it? https://indiewebcamp.com/s/10M8
#
gRegorLove microformats-shiv is http://microformatshiv.com/
#
loqi.me created /microformats-shiv (+75) "prompted by gRegorLove https://indiewebcamp.com/irc/2016-05-26/line/1464307287228 and dfn added by gRegorLove" (view diff)
#
Loqi ok
#
willnorris aaronpk: tests repo is indeed just the data. The question/concern comes from what the expected values (captured in the *.json files) are.
#
gregorlove.com moved /microformats-shiv to /microformat-shiv "canonical name"
#
aaronpk oh, I thought the .json files represented the canonical expected output of any parser
#
willnorris yes, that's true, but how do you decide what that canonical expected output should be?
#
willnorris today, it's basically whatever the nodejs or shiv implementation outputs
#
gRegorLove Based on inspecting one test, I'm pretty sure the json is from MF Node parser
#
willnorris the flip side of all this of course, is that unless you actually have multiple implementations regularly running the test suite, the current situation is the only outcome possible
#
aaronpk i thought the mf2 parsing algorithm described a canonical output
#
gregorlove.com edited /microformat-shiv (+97) "dfn, link" (view diff)
#
willnorris See https://github.com/microformats/tests/pull/48 for examples of what I mean where the tests repo does not actually match the spec (or the behavior of the other common libraries)
#
aaronpk sounds like a bug in the tests then?
#
willnorris doing value-class-pattern on datetime values is also a bit all over the place. Part of that is from some of the -issues stuff not getting integrated into the core spec pages, which tantek has since fixed
#
willnorris well, "all over the place" from my limited perspective. It may actually be completely fine, but the tests were actually ahead of the spec for a while
#
willnorris yes, that one was definitely a bug in the tests
#
gregorlove.com created /microformat_node (+194) "stub" (view diff)
[kylewm] joined the channel
#
[kylewm] can confirm, there are bugs in the test suite
tantek joined the channel
#
gRegorLove But at least we've decided on one suite to fix the bugs in? Heh
#
willnorris gRegorLove: definitely true :)
#
gRegorLove microformat node "is the same codebase as microformat-shiv project, but used the ultra-fast HTML DOM cheerio."
#
tantek genuinely glad to have you around these parts again willnorris
#
willnorris tantek: thanks. happy to be back (for a time, anyway :) )
#
tantek willnorris - re: formally accepted mf2 parser issues, I look for two things, rough consensus among the discussion of an issue, and then at least one parser implementation of it (i.e. someone demonstrates the resolution is implementable)
#
tantek that being said, I very well may be behind on incorporating some resolved mf2 parser issue into the spec
#
tantek when you find specific instances of that, please raise them to my attention. really appreciated.
#
willnorris tantek: that sounds completely reasonable. My concern is more when things hat do not yet have consensus or are not yet integrated into the spec make it into the test suite. Then other implementations have to added support for what may be experimental features if they want to continue passing the test suite. It's basically the vendor prefix problem
#
tantek yikes hopefully not as bad as vendor prefix problem
#
willnorris will do. I've still got a bit more work to do to get the go library fully passing, so may find more of these
#
willnorris (no probably not, but I knew that would make my point)
#
tantek I believe glennjones has been updating the test suite in sync with his microformats node implementation which means when he implements a resolved issue, there is a time when the test suite may be ahead of the spec
#
willnorris and that's my concern
#
willnorris if these were tests that were just for the node implementation, then I wouldn't care. But they're presented as THE test suite which all implemtnations should seek to pass
#
willnorris it's unclear to someone like me who comes along and just sees that the tests don't match the spec. it's very difficult for me to know if that's because the tests are wrong, or because the spec is just lagging a little bit
#
tantek they were just the node implementation, but since they are so far ahead of any other microformats test suite attempt, they are becoming the defacto canonical test suite
#
willnorris ah, interesting. I didn't know that histry
#
tantek willnorris: understood about the unclear / difficult to know aspect
#
tantek we had similar challenges with CSS1 and the CSS1 Test Suite and browsers :)
#
tantek this is part of the reality of polishing a standard
#
tantek there is some time lag between such different piees of the puzzle
#
tantek s/piees/pieces
#
Loqi tantek meant to say: there is some time lag between such different pieces of the puzzle
#
willnorris yeah, understood
#
tantek and that time lag is preferable to strict bureaucracy to keep everything in sync which would actually slow *everything* down, and thus ironically contribute to more legacy problems
#
tantek I think in "data" circles this is called "eventual consistency" or something?
#
willnorris fair point. And I certainly don't mean to sound ungrateful. The fact that this test suite exists *at all* is amazing, and way better than not having anything
#
tantek yup - no ungratefulness taken :) the questions you have a perfectly reasonable
#
tantek s/have a/have are
#
Loqi tantek meant to say: yup - no ungratefulness taken :) the questions you have are perfectly reasonable
#
loqi.me created /happening (+17) "prompted by aaronpk and dfn added by tantek" (view diff)
arlen joined the channel
#
tantek considering adding HWC, in particular a link to /next-hwc to https://en.wikipedia.org/wiki/Wikipedia:Meetup/San_Francisco#Upcoming (since we do encourage wiki editing/writing during quiet writing hour) - thoughts?
[kevinmarks] joined the channel
#
[kevinmarks] Sounds like we should iterate on the shared tests again
#
tantek there is no again, the iteration should be continuous
#
tantek Kevinmarks, your thoughts on ^^^ (link to /next-hwc) ?
#
[kevinmarks] Yes, I mean that Glenn built the "parse this with all the things and compare" tool before
#
[kevinmarks] And we should do that again and make it a ci thing if we can
#
tantek oh right, what willnorris is now building via all the JSONP support
#
tantek alright, adding /next-hwc - we'll see if it sticks!
#
aaronpk here's an example of a receiver test for webmention.rocks where i'm not sure the best way to test for this: "Verifies that target is a valid resource for which the receiver accepts Webmentions"
#
tantek you test the opposite
#
aaronpk how do I know what resources the receiver accepts webmentions for? I could assume that no legitimate receiver would accept webmentions for example.com but that isn't necessarily true
#
tantek that is, you send them a target of their domain with some total and utter nonsense
#
[kevinmarks] If we can make the parsers update with warnings when tests change
#
aaronpk i might want to accept webmentions of URLs on my domain that don't necessarily correspond to a file or post on my site though, just to see what people are linking to
#
[kevinmarks] We do have wildcard receivers
#
aaronpk as a way to catch my own mistakes for example, if i forget to properly handle redirects when i move files
#
tantek not quite, we have receivers that will accept on behalf of a particular set of domains
singpolyma joined the channel
#
[kevinmarks] Mention-tech will receive anything
#
aaronpk webmention.io also accepts anything
#
tantek aaronpk: one way would be to try a URL that will likely return a 404, then assuming it does, use that as the target
#
[kevinmarks] I think webmention.herokuapp.com might too
#
aaronpk nothing says you shouldn't accept webmentions for URLs that return 404
#
GWG How do you display a 404?
#
tantek aaronpk, you may also be running into a need to explicitly specify the difference in implementation class(es) between "receivers" and "proxy receivers"
#
tantek GWG, lots of debate on that one, up to the site
#
aaronpk that's a good point, that might make this more obvious
#
tantek you can start by opening an issue on it
mblaney joined the channel
#
KartikPrabhu GWG: see https://kartikprabhu.com/404
#
tantek differentiate implementation class of "receiver" and "proxy receiver"
#
tantek and then start capturing sections/quotes of spec text that may apply differently to each
#
tantek e.g. "Verifies that target is a valid resource for which the receiver accepts Webmentions" applies to receivers but not proxy receivers
#
tantek there may be other such differences
#
tantek and for now, since the spec doesn't specify "proxy receiver" as an implementation class, we acknowledge that limitation and document it
#
aaronpk so, which side should the webmention.rocks test take?
#
tantek the existing defined implementation class, "receiver"
#
tantek so the issue becomes
#
tantek Define new implementation class "proxy receiver" and how it differs in requirements from "receiver"
#
aaronpk it currently leaves it open to proxy receivers
#
tantek it currently doesn't test receivers
#
aaronpk the spec i mean
#
tantek no it does not, per the quoted text
#
tantek "Verifies that target is a valid resource for which the receiver accepts Webmentions"
#
tantek that's the whole reason we're having this discussion
#
aaronpk right. that means the receiver decides what resources it accepts.
#
aaronpk webmention.io decides to accept all resources
#
tantek proxy receivers have no way of complying with "Verifies that target is a valid resource for which the receiver accepts Webmentions"
#
aaronpk there's also this in the spec: "A Webmention Receiver is an implementation that receives Webmentions to one or more target URLs on which the Receiver's Webmention endpoint is advertised."
#
aaronpk which is a little better defined
#
tantek right
#
aaronpk but there's no way to make an automated test to check that
#
aaronpk or rather to check the inverse
#
tantek the negative of the inverse is not required by the spec
#
tantek that's the point
#
aaronpk is beginning to think that tonight's task is going to be to simply document all the tests required rather than implement any of them
#
tantek i.e. the spec doesn't say "A Webmention Receiver is an implementation that MUST reject Webmentions to any target URLs on which the Receiver's Webmention endpoint is NOT advertised."
#
aaronpk it actually does: "The receiver SHOULD check that target is a valid resource for which it can accept Webmentions."
#
aaronpk i guess that isn't the same
#
tantek right
#
aaronpk but there's no way to test it except testing the inverse
#
aaronpk the point of the implementation report is to have a checkbox for every part of the spec that means something. so i have a checkbox for that sentence. but i can't make a test to test that sentence without testing the inverse of it.
#
tantek right.
#
aaronpk so, no test then?
#
tantek it's worth capturing as an issue
#
tantek there is a spec assertion for which we need a test
#
aaronpk what's the issue?
#
tantek https://github.com/aaronpk/webmention.rocks/issues/10
#
tantek does that help?
#
aaronpk yes, I am making issues for the other checkboxes as well
#
aaronpk https://github.com/aaronpk/webmention.rocks/issues/9
#
tantek great!
#
tantek github issues as to do list :)
#
aaronpk the next interesting one is "it must perform an HTTP GET request on source, and follow any HTTP redirects (up to a self-imposed limit such as 20)"
#
aaronpk the self-imposed limit part
#
aaronpk one way to handle this is to first have someone describe their implementation features like this, then test against that
#
bear why a GET - does HEAD not satisfy the requirement?
#
tantek bear++
#
Loqi bear has 137 karma
#
aaronpk GET is required to find the link in the HTML
#
GWG aaronpk: Didn't I ask about the HEAD as a amplification attack prevention measure?
#
tantek is that after the SHOULD do an HTTP HEAD first?
#
aaronpk ah
#
bear yea, I thought it was HEAD first, validate that it's ok, GET to retrieve html and process
#
aaronpk oh i see what happened
#
aaronpk it used to not mention following redirects
#
aaronpk we added that, but didn't clarify the HEAD/GET issue
#
bear IMO setting a limit of 20 is unnecessary
#
bear (for the redirects)
#
aaronpk you gotta pick some limit tho
#
bear unless we survey all http handling libraries in all of the languages to find out what the average is ...
#
aaronpk some of my own URLs have 2-3 redirects because of the way I deal with changing my permalinks
#
aaronpk and if you don't set a limit you can end up in an infinite loop
#
aaronpk or just a DoS situation
#
bear and that is an implementation issue
#
bear MUST follow redirects; SHOULD limit number
#
aaronpk which is why the spec says "up to a self-imposed limit such as 20"
#
bear right, i'm just saying that *20* isn't needed IMO, just say "SHOULD limit the number"
#
aaronpk ah
#
bear but i'm not a spec writer so I'm not going to press on this
#
aaronpk well the good news is that sounds like an editorial change, so that's easier :-)
#
bear the limit can be mentioned in the discussion of ddos mitigations
#
tantek yes
wolftune joined the channel
#
aaronpk captured https://github.com/aaronpk/webmention/issues/45
#
aaronpk okay the HEAD-first thing is a separate issue
#
bear HEAD first can be in the security section
#
aaronpk i agree it's kind of misleading to have the text say " it must perform an HTTP GET request on source" when it actually is fine to do HEAD first
#
bear the point for me is this -- since the purpose of the request is to verify that the resource exists, HEAD is the proper call
#
bear later in the process when the HTML is required, then a GET can be performed
#
tantek completely agreed with bear
#
tantek that's worth fixing in the spec
#
aaronpk agreed
#
aaronpk is this editorial as well?
#
tantek no, that's definitely a normative requirement change
#
tantek a fix to comply with what was intended and what implementations are doing (checking HEAD before GET)
#
aaronpk it's suggesting a possible improvement to handling, but not all implementations are checking HEAD first
#
tantek they should
#
bear I really want to add my implementation but man, the itchy feeling I get approaching this WG makes me pause
#
tantek it's much kinder on the target server
#
GWG I set mine up to...but I'm still waiting for the merge.
#
tantek bear, implementers don't have to be part of the WG
#
bear IMO any implementation not doing a HEAD is wrong (meant to be strong wording)
#
tantek anyone can submit an implementation report
#
tantek bear, I agree with you
#
tantek implementations SHOULD do a HEAD request first
#
tantek that leaves an out for someone not doing a HEAD request, but only if they provide a *very good reason*
#
tantek can still test SHOULDs in the test suite too
#
bear really, if the spec says HEAD and someone does a GET, then they are just doing it badly but it's not the end of the world
#
bear but if the spec says GET that means I can't really do a HEAD
#
tantek hence SHOULD
#
tantek exactly
#
aaronpk anyone want to write this up? https://github.com/aaronpk/webmention/issues/new
#
bear I'll take a stab
singpolyma joined the channel
#
tantek bear++
#
Loqi bear has 138 karma
#
aaronpk bear++
#
Loqi bear has 139 karma
#
bear what is the section that references this - don't have links to the spec handy
#
aaronpk https://www.w3.org/TR/webmention/#webmention-verification
#
Loqi [Aaron Parecki] Webmention
#
bear ta
#
mblaney tantek: regarding fragment identifiers in urls to fetch specific comments, that technique would work for multiple h-feeds on a page too.
#
tantek indeed
#
mblaney I'm guessing that's something we need all the mf2 parser libraries to add?
#
tantek I think it's a very useful nice to have
#
bear issue https://github.com/aaronpk/webmention/issues/46 filed
#
tantek.com edited /events/2016-06-01-homebrew-website-club (+747) "featured photo and summary first so people get an idea sooner" (view diff)
#
tantek updated with featured photo: https://indiewebcamp.com/events/2016-06-01-homebrew-website-club
#
tantek considers begging kylewm for an indie event and FB POSSE copy
#
aaronpk bear++ thanks
#
Loqi bear has 140 karma
#
bear aaronpk - your very welcome
[kylewm] joined the channel
#
[kylewm] I have a feeling there are all sorts of nasty edge cases with representative h-entry and authorship when you get into parsing from fragment identifiers
#
[kylewm] tantek++ the banner on /next-hwc looks great!
#
Loqi tantek has 292 karma
#
Loqi [indieweb] "SF Homebrew Website Club" https://kylewm.com/2016/06/jun-01-sf-homebrew-website-club
#
tantek kylewm let's find the nasty edge cases! :)
shiflett joined the channel
#
tantek !tell snarfed looks like https://snarfed.org/2016-05-22_kyle-mahan-instead-of-asking-for-clarification-or-arguing is not showing kylewm's follow-up reply: https://kylewm.com/2016/05/frequent-pattern-1-someone-tweets-unclear-insults - problem showing webmentions? (yes I know I'm not one to speak :P)
#
Loqi Ok, I'll tell them that when I see them next
#
[kylewm] tantek: I see my reply on there
#
[kylewm] it stripped the bullet points, but...
#
aaronpk me too
[kevinmarks] joined the channel
#
[kevinmarks] My SVG/png cases are good for rhis
#
[kevinmarks] As they have Link in Head
#
tantek wow I totally misread it because it wasn't 1 2 3 4 list items
#
[kevinmarks] (as receivers)
#
Loqi Just generated the first draft of this week's newsletter! http://indiewebcamp.com/this-week/2016-05-27.html I'll generate it again tomorrow.
#
kylewm.com edited /events/2016-06-01-homebrew-website-club (+88) "/* San Francisco */ URLs" (view diff)
#
aaronpk you're welcome
#
[kevinmarks] I put the headers in the get too
#
[kevinmarks] The point is that wildcard receivers and opaque get is a viable combination
#
[kylewm] tantek: you didn’t have to beg! the threat of begging was enough :wink:
#
tantek !tell snarfed nevermind apparently I parse ordered lists diferently than stream of prose
#
Loqi Ok, I'll tell them that when I see them next
#
aaronpk why do i find myself using the british spelling of words more lately?
#
bear i've worked with some many canadians and EU folk that colour is normal to me now
#
[kylewm] adding a little flavour to your life?
#
tantek aaronpk: because you were in Europe a lot recently?
#
aaronpk i just typed "behaviour"
#
bear colour flavour behaviour labour
#
aaronpk also things like organize / organise
#
aaronpk advertize / advertise
#
tantek did you switch your keyboard?
#
tantek (auto-correct?)
#
aaronpk the webmention spec uses "advertise" everywhere
#
aaronpk google says "advertize" is the correct "american" spelling
#
bear I never got used to ize for those
#
aaronpk i think advertize looks wrong
#
tantek same
#
bear yea, the z is so jarring
KartikPrabhu and mlncn joined the channel
#
Salt hey, anyone around who can help me debug my h-card? I want to get the rsvp sent but something isn't validating...
#
aaronpk sure!
#
aaronpk have you tried it on http://indiewebify.me yet?
#
Salt yeah
#
Salt but it is returning no h-card found
#
Salt could be it is looking at a cached version...
#
aaronpk it shouldn't be caching. what's the URL?
#
Salt just pm'ed you
#
aaronpk interesting. are quotes required around HTML attribute values now?
#
aaronpk I see stuff like "<div class=h-entry>" in your source
#
Salt let me check where, everything is being generated in middleman so it may be doing weird haml parsing
#
Salt aaronpk, where do you see that?
#
aaronpk if you check your URL in http://pin13.net/mf2/ you can check whether the parser sees anything at all
#
Salt well with class that is definitely off
#
Salt right, and it doesn't
#
aaronpk i just viewed source on the url you gave me
#
KartikPrabhu looking up https://www.w3.org/TR/html-markup/syntax.html#syntax-attributes
#
Salt I don't see the weird quotes you are mentioning
#
Salt except in one place, and fixing that
#
Salt will clean up and reping
#
Salt things
#
Salt thanks*
#
KartikPrabhu Salt: aaronpk means that you're attribute value is unquoted
#
KartikPrabhu class=h-entry vs class="h-entry"
#
aaronpk whoa why doesn't curl un gzip ?
#
Salt KartikPrabhu, that is incorrect
#
aaronpk Salt: do you mind if I share your URL here?
#
aaronpk (it's going to appear on 2016.indieweb.org as soon as this works anyway)
#
Salt let me fix the one bug I did catch, then yes
#
Salt and I can share :)
#
KartikPrabhu so it seems HTML allows unquoted attributes but they can not have a space in them i.e. no multiple values
#
aaronpk maybe the gzip is throwing off the parser?
#
Salt possible
#
aaronpk i thought that was handled by the http client pretty seamlessly, but...
#
Salt alright, so here is the website with hcard located on homepage, haven't tried to scan the rsvp post yet
#
Salt http://www.altsalt.net/
#
aaronpk here's what i see when i view source https://aaronparecki.com/uploads/Screen-Shot-2016-05-26-20-16-10.png
#
Salt I can attempt to upload a non gzipped version, pretty easy to redeploy
#
aaronpk notice that most of the attributes are unquoted
#
aaronpk oh, easy way to test this, one sec
#
aaronpk the parser handles unquoted attributes just fine :-) http://pin13.net/mf2/?id=20160527031649047
#
Salt ah, yeah, I do see what you mean, all of the h-card stuff does have quotes, interesting that it isn't inserted on the others
#
aaronpk this is weird tho https://aaronparecki.com/uploads/Screen-Shot-2016-05-26-20-18-18.png
#
Salt aaronpk, see, that is what I expected to see!
#
Salt woah, weird
#
Salt must be gzipped is all
#
aaronpk it is sending the "Content-Encoding: gzip" header so i'm not sure why that doesn't work
#
Salt just add a "| gunzip"
#
Salt curl is a unix tool, it doesn't do the un gzipping for you
#
aaronpk i thought i normally don't have to do that, and curl takes care of it
#
Salt *shrug*
#
Salt mac curl :P
#
Salt so yeah, odd that http://indiewebify.me still can't find the h-card
#
aaronpk and pin13.net/mf2 still can't find anything
#
aaronpk anyone with more low-level http experience have any ideas?
#
Salt I am going to try uploading a non gzipped version, my guess is it isn't unpacking
#
Salt though that would be a shame if it broke my rsvp capabilities
#
aaronpk i find it hard to believe nobody has tried using gzip on their site with mf2 parsers before
#
Salt truth
#
Salt LMFAO
#
Salt YUP
#
Salt aaronpk, working now that I disable gzip
techlifeweb joined the channel
#
aaronpk weird!
#
Salt that's bad...
#
Salt excited to talk about this at the summit :P
#
Salt so, the question becomes should I keep it ungzipped for now?
#
aaronpk you could always send your rsvp webmention and then re-enable gzip 😂
#
Salt aaronpk, kay, will fix some h-entry then do that, thanks
wolcen_ joined the channel
#
gregorlove.com edited /WebFaction (+31) "link, period" (view diff)
#
gregorlove.com edited /meetup.com (+37) "dfn, link" (view diff)
#
Salt aaronpk, know if it is best practice to include the article title inside the <article> element?
#
aaronpk that's the idea as far as html is concerned, but it doesn't matter for microformats2
#
tantek Salt, I'm not sure article markup actual makes any difference any ore
#
tantek s/ore/more
#
Loqi tantek meant to say: Salt, I'm not sure article markup actual makes any difference any more
#
tantek I used to have it and dropped it
#
Salt sure, but trying to suit both, just that the validator isn't picking up a couple of tags, pin13 does, but not the other
#
tantek it was optimistically added to the HTML5 spec years ago (along with a bunch of new semantic elements) but browsers etc. never did anything with it
#
Salt http://indiewebify.me/validate-h-entry/?url=http%3A%2F%2Fwww.altsalt.net%2Fblog%2Findieweb-summit-2016-rsvp%2F
#
Salt vs http://pin13.net/mf2/?url=http%3A%2F%2Fwww.altsalt.net%2Fblog%2Findieweb-summit-2016-rsvp%2F
#
aaronpk looks like there's two elements with h-entry on them
#
Salt sure, iirc sub h-entries are allowed
#
aaronpk "allowed" is different from whether anything will recognize it :-)
#
Salt basically I want every article to have an h-entry, whereas that post specifically has the rsvp h-entry
#
tantek sounds like too much markup
#
Salt lol
#
tantek an RSVP really should *just* be that, an RSVP
#
Salt so I don't need the sub h-entry, makes enough sense
#
tantek I think it's good to require that only the simple things work
#
Salt tantek, an rsvp is just like any other post on my site, blogs, tweets, etc, all one template
#
tantek good to discourage excessive nesting / tables / hierarchy etc.
#
tantek a permalink to a post only needs one h-entry for that post, so it should have only one
#
Salt and in this case, I am trying to include the rsvp with some post content around it
#
tantek anything nested should at most be comments etc.
#
tantek you can RSVP with a comment sure
#
Salt will see what parses when I remove that
#
tantek I foten do that
#
tantek s/foten/often
#
Loqi tantek meant to say: I often do that
#
Salt alright, well that did collapse it a bit, still isn't registering e-content or p-name
#
aaronpk "h-entry" and "e-content" can't go on the same element
#
tantek that's a common enough bug that the validator should check for it and tell people
#
aaronpk https://github.com/indieweb/indiewebify-me/issues :-)
#
tantek root + property on the same thing
#
Salt :P
#
Salt I was thinking that might be the case, but didn't see a decent way to go around it
#
tantek it makes no sense if your post is an actual article with an article name
#
tantek that belongs outside the e-content anyway
#
tantek as a p-name inside the h-entry
#
Salt p-name I know what is wrong, e-content I am still not clear on
#
Salt so should I do some sort of e-content div that wraps all of the paragraphs?
#
tantek yes
#
Salt gotcha
#
Salt yay, hit the minimum recommended
#
Loqi does a happy dance!
#
Salt sooo, anyone care to do a once over before I rsvp this up?
#
gRegorLove Sure, what's the link?
#
gRegorLove Though you can always adjust and re-send the webmention; it will update.
#
Salt http://www.altsalt.net
#
Loqi William Hale
#
gRegorLove Looks good!
#
Salt kinda hack and slashing to get this rsvp out, I have ~10 95% finished posts that I should edit and put up first, but I want to get this in incase slots fill up
#
Salt gRegorLove, thanks
#
Salt boom
#
Salt looking forward to refining this stuff and spending the hack day working on more middleman incorporation... I really should try to find the time to upgrade to v4 before doing too much hacking on it
#
Salt so I did the ti.to registration but as an RSVP, though it seems to have ticked off one of the via ti.to only instead of RSVP slots...
#
aaronpk ticked where?
#
aaronpk the tickets on ti.to are linked so the counts for both kinds of tickets will always match
#
Salt right now it says 13/12
#
Salt so maybe cron needs to run *twiddles*
#
aaronpk where are you seeing that?
#
Salt hmmm, looks like I am missing the webmention part
#
Salt https://indiewebcamp.com/2016/Guest_List
#
gRegorLove Oh, that isn't automatically updated
#
aaronpk ohh that's the "manual cron" you're waiting for ;-)
#
aaronpk maybe we shuld just drop the count from that page if it's confusing
#
gRegorLove Did you send a wm to http://2016.indieweb.org/? I'm not seeing your rsvp
#
Loqi IndieWeb Summit
#
aaronparecki.com edited /2016/Guest_List (+143) "/* Official Guest List */" (view diff)
#
Salt gRegorLove, no, rying to figure that out
#
aaronpk the microformats on your RSVP look good now tho!
#
Salt historically I am paid to make sites for other people and care about these things, but I am crud at doing it for myself :P
#
Salt I could have sworn I saw an easy "enter link here to webmention" somewhere along this evenings links
#
gRegorLove #2 on indiewebify.me you can just plug in your rsvp url to send a webmention
#
Salt there we go
#
Salt thanks
#
aaronpk also http://2016.indieweb.org/webmention.php
#
Salt now do I link my site or the post
#
aaronpk the post
#
Salt there we go
#
gRegorLove Nice typography btw, Salt
#
Salt thanks
#
aaronpk woohoo congrats!
#
Salt I keep wanting to put time into doing more
#
Salt from what I understand this is a good way to get that!
#
Salt hmm, avatar didn't link
#
Salt cest la vie, will work on this more later
#
aaronpk I don't see an avatar on http://www.altsalt.net/blog/indieweb-summit-2016-rsvp/
#
Loqi IndieWeb Summit 2016 RSVP
#
aaronpk but yes you can always change the page and send the webmention again!
#
Salt oh... does the hcard need to be on EVERY page?
#
Salt *grumbles*
#
aaronpk not exactly
#
gRegorLove Not necessarily
#
Salt I need a new avatar anyway, that one has way less hair than is currently the case :P
#
bear salt - reading the scroll back my only thought/suggestion is that content-encoding is not the same as content-type so anything processing your site for mf2 will not automatically gunzip it
#
Salt bear, so the solution?
#
Salt aaronpk, gRegorLove you say not necessarily, well right now it doesn't have name or avatar...
#
bear Content-Type=application/tar-gzip (from memory)
#
gRegorLove Salt: I think you can just link to the page that has your h-card (homepage) with u-author class.
#
aaronpk i'm looking for an example without sending you down too deep of a rabbit hole :)
#
Salt bear, so I have removed the gzip stuff I was doing, I can reup as that if you want to take a look
#
bear what is generating your site?
#
Loqi It looks like we don't have a page for "generating your site" yet. Would you like to create it? https://indiewebcamp.com/s/10MA
#
Salt gRegorLove, good idea
#
Salt bear, middleman
#
bear but sure, I can take a look
#
gRegorLove That's what aaronpk does, though he might find a simpler example: <a class="u-author" href="/"></a>
#
Salt bear, sec, will reup and fix the author issue, see if it does parse
#
aaronpk yeah i do that but my h-card is down at the bottom of my site
#
Salt gRegorLove, just adding u-author to my header link
#
aaronpk the trick is you'll need to make sure there's an "author" property *inside* the h-entry, but that only needs to link to your home page which is where your h-card is
#
Salt bah
#
Salt aaronpk, good point
#
Salt hmm
#
aaronpk that's why i have the blank link like gRegorLove pointed out
#
aaronpk it's a weird hack that tantek probably wouldn't like ;-) but i can't think of a better solution
#
bear are you using Rack::Deflater to get middleman to gzip the html?
#
bear or `activate :gzip` in your :build configure?
#
Salt active
#
Salt activate*
#
gRegorLove rel-author is another option for permalinks, Salt
#
gRegorLove That doesn't have to be in the h-entry; usually in the <head>
#
bear what do you use as your webserver then - nginx, or something else?
#
aaronpk checks if XRay implements rel-author
#
tantek really prefer to avoid such empty markup if at all possible
#
gRegorLove aaronpk: Why don't you put your name in your blank link?
#
aaronpk because i don't have anywhere on the page where i want my name to show up except in the footer which is the h-card
#
Salt gRegorLove, I have that actually
#
tantek I'm always surprised when people don't sign their posts with at least their name
#
Salt I tried to implement all of opengraph
#
aaronpk my site has my name on it more than enough already
#
Salt tantek, the website is the signature
#
gRegorLove Hm, I don't see rel-author on the rsvp permalink
#
Salt bear, AWS
#
Salt just re-upped with gzip if you want to take a look
#
bear my main feed mentions my name once, but each post has my name in the byline
#
aaronpk looks like XRay does implement rel=author so if you do that it should find it
#
tantek Salt that doesn't make any sense
#
tantek it's pretty normal authoring convention to sign what you write, visibly so
#
tantek every social media site does it for you automatically
#
tantek "site has my name on it more than enough" is irrelevant - people go to your permalinks directly from search engines
#
tantek cannot assume anyone will have the context of your "site"
#
aaronpk tantek: take a look at any of my permalinks and tell me if you think it needs more occurrences of my name ;-)
#
Salt lol
#
tantek only one should be needed is my point
#
gRegorLove Did /this-week get moved to Thursdays?
#
tantek the rest if just figuring out where to put the markup
#
tantek uh what
#
Salt tantek, interested in debating this point, but not tonight, been up since 5a
#
tantek what is this week
#
Loqi This Week in the IndieWeb is a weekly digest of activities of the IndieWebCamp community https://indiewebcamp.com/this-week
#
aaronpk gRegorLove: i made it generate the first draft on thursdays
#
tantek Salt - I appreciate, design debates can go on :)
#
Salt anyhoo
#
aaronpk tantek: i do the same thing so feel free to suggest how i can improve my markup
#
bear k, Salt - your config seems sane - all the proper headers are in place
#
gRegorLove aaronpk: Ah. It popped up in woodwind, so was confused
#
tantek sleep well and see you in the morning! congrats on getting an indie rsvp working!
#
aaronpk gRegorLove: ohh i should hide it from the feed until friday
#
Salt bear, right? so I presume the parser's fault
#
tantek tomorrow you can add yourself to /RSVP#IndieWeb_Examples :)
#
bear the pattern others are seeing is that it looks like your using middleman's minify html - which strips all unnessary chars
#
Salt tantek, yay only paying for gas to attend conferences :)
#
Salt some day I will have the funds to do more
#
Salt bear, right
#
Salt but turning off gzip fixes the parser
#
bear whose parser?
#
aaronpk at least the php one
#
bear hmm, the headers look ok but my browser is sending and accept header
#
aaronpk but what's suspicious is that curl also gives me a bunch of junk back
#
aaronpk https://aaronparecki.com/uploads/Screen-Shot-2016-05-26-22-08-28.png
#
bear did you tell curl to send an accept
#
Salt if you pipe the curl to gunzip it is legible
#
bear yea, I suspect this is because s3 is being used as a cdn
#
aaronpk i just thought curl handled unzipping stuff automatically
#
bear have curl dump what the response headers were
#
bear so we can see what s3 returned
#
aaronpk https://aaronparecki.com/uploads/Screen-Shot-2016-05-26-22-09-37.png
#
bear curl is behaving correctly
#
tantek what is accept headers
#
Loqi It looks like we don't have a page for "accept headers" yet. Would you like to create it? https://indiewebcamp.com/s/10MB
#
tantek what is response headers
#
Loqi It looks like we don't have a page for "response headers" yet. Would you like to create it? https://indiewebcamp.com/s/10MC
#
tantek what is gzip
#
Loqi It looks like we don't have a page for "gzip" yet. Would you like to create it? https://indiewebcamp.com/s/10MD
#
tantek what is unzipping
#
Loqi It looks like we don't have a page for "unzipping" yet. Would you like to create it? https://indiewebcamp.com/s/10ME
#
bear what's wrong that it's also sending back Content-Type text/html
#
Salt oooh, interesting
#
bear let me dig deeper into middleman and s3
#
Salt bear, thanks much!
wolftune joined the channel
#
bear i'll do a !tell to you when I find something
#
Salt hey wolftune, lots of people here helping get my rsvp properlike :)
#
Salt bear, sounds good, probably going to stop messing with the site for now, maybe finish editing a few of these back-posts
#
wolftune Salt: have you never done IndieWebCamp anything before?
#
Salt wolftune, nope, hadn't heard of it until you mentioned it
#
Salt I did opengraph stuff, and html5 content types, but not micro
#
wolftune oh! Well, it's superb, glad you're connected! The website is a hugely valuable store of resources/links too
#
tantek Salt, opengraph makes sense since FB consumes it and it has an impact immediately
#
Salt yeah, I am very excited about this stuff, it is basically what I built my site for
#
tantek I'm curious what motivated "html5 content types"
#
Salt plus, going to the summit, I will have time to build POSSE into middleman
#
bear salt - later when you swing back around try this: remove the .gz extension when pushing to s3 - see this https://forum.middlemanapp.com/t/gzip-prep-for-s3-deploy/1173 for why
#
Salt tantek, I have always tried to keep to strict validation and was following the spec as it was coming out
#
wolftune still needs to do all this more himself, but is still novice in the tech side, never run a server ever yet…
#
tantek Salt, yes, that used to be my methodology too
#
tantek now I'm more of a minimal tagist
#
tantek still valid of course, just using fewer and fewer tags (only those that make a difference somehow)
#
tantek no markup purely for markup's sake
#
Salt wolftune, I will be moving things off of AWS eventually, part of why I don't have analytics is no consistent server to run piwik
#
bear Salt and also, are you using s3cmd to upload to S3? if not then you may have to set the attributes on the file so that S3/CloudFront know how to return the content
#
tantek I realized that's one of the antipatterns that drives people into semweb astronomy
#
Salt :P
#
tantek I mean if that's your hobby then sure, more fun to you!
#
bear see this http://sukharevd.net/gzipping-website-in-amazon-s3-bucket.html for examples of uploading to s3 with gzip'd files
#
Salt bear, I will attempt to look at this through blurry eyes >_>
#
tantek I've just found I have too many real world itches to spend time to bother with markup for markup sake anymore
#
bear read it later - it's not going to be a quick fix :)
#
Salt tantek, yeah...I have many many things taking time
#
Salt tantek, this is the person site, it should be a hobby imho
#
Salt bear, mind if I gunzip it for now and look at this later? or were you going to keep poking?
#
bear gunzip it now and carry on
#
Salt cool, thanks for the links :)
#
bear I'm around at any time for server side ops help
#
Salt hahaha, I JUST bout uploaded all of my unfinished posts
#
bear yea, sleepy ops is bad ops
#
aaronpk I should get that framed
#
bear you think I would already have that on a tshirt
#
Salt right along with the git stash fire alarm
#
Loqi I agree
#
Salt hrm, the empty link does not appear to be parsing
#
Salt okay, going to fade to sleepy ops and check back later, g'night all
#
bear o/ Salt
#
Salt \o
Pierre-O joined the channel
#
kylewm What's the git stash fire alarm? That sounds like something I need
#
gRegorLove https://github.com/qw3rtman/git-fire ?
sandro joined the channel
#
Loqi [indieweb] "Bookmarklets for #Indieweb Favorites and Retweets" by Scott Kingery http://techlifeweb.com/bookmarklets-for-indieweb-favorites-and-retweets/
#
Loqi [indieweb] "Comment on Using WordPress for IndieWeb by Bookmarklets for #Indieweb Favorites and Retweets |" by Bookmarklets for #Indieweb Favorites and Retweets &#124; http://techlifeweb.com/using-wordpress-for-indieweb/#comment-119437
#
Loqi [indieweb] "Comment on by Scott Kingery" by Scott Kingery http://techlifeweb.com/14030-2/#comment-119437
cweiske joined the channel
#
@TechLifeWeb Bookmarklets for #Indieweb Favorites and Retweets http://techlifeweb.com/bookmarklets-for-indieweb-favorites-and-retweets/ (twitter.com/_/status/736084635273158660)
nitot, jihaisse, Pierre-O, cyberjar09 and KevinMarks joined the channel
#
colintedford aaronpk: I feel weird asking, but is there a reason you couldn't put the h-entry higher up (e.g. on <body>)?
#
colintedford & then use either your footer h-card or header home link for the author property.
cweiske, Pierre-O, mapkyca, cyberjar09, arthurspooner and dogada joined the channel
#
Loqi [indieweb] "IndieWebCamp Düsseldorf:" http://eay.li/2ta
adactio, cyberjar09 and Pierre-O joined the channel
#
@cdevroe I added a barebones display of WebMentions (pings, trackbacks) to blog posts. Example: http://cdevroe.com/2016/05/20/owning-my-words-and-photos-and-audio-bits/ via: http://cdevroe.com/statuses/3440-2/ (twitter.com/_/status/736148496252870657)
Garbee, mlncn, friedcell, Pierre-O, sergio_101, j12t, singpolyma, tvn and snarfed joined the channel
#
@calum_ryan What came through the letterbox this morning from the US #indiewebcamp (twitter.com/_/status/736197350071369728)
webuser joined the channel
#
@calum_ryan What came through the letterbox this morning from the US #indiewebcamp https://twitter.com/calum_ryan/status/736198063296024576/photo/1 (twitter.com/_/status/736198063296024576)
mlncn and j12t joined the channel
#
kylewm colintedford: I think aaronpk likes being a forcing function for us to implement all of /authorship
hs0ucy, frzn, nitot_ and wolcen_ joined the channel
#
aaronpk i guess if i moved my h-entry tag higher up, then for my list pages i'd turn it into an h-feed
[kylewm] joined the channel
#
[kylewm] ha, did the IWC t-shirts get to the UK before California??!
#
aaronpk hah wow
#
aaronpk I got the big bag of them in portland yesterday!
hs0ucy and tantek joined the channel
#
bear that is very fast, I wonder if they have a printing setup somewhere in the UK
#
aaronpk i think they do. when you set up a campaign you can choose to have it "based out of" europe
#
bear yea, I see the shipping faq mentioning that
#
ben_thatmustbeme as far as the head thing, i think a copy-pasta of the section of 3.1.1 saying you can do head first would solve the issue
#
ben_thatmustbeme woops, wrong room
#
tantek copy-pasta? this is the right place for that
#
tantek ;)
AndroUser joined the channel
#
ben_thatmustbeme no but this is the right room to mention i'm going to be working on an offline html5 MP client
#
ben_thatmustbeme service workers to cache it all
#
ben_thatmustbeme it will take some time as i don't have much free time, and i'm learning as i go
#
ben_thatmustbeme but mobilepub will be reborn as a website
#
aaronpk cool!
#
tantek at first I read that as HTML5 MP3 client
#
tantek s/HTML5/offline HTML
#
Loqi tantek meant to say: at first I read that as offline HTML MP3 client
#
aaronpk that would be...interesting
#
ben_thatmustbeme lol
AndroUser joined the channel
#
tantek thisismyjam had a mobile web app that played your friends's jams
#
tantek so there is prior art here
AndroUser joined the channel
#
tantek aaronpk: this would actually be a good thing to test in webmention.rocks - that the requester explicitly does a HEAD request FIRST, and then two subtests: one where the HEAD has the link header and then the sender sends a webmention WITHOUT doing a GET, and another where the HEAD lacks the link header, and the sender does a GET for endpoint discovery.
gRegorLove joined the channel
#
aaronpk interesting idea
#
tantek the thing is, if the test suite has it, implementations will almost certainly all do it, because test suites in practice tend to drive implementation details MUCH more than any spec prose details
#
tantek (also discovered from CSS1)
calumryan joined the channel
#
tantek (where the test suite explicitliy tested some features not documented in CSS1, everyone implemented them - interoperably! - and then the CSS1 spec had to be updated accordingly)
#
tantek (and it took like a year for folks to discover the spec/test discrepancy)
j12t, wolftune and snarfed joined the channel
#
calumryan IWC t-shirt came today - did have a customs declaration but seemed to have come from inside EU - Germany
[kevinmarks] joined the channel
#
[kevinmarks] Test driven development for specs
#
[kevinmarks] There is the appengine issue where if you don't implement HEAD explicitly you get a 5xx error
shiflett joined the channel
#
@scott_gruber IndieWebify.Me - a guide to getting you on the IndieWeb. cool page part of getting ready for @indiewebcamp http://indiewebify.me/ (twitter.com/_/status/736237195531411456)
#
bear ^^ I tend to forget how exiciting IndieWebify.me was when I started
#
aaronpk we need to do a code sprint on it
bret joined the channel
#
snarfed [kevinmarks]: re app engine and head, i believe it's a 405, not 5xx
#
Loqi snarfed: tantek left you a message 14 hours, 22 minutes ago: looks like https://snarfed.org/2016-05-22_kyle-mahan-instead-of-asking-for-clarification-or-arguing is not showing kylewm's follow-up reply: https://kylewm.com/2016/05/frequent-pattern-1-someone-tweets-unclear-insults - problem showing webmentions? (yes I know I'm not one to speak :P) http://indiewebcamp.com/irc/2016-05-26/line/1464316050084
#
Loqi snarfed: tantek left you a message 14 hours, 17 minutes ago: nevermind apparently I parse ordered lists diferently than stream of prose http://indiewebcamp.com/irc/2016-05-26/line/1464316332105
#
tantek sorry for the noise snarfed
#
tantek re webmention and head requests
#
tantek huh I didn't even consider HEAD for link verification
#
tantek but it is theoretically possible!
#
tantek you could return LINK headers for everything you claim your content links to
#
aaronpk that sounds almost as bad as a sidefile
#
tantek and that SHOULD be trustworthy enough to not have to bother doing a GET. HOWEVER, yes it is kind of side-file like
#
aaronpk also way easy to spam since it's invisible data
#
tantek and worse, it quickly runs out of steam in terms of upgrading the "mention" to be more meaningful
#
bear yea, even as a user of "sidefiles" :P -- this one has a DRY smell
#
tantek I mean you're limited to rel values
#
tantek bear I think you meant WET smell ;)
#
tantek (Write Everything Twice+)
#
bear LOL - that just made my day
#
tantek (forgot where I saw that)
#
tantek oh boy just going to brainstorm this outloud so it gets me into trouble later
#
tantek Link: <http://example.com/permalink>; rel="like-of"; title="Tantek likes this permalinkthing"
#
tantek or even more sketchy, we extend the params to any h-entry property
#
tantek e.g.
#
aaronpk runs away
#
tantek Link: <http://example.com/eventpermalink>; rel="in-reply-to"; title="Tantek is going to eventname"; rsvp="yes"
#
aaronpk why is this inivisible data a good idea?
#
tantek better to state it and debunk it than let someone else think they came up with the idea later
#
tantek "oh yeah we already thought about that ;)"
#
tantek or I'll just collect it for the inevitable "metaformats" spec
#
bear running down bad ideas is very useful for docs and specs
#
bear so much bikeshedding is prevented by being able to point people to documentation of why things are not done
#
tantek right, idea considered and rejected. do you have new information? etc.
#
aaronpk so the github thread that bear opened was *only* about HEAD requests on webmention verification
#
aaronpk not about HEAD requests on discovery
#
snarfed loves that ben_thatmustbeme is planning to implement https://snarfed.org/posting-to-the-indieweb-from-your-phone
#
snarfed ben_thatmustbeme++
#
Loqi [Ryan Barrett] Posting to the indieweb from your phone
#
Loqi ben_thatmustbeme has 146 karma
#
snarfed go ben_thatmustbeme go!
#
tantek aaronpk - right, hence my LINK header examples are for verification, not discovery
#
tantek I'm sure all the js;dr folks would love it too ;)
#
aaronpk you'd have to pack JSON data into the HTTP headers for them to like it ;-)
#
tantek haha easy, add a ;json="items:[{ ...}]"
#
aaronpk on boy
#
aaronpk h
nitot joined the channel
#
@jimpick I'm heading down to https://indiewebcamp.com/2016 in Portland. Anybody else heading there from Vancouver, BC? Thinking of doing Amtrak. (twitter.com/_/status/736243628348039169)
dogada and wolftune joined the channel
#
ben_thatmustbeme snarfed: i don't know if you saw, but at the last IWC @ MIT I made my site save offline posts made on my web client , but it doesn't cache the site
#
snarfed ben_thatmustbeme: cool!
#
tantek what is u-uid
#
Loqi It looks like we don't have a page for "u-uid" yet. Would you like to create it? https://indiewebcamp.com/s/10MF
#
tantek u-uid is a microformats2 property for representing a unique identifier for the item, e.g. an unchanging permalink, like a u-url but without any user-authored parts like a slug.
#
loqi.me created /u-uid (+203) "prompted by tantek https://indiewebcamp.com/irc/2016-05-27/line/1464370475717 and dfn added by tantek" (view diff)
#
Loqi ok
#
tantek.com edited /u-uid (+628) "why brainstorming see also" (view diff)
[scottgruber] joined the channel
#
[scottgruber] Hmm. Scratching my head whilst working to post my name and photo to http://2016.indieweb.org. I set up an h-card this morning (validates here http://indiewebify.me/validate-h-card/?url=scottgruber.me). Sent an update webmention here http://2016.indieweb.org/webmention.php but my name and photo aren’t showing up. Must have missed something or need to clean up some html. Simple stuff I’m sure, but any hints how to markup name and photo
#
[scottgruber] webmention? I’m using p-name on an anchor tag and <img class="u-photo”
#
Loqi IndieWeb Summit
#
aaronpk [scottgruber]: your post is missing an author property
#
tantek scottgruber, see here: http://pin13.net/mf2/?url=http://scottgruber.me note there is no "author" inside the "h-entry"
#
aaronpk indiewebify.me shows you're missing author too http://indiewebify.me/validate-h-entry/?url=https%3A%2F%2Fscottgruber.me%2Fnotes%2Findiewebcamp
#
tantek but really what you should be checking is this: https://scottgruber.me/notes/indiewebcamp
#
Loqi Indiewebcamp
#
tantek your RSVP permalink :)
#
tantek I'm guessing that author/footer is the same for your whole site?
#
tantek scottgruber, put the class="h-entry" on body
#
tantek e.g. <body class="h-entry">
#
tantek for your permalinks
#
tantek and then add p-author *on* your h-card directly so it links it up
#
tantek <div class="p-author h-card vcard">
#
tantek that ought to do it
#
tantek without having to add any more elements etc.
#
tantek also you can drop the p-author here: class="p-name p-author u-url fn url" to just class="p-name u-url fn url"
#
tantek also you want <time> not <date>. there's no <date> element
#
tantek <time>2016-05-23</time>
#
[scottgruber] Much thanks. Working it up now.
#
tantek lastly you probably want to change <p>RSVP: to <p class="p-name">RSVP:
#
tantek and <time class="dt-published">2016-05-23</time>
#
tantek assuming that's the date you published the RSVP
snarfed joined the channel
#
ben_thatmustbeme wow, actually, most all of this is going to be exactly the same... its just a matter of getting the login mechanism working correctly
#
[scottgruber] Magic.
#
[scottgruber] Thanks again. I’ll post my notes so I won’t forget and improve my site’s html.
#
aaronpk [scottgruber]: nicely done!
#
[scottgruber] dance hands
KartikPrabhu joined the channel
#
tantek looks great scottgruber! http://pin13.net/mf2/?url=https://scottgruber.me/notes/indiewebcamp
#
tantek scottgruber++ well done!!!
#
Loqi scottgruber has 1 karma
#
tantek if you'd like, you can also add yourself to this list: https://indiewebcamp.com/rsvp#IndieWeb_Examples
#
tantek as an implementer of RSVP posts!
#
tantek also on your home page, since it's primarily a feed, you could add <body class="home h-feed">
#
tantek scottgruber, interesting, in your RSS feed you have <title>Scott Gruber's Blog</title>, however no equivalent on your home page
#
[scottgruber] I need to add canonical urls and fix up the header html, setting up my building blocks as part of my indieweb goals.
#
tantek you could add an <h1 class="p-name">Scott Gruber's Blog</h1> on your home page
#
[scottgruber] i have a [blog](https://scottgruber.me/blog) that has the rss that I removed from my nav but I left rss in head. Is it ok to have multiple rss feeds, 1 for articles rss and one for notes rss and one for blog (if I publish it.)
#
tantek sure!
#
tantek we also encourage composite streams
#
tantek what is a composite stream?
#
Loqi A composite stream is a stream that interleaves multiple types of posts, such as notes and articles https://indiewebcamp.com/composite_stream
#
colintedford.com edited /jam (+60) "/* See Also */ ♫ "Jam On It" by Newcleus" (view diff)
#
scottgruber.me edited /rsvp (+144) "/* IndieWeb Examples */" (view diff)
j12t joined the channel
#
tantek oldest live like example appear to be snarfed's
#
snarfed wins!
#
tantek (not counting Bridgy ;) )
#
snarfed aww there are bridgy likes older than the oldest known indie like?
#
snarfed boo :P
#
KartikPrabhu bridgy++ for finding older backfeeds, more like
#
Loqi bridgy has 41 karma
#
tantek snarfed, Bridgy supposedly started sending backfeed likes as of end of 2013
#
tantek yours is from 2014
#
snarfed yup. still boo.
j12t joined the channel
#
tantek.com edited /like (+214) "/* IndieWeb Examples */ comment out sandeep and benwerd as their permalinks either 404 or lack like markup respectively" (view diff)
#
snarfed (bridgy was actually backfeeding as early as 2012, just not via mf2/wm)
#
gRegorLove gives snarfed the silver medal
#
ben_thatmustbeme darn... i was going to run my new client entirely off of github pages, but i can't login from JS because of CORS, and I can't host the php i need on github.io
#
tantek gets bridgysplained
#
snarfed heh
#
tantek bumping Bridgy down to its own category - Services
#
tantek wait is bridgy still publishing u-like? I thought I sent a pull request to fix that
#
snarfed right, i think it doesn't now
#
tantek.com edited /like (-19) "move bridgy to services section, note only publishes u-like-of markup, need an early permalink" (view diff)
#
tantek !tell barnabywalters can you dig up your first "like" post (likely in 2014? maybe 2013?) permalink and add it to https://indiewebcamp.com/like#Barnaby_Walters ? Thanks!
#
Loqi Ok, I'll tell him that when I see him next
#
tantek.com edited /like (+96) "/* Kyle Mahan */ u-like-of h-cite" (view diff)
#
snarfed heh, memories, first public bridgy commit, dec 2011: https://github.com/snarfed/bridgy/commit/cd6c6279c749f01365f6f7317309658759b1f4fb
#
snarfed (i'd been working on it since summer 2011, but in a private repo)
#
tantek !tell benwerd your early RSVP posts appear to have broken, i.e. lack p-rsvp markup now, e.g. http://werd.io/view/51ca1213bed7de1e7763d71f
#
Loqi Ok, I'll tell them that when I see them next
#
snarfed and the old cartoony logo! https://snarfed.org/2012-01-08_bridgy_launched
#
Loqi [Ryan Barrett] Bridgy up and running
#
tantek !tell aaronpk your early RSVP posts appear to have broken, i.e. lack p-rsvp markup now, e.g. https://aaronparecki.com/2013/06/25/4/indiewebcamp
#
Loqi Ok, I'll tell them that when I see them next
#
aaronpk mine too??
#
aaronpk weird
#
KartikPrabhu if posts are not written as templates how does one keep the markup up-to-date?
#
tantek working code?
#
KartikPrabhu for say a static site?
#
tantek my posts are not written as templates
#
aaronpk just that one is broken
#
tantek KartikPrabhu: there is no up to date problem
#
KartikPrabhu right so how do you keep old posts updated with new markup?
#
tantek these are all posts that *used to* have p-rsvp
#
tantek but don't have it any more
#
KartikPrabhu oh ok, different problem then :)
#
tantek KartikPrabhu: I don't understand your question
#
aaronpk fixed
#
tantek aaronpk: suspicious, why just that one?
#
tantek what was your second RSVP?
#
Loqi It looks like we don't have a page for "your second RSVP" yet. Would you like to create it? https://indiewebcamp.com/s/10MG
#
tantek uh oh
#
aaronpk https://aaronparecki.com/rsvps?before=20130926T163319-0700
#
KartikPrabhu tantek: I write articles in HTML and so had to go an update all older posts by hand when adding say "u-featured"
#
aaronpk not sure why that one didn't import right, but in yaml, "yes" is an alias for "true", and that post had "true" stored for the rsvp value instead of the string value "yes"
#
aaronpk so because true != "yes", my code didn't add the rsvp markup
#
tantek sees an advantage to "going to" in the start of the prose text of a reply meaning "rsvp:yes" ;)
#
tantek SuperDRY*
#
tantek is tempted to pick up some Superdry branded clothing for the next W3C event.
#
aaronpk the rest of the RSVPs imported correctly, so no idea why that one didn't
#
tantek.com edited /rsvp (+54) "/* IndieWeb Examples */ benwerd lost his p-rsvp" (view diff)
#
tantek and that makes bret the oldest RSVP still working
#
tantek https://indiewebcamp.com/rsvp#IndieWeb_Examples
#
tantek and with 10+ publishers, and 5+ consumers, p-rsvp is so stable it would be harder to change it than not change it, so it just got promoted to core h-entry
#
tantek we have more different documented RSVPs than likes, reposts, or photos
#
aaronpk what!
#
tantek only 9 /photo#IndieWeb_Examples
#
aaronpk are there any rsvp consumers listed on /rsvp?
#
tantek compared to what, 13 RSVPs on 2016.indieweb.org
#
tantek not counting the no photo no permalink
#
tantek aaronpk: hah - rsvp consumers are called /event posts :)
#
aaronpk hmm
#
aaronpk i would expect to see a "How to consume RSVPs" on that page to go along with "How to publish"
#
tantek true!
#
tantek we have a How to consume/show likes on the /like page
#
tantek want to stub it?
#
aaronpk also /event doens't actually have any info on how to consume rsvps ;-) https://indiewebcamp.com/event#How_to_accept_RSVPs
#
gRegorLove I publish events, but don't consume rsvps yet.
#
tantek huh
#
gRegorLove Mostly because I'm creating a log of events I've attended.
#
tantek.com edited /posts (+26) "/* Kinds of Posts */ rsvps way ahead of everything but note/article/reply" (view diff)
#
tantek I know GWG has not yet added himself to a lot of the IndieWeb Examples lists
#
tantek but he's been doing likes for a while (pretty sure)
#
tantek !tell GWG what's your oldest "like" post, i.e. both semantically (like you menat it as a like, not just a bookmark), and with u-like-of markup? can you add it to https://indiewebcamp.com/like#IndieWeb_Examples
#
Loqi Ok, I'll tell them that when I see them next
#
aaronparecki.com edited /rsvp (+401) "stub how to accept an RSVP" (view diff)
#
tantek aaronpk: re: "how to consume an rsvp" not well defined, should it be on its own? or should it be just one part of "how to interpret a valid webmention" ?
#
aaronpk both
#
aaronpk it's useful to have the reference to it on the rsvp page, but also should be part of the normal process of handling webmentions
#
tantek like two steps? 1 is this an rsvp? 2 what is the rsvp?
#
Loqi An RSVP is a type of post that is a reply to an event post https://indiewebcamp.com/rsvp
#
aaronpk i do it the other way around. first is this a valid webmention, then is this a reply, then is this an RSVP?
#
tantek is "how to interpret a valid webmention" just a subset of /post-type-discovery ?
#
tantek that's not the other way around
#
aaronpk an RSVP is an upgrade of a reply
#
tantek that's exactly what I'm talking about
#
aaronpk it sounded like you were checking if the post is a reply first
#
tantek " is this a reply, then is this an RSVP?" is just part of a larger "how to interpret a valid webmention"
#
aaronpk sure. i don't really see where this is going
#
tantek what? seriously? this is a big impact on /post-type-discovery
#
tantek as in, /post-type-discovery may have a second *simpler* algorithm specifically for webmention-type-discovery
shiflett, arlen and snarfed joined the channel
#
@camcavers RT @jimpick I'm heading down to https://indiewebcamp.com/2016 in Portland. Anybody else heading there from Vancouver, BC? Thinking of doing Amtrak. (twitter.com/_/status/736277801808977920)
#
Loqi [indieweb] "IndieWebify.Me" https://indiewebify.me/
tvn and KevinMarks joined the channel
#
KevinMarks Gillmor gang time
#
tantek Webmention is a CR is the obv big news
#
tantek hey when do we see /this-week draft?
#
aaronpk yesterday
#
tantek hmm Upcoming events doesn't show the featured image for HWC on 2016-06-01 https://indiewebcamp.com/this-week/2016-05-27.html
#
tantek really need to build a 2015 grid
#
aaronpk it should be possible to automate generating it
#
aaronpk from the h-cards on all the event pages
#
tantek and then the de-duping
#
tantek and then the manual image fixups
#
tantek actually, that's where we're at!
#
tantek https://indiewebcamp.com/2015-review#IndieWebCampers
#
tantek LOTS of manual images to fix
#
tantek e.g. from their linked site, or at worst their Twitter avatar
#
bret woop
#
bret yay for static sites
#
Loqi bret: tantek left you a message on 5/12 at 1:02pm: are you coming to the Leaders Summit? Can you RSVP here accordingly? https://indiewebcamp.com/2016/Leaders#RSVP http://indiewebcamp.com/irc/2016-05-12/line/1463083374469
#
bret i want to but I'm on the fence about organizing, i've been dealing with some personal issue lately
#
KevinMarks what is this week?
#
Loqi This Week in the IndieWeb is a weekly digest of activities of the IndieWebCamp community https://indiewebcamp.com/this-week
#
tantek What is verisignlabs?
#
Loqi It looks like we don't have a page for "verisignlabs" yet. Would you like to create it? https://indiewebcamp.com/s/10MH
#
tantek verisignlabs is a user identity silo owned by [[Symantec]] that is discontinuing on 2016-09-12. https://twitter.com/jcn/status/729066788898672640
#
loqi.me created /verisignlabs (+172) "prompted by tantek https://indiewebcamp.com/irc/2016-05-27/line/1464380136135 and dfn added by tantek" (view diff)
#
Loqi ok
#
@jcn Sigh, the final nail in the coffin that houses all of my OpenID dreams. RIP https://pip.verisignlabs.com/ [photo] (twitter.com/_/status/729066788898672640)
#
aaronpk whoa really?
#
aaronpk does this mean indieauth.com is one of the only openid providers left?
#
@kevinmarks @jcn did you know that https://indieauth.com/ will handle OpenID login for you? #indieweb (twitter.com/_/status/736290530171052032)
#
tantek I have no idea what this is, just saw @jcn's tweet and felt it was site-death worthy
#
tantek kevinmarks is on it!
#
aaronpk huh where is that documented
#
aaronpk oh that should be moved from /IndieAuth to /indieauth.com
#
tantek where is what documented?
#
aaronpk using indieauth.com as your openid provider
#
aaronparecki.com edited /indieauth.com (+568) "move OpenID instructions to indieauth.com" (view diff)
#
aaronparecki.com edited /IndieAuth () "(-518) move OpenID instructions to indieauth.com" (view diff)
tvn_ joined the channel
#
KevinMarks is it documented on indieauth,com itself?
#
aaronpk apparently not
#
aaronparecki.com edited /indieauth.com (+6) "/* OpenID */ use openid.indieauth.com to avoid this bug https://github.com/aaronpk/IndieAuth.com/issues/99" (view diff)
#
@kevinmarks @jcn more info here: http://indiewebcamp.com/indieauth.com (twitter.com/_/status/736291705129828352)
#
@hvdsomp @blahah404 @invisiblecomma @gbilder See my Webmention intervention. https://github.com/aaronpk/webmention/issues/42 (twitter.com/_/status/736292198757609472)
#
aaronpk lol!
#
@kevinmarks @hvdsomp @blahah404 @invisiblecomma @gbilder a brid.gy style proxy for webmentions of paper references would be a good approach (twitter.com/_/status/736292939437998084)
#
Loqi Just generated this week's newsletter! You still have a few minutes to make changes, and I'll re-generate it 10 minutes before it gets sent out at 2pm. http://indiewebcamp.com/this-week/2016-05-27.html
#
aaronpk oh tantek last time the newsletter included the featured image you said it should not!
#
aaronpk so it only looks for "photo" now
#
tantek no that was contextual!
#
tantek sigh
#
tantek I meant for PAST events it should not show a u-featured
#
tantek because obviously the u-featured was not a photo taken at that tevent
#
tantek and past events with photos implies very strongly that the photo was from that event
#
tantek presentationally
#
tantek but for *upcoming* events, u-featured photos make sense!
#
tantek just as they show up for upcoming events in like every silo ever
#
tantek or something :p
#
aaronpk well that was very not clear from last time
#
Loqi Just generated this week's newsletter! You still have a few minutes to make changes, and I'll re-generate it 10 minutes before it gets sent out at 2pm. http://indiewebcamp.com/this-week/2016-05-27.html
#
tantek.com edited /2015-review (-295) "/* IndieWebCampers */ fix a few (7) photos" (view diff)
#
tantek ^^^ help ? I fixed a few photos, requires manually looking up the person *somewhere* (e.g. their website, their twitter etc.) and putting in an image from there
#
tantek dang we really should have a featured image for IWS
#
tantek since this is the last newsletter that will go out before the summit!!
#
aaronpk there's still time!
#
KevinMarks get one in
#
tantek finds http://2016.indieweb.org/assets/2014-indieweb-movement.jpg :P
[shaners] joined the channel
#
[shaners] How about a gif of each IWC’s group photo?
#
aaronpk put the new logo on top of that!
#
tantek.com edited /2016/Schedule (+148) "add a u-featured" (view diff)
#
tantek puts content before style :P
#
aaronpk it has to be at https://indiewebcamp.com/2016 since that's the URL listed at /events
#
tantek shaners sounds like unvolunteered work :P
#
aaronpk it sounded like [shaners] was volunteering ;-)
#
[shaners] If someone else found all of the photos, I’ll put together the gif.
#
tantek.com edited /2016 (+148) "this u-featured" (view diff)
#
[shaners] When does the newsletter go out?
#
tantek shaners, step one, help with filling in such photos on /IndieWebCamp
#
aaronpk 2pm
#
[shaners] Ok. Nevermind. This is out of scope for right now.
#
[shaners] I’ll work on it at IWS.
#
tantek.com edited /2016 (+73) "style it like a header image" (view diff)
#
tantek how's that? https://indiewebcamp.com/2016
#
aaronpk looks good but the email will have it at full height
j12t_ joined the channel
#
Loqi Generated the final version of the newsletter! This will be sent out at 2pm. http://indiewebcamp.com/this-week/2016-05-27.html
#
tantek yay lots of images!
#
Loqi yay!
#
tantek KevinMarks: Two big things for GG: Webmention CR (27+ implementations), 6th Annual IndieWeb Summit is next Friday-Sunday
#
tantek note only 16 spots left for IndieWeb Summit!
j12t joined the channel
#
@gRegorLove Looking forward to IndieWeb Summit in Portland next weekend! There’s still 16 tickets available: http://2016.indieweb.org (twitter.com/_/status/736302429210214401)
#
@indiewebcamp This week in the #indieweb https://indiewebcamp.com/this-week/2016-05-27.html https://twitter.com/indiewebcamp/status/736304729425223680/photo/1 (twitter.com/_/status/736304729425223680)
#
@RikMende RT @indiewebcamp This week in the #indieweb https://indiewebcamp.com/this-week/2016-05-27.html https://twitter.com/indiewebcamp/status/736304729425223680/photo/1 (twitter.com/_/status/736304788334317569)
KevinMarks joined the channel
#
KevinMarks can you check I'm not consuming a spot by remote attending?
#
@kevinmarks RT @indiewebcamp This week in the #indieweb https://indiewebcamp.com/this-week/2016-05-27.html https://twitter.com/indiewebcamp/status/736304729425223680/photo/1 (twitter.com/_/status/736304865706532865)
yakker joined the channel
#
sknebel ben_thatmustbeme: re your github issue, you can generate an oauth token in your github account and then use that from client-side JS... not that nice a login flow, but works
#
Loqi [indieweb] "WeeklyUpdates/2016-05-30" by Tantek https://wiki.mozilla.org/index.php?title=WeeklyUpdates/2016-05-30&diff=1134839&oldid=1134833
#
Loqi [indieweb] "WeeklyUpdates/2016-05-16" by Tantek https://wiki.mozilla.org/index.php?title=WeeklyUpdates/2016-05-16&diff=1134838&oldid=1132861
#
@jkphl RT @indiewebcamp This week in the #indieweb https://indiewebcamp.com/this-week/2016-05-27.html https://twitter.com/indiewebcamp/status/736304729425223680/photo/1 (twitter.com/_/status/736309955582676992)
#
Loqi [indieweb] "WeeklyUpdates/2016-05-30" by Tantek https://wiki.mozilla.org/index.php?title=WeeklyUpdates/2016-05-30&diff=1134840&oldid=1134833
j12t, KartikPrabhu and KevinMarks joined the channel
#
tantek.com edited /create (-23) "/* p3k event */ fix png" (view diff)
#
tantek goes down an input type=file capture rabbithole
#
tantek discovers it's not in MDN :P
KartikPrabhu joined the channel
#
KevinMarks it is a bit of a neglected api; benwerd had trouble with it
#
tantek API? I mean just the 'capture" attribute
#
@m_ott IndieWebCamp Düsseldorf 2016 POSSEing test https://indiewebcamp.com/2016/D%C3%BCsseldorf (twitter.com/_/status/736328164801138688)
snarfed1, snarfed and [benatwork] joined the channel
#
[benatwork] it’s handled in a very inconsistent way by mobile browsers
#
tantek I'm having trouble even finding a bug on it in Bugzilla
#
tantek so I'm going to file a new one :P
#
tantek rabbit rabbit rabbit
#
tantek wow, the "check for duplicates" search when filing a big is so much better than actually searching bugzilla
#
tantek.com edited /photo_upload (+99) "/* Desktop upload */ note feature request bug for Firefox" (view diff)
finchd and wolftune joined the channel
#
tantek.com edited /create (+1759) "brainstorming / photo - some thoughts" (view diff)
wolcen_ joined the channel
#
tantek.com edited /photo_upload (+80) "see also create photo" (view diff)
snarfed1 joined the channel