#indiewebcamp 2016-02-29

2016-02-29 UTC
#
kylewm
(yeah i realized it was more appropriate to ask Google, 2 seconds after i asked here)
#
kylewm
if anyone else is curious, the 3rd answer is the best one there
#
aaronpk
so yeah I can't make my tests for XRay pass until I use the fixed mf2 parser
#
aaronpk
otherwise I either have broken HTML escaping handling, or the contents of <script> tags ends up in the plaintext
shiflett and wolftune joined the channel
#
aaronpk
Any objections to me merging gregor's PR? (We have to get barnaby to do an actual release of a point version, but ppl can use the master branch if they want in the mean time)
#
tantek
looks
#
tantek
why did the Travis build fail?
#
aaronpk
gregor mentioned that in his comment
#
aaronpk
there's one failing test in php 5.4 for resolving relative URLs
#
tantek
why are we testing usernames in URLs?!?
#
aaronpk
it's part of the relative URL resolving testing
#
tantek
wow ok
#
aaronpk
URLs are complicated
#
aaronpk
i'm tempted to change the test to allow either of those though
#
tantek
I'm not even sure which is correct per the "URL Standard"
#
aaronpk
i don't think anybody does
#
aaronpk
it's not like the URL standard changed between PHP 5.4 and 5.5
#
kylewm
php5.4--
#
kylewm
seems like they really got their act together in 5.5
#
aaronpk
yeah also 5.4 was EOL in september 2015
#
aaronpk
well now i'm confused. the contents of the script tag are still being returned
#
kylewm
still a *ton* of shared hosts on 5.3 :(
#
Loqi
aww, cheer up
#
aaronpk
yeah it'll be a while til we can actually drop 5.4 support
#
aaronpk
ohhh it removes it from p-content but not from e-content
#
aaronpk
well that's another question actually. should <div class="e-content"><script>alert("hi");</script></div> contain the whole script tag in the html value?
#
aaronpk
i think technically it should, but the plaintext value should not include the alert("hi"); text
benwerd and shiflett joined the channel
#
tantek
aaronpk that sounds like a reasonable resolution
#
tantek
I would also be ok with a blacklist of tags to strip from HTML
#
aaronpk
I think that gets into dangerous territory
#
aaronpk
like, false sense of security ish
#
tantek
not if we do it for specific reasons e.g. script / security
#
tantek
it's not false, we can use similar logic as CSP
#
aaronpk
IMO the HTML returned by the parser should not be assumed to be safe for rendering as is
#
tantek
which has special casing for script, style, iframe, object
#
tantek
the point is not to be "safe" for rendering, but rather, to block things that are known (per CSP special-casing) to often be trouble
#
tantek
CSP is still a good thing even though it is not "complete" by any measure
#
aaronpk
well I just added a failing test for the script tag behavior i described https://github.com/indieweb/php-mf2/pull/83
#
tantek
so the point is, we don't make up our own list, we go by the list of things that CSP thinks are "special" (worthy of blocking), and thus has been reviewed by all the security people that worked ont hat
#
tantek
you want <script> included or not?
benwerd joined the channel
#
aaronpk
Basically I want this: {"content":[{"html":"<b>Hello</b><script>...</script>","value":"Hello"}]}
#
tantek
so you want the <script> ?
#
aaronpk
I mean I would be fine with having it remove the script tag completely, but that's a bigger change
#
aaronpk
Right now it parses as {"content":[{"html":"<b>Hello</b><script>...</script>","value":"Hello..."}]}
#
tantek
I feel that's more in the spirit of the issue that was filed
#
aaronpk
where's that issue?
#
Loqi
microformats2-parsing-issues
#
aaronpk
well I will say that I am sanitizing the HTML returned from the parser anyway, so the first thing that happens is I remove script and style tags anyway
#
aaronpk
so it wouldn't hurt if the parser didn't include those in the first place
#
Loqi
microformats2-parsing-issues
benwerd, CaptainCalliope, j12t, davidmead, link0_, lukebrooker, snarfed and Zegnat joined the channel
#
@mathpunk
@atduskgreg @m1k3y Making an #indiewebcamp note that soundcloud feed subscription is a desirable feature...
(twitter.com/_/status/704159023264768000)
snarfed joined the channel
j12t, lukebrooker, dogada, begriffs and tantek joined the channel
#
tantek
good evening #indiewebcamp!
#
begriffs
good evening
#
KartikPrabhu
good very late time tantek
#
KartikPrabhu
kylewm: KevinMarks snarfed tommorris: sorry for not looking deeply at the mf2py performance issue. been pretty busy with "IRL" work; though kylewm seems to have it under control
j12t and jihaisse joined the channel
#
@VSChawathe
I just queued up "IndieWeb Reply" to watch later on @Vimeo: https://vimeo.com/48097093
(twitter.com/_/status/704215877282045952)
#
bear.im
edited /User:Bear.im (+108) "update project list"
(view diff)
#
bear.im
edited /palala (-25) "fix dfn"
(view diff)
Zegnat joined the channel
#
KartikPrabhu
bear: are you moving away from Japanese names?
#
bear
only for this project - it's a very old name for me
#
KartikPrabhu
there is always an exception to a rule ;)
#
bear
always
#
bear
but I may rename it once i'm happy with how it looks ;)
#
KartikPrabhu
haha! touche'
#
Loqi
awesome
Pierre-O joined the channel
#
bear
wow, did not realize how late it was
#
bear
but I now have a basic api and webmention handler for https://indieweb.news
#
Loqi
bear has 106 karma
#
bear
now to read/borrow some code for rendering the mf2 I get from a submitted post
#
bear
thanks :)
#
KartikPrabhu
bear: i should recommend mf2util
#
bear
wow - that is exactly what I was worried I would have to write!
#
KartikPrabhu
it gives an interpretation of the micro
#
bear
cool - so I can wire up some basic mf2parser and this and get something done soon
#
KartikPrabhu
I use that to interpret webmentions
#
bear
yea, I should use that for my site also
#
KartikPrabhu
kylewm++ for mf2util
#
Loqi
kylewm has 297 karma
#
bear
kylewm++ for mf2util
#
Loqi
kylewm has 298 karma
finchd, sanduhrs, friedcell, nelson, arthurspooner, Pierre-O, adactio and j12t joined the channel
#
bear
k, the idea of using mf2util was so interesting I couldn't stop...
#
bear
so I now use mf2util and mf2py to parse each webmention, store the comment json blob and use that to fill in the template
#
bear
my test webmention is very simple, will have to flex some templating skills to get a more information displayed - but it's working - woo!
Tino, friedcell, arthurspooner1, peacekeep3r and arthurspooner joined the channel
arthurspooner, prtksxna, M-kegan, thehighfiveghost, M-Kodo, M-RyanRix, sanduhrs, squeakytoy, jonnybarnes, friedcell, mxuribe, nitot, j12t, tantek and Pierre-O joined the channel
#
tantek
good morning #indiewebcamp
#
tantek
Happy 60th day of the year!
#
Loqi
[mention] Peter Molnar commented 're: http://justhugo.com/2015/10/16/freedom/ I’m not surprised about this; unf...' on a post that linked to http://indiewebcamp.com/ (https://petermolnar.eu/nwbcxx/)
#
Loqi
[mention] Peter Molnar commented 're: http://justhugo.com/2015/10/16/freedom/ I’m not surprised about this; unf...' on a post that linked to http://indiewebcamp.com/webmention (https://petermolnar.eu/nwbcxx/)
#
petermolnar
sorry, cleanup noise
#
tantek.com
edited /palala (+134) "stub, dfn fix/expansion, a few minor tweaks, linky"
(view diff)
#
cweiske.de
edited /SubToMe (+83) "/* Is SubToMe another service to depend on */"
(view diff)
Pierre-O joined the channel
#
tantek
this page needs some work to be updated: https://indiewebcamp.com/planet
#
tantek
I think it's mostly historical now?
#
voxpelli
tantek: feels like indienews has superseded it?
#
tantek
no - different efforts completely
#
tantek
planets do not require explicit submission
#
voxpelli
oh, the only ones I've been publishing to has been based on subscription to a tag on my blog, and IndieNews is more or less the same
#
tantek
IndieNews does not subscribe to anything
#
tantek
it's not "the same" to anything subscription based at all
#
voxpelli
the goal is the same at least, to gather posts related to a specific topic
#
voxpelli
in practice IndieNews could opt to include posts based on tag-filtered subscriptions, but I guess so far it hasn't felt the need for it
#
tantek
my understanding is that it is a principle of IndieNews to be explicitly (human) submission based
#
tantek
what is IndieNews?
#
Loqi
IndieNews is an indieweb news aggregator, similar to Hacker News, which only accepts submissions via webmention https://indiewebcamp.com/IndieNews
#
voxpelli
since it does two things: Both establishes the taxonomy that things are tagged with and aggregates things using that taxonomy then WebMentions is enough for it
#
voxpelli
if someone else would like to reuse its taxonomy then those would need to rely on feed subscriptions
#
voxpelli
but yeah, it feels like an evolution of the planet pattern
ttepasse joined the channel
#
tantek
I'm not sure how submission based aggregators evolved
#
tantek
there does seem to be appeal to people for both kinds
#
tantek.com
edited /planet (+817) "dfn, move past content to brainstorming, add why, how, examples stubs"
(view diff)
#
tantek
bear, I've tried to help out on /palala and /planet - please review and see if they make sense per your expertise / focus
Schnitz joined the channel
#
Schnitz
hey there
#
tantek
Schnitz, hello! Do you use any planets?
#
Schnitz
planets? What do you mean?
#
Schnitz
hi tantek :-)
#
Schnitz
what are planets?
#
Loqi
It looks like we don't have a page for "planets" yet. Would you like to create it? https://indiewebcamp.com/s/10FN
#
tantek
planets are /planet
#
loqi.me
created /planets (+19) "prompted by Schnitz https://indiewebcamp.com/irc/2016-02-29/line/1456758614252 and dfn added by tantek"
(view diff)
#
tantek
try again :)
#
Schnitz
what are planets?
#
Loqi
A planet, in the context of the indieweb, and blogs/feeds for even longer, is a site that aggregates feeds/updates from a variety of sources, typically focused on a particular topic or community https://indiewebcamp.com/planets
#
Schnitz
yes interesting
#
Schnitz
need a planet
#
ben_thatmustbeme
thats an interesting term
#
ben_thatmustbeme
had not heard that before
Sal joined the channel
#
kodfabrik.se
edited /planet (+76) "/* Examples */ Added Drupal Planet as an example"
(view diff)
#
voxpelli
Are there any example of planets created past the introduction of Digg/Reddit/Hacker News-like sites?
#
tantek
voxpelli: good question. I wonder if new planets have been another one of the casualties of the transition from feeds/readers to social media/silos
#
tantek
since planets were all AFAIK cross-site in contrast to silo reading experiences
#
voxpelli
that's why I'm thinking IndieNews might be the new "planet", because it takes the social media approach to "planets" and and brings it back to the indieweb
#
tantek
it's not really a social media approach though
#
tantek
neither is Reddit/HN
#
tantek
they don't require you to put a copy of your content there
#
tantek
unlike social media / silos
#
tantek
which now even ask for a copy of your content in special formats :P
#
tantek
(AMP, FB Instant Pages)
#
voxpelli
they require an account though, but yeah, more "silo" than "social media" I guess, but still a child of the social media era
#
voxpelli
the submission and votes are still siloed – unlike with IndieNews others can't find out whether an article has been submitted to eg. HN without asking HN
#
tantek
I don't know about that either. Digg/Reddit/Hacker News are all more derivatives of Slashdot than any social media / silo
#
tantek
parallel evolution as it were
#
tantek
Slashdot predates FB etc. by quite some time
#
tantek
https://en.wikipedia.org/wiki/Slashdot - at least 6 years to FB launch, a few more to FB being open to anyone creating an account
#
voxpelli
I think Digg/Reddit/HN takes inspiration from both Slashdot, which are fully curated by non-authors, and Planets, which are fully curated by authors
#
voxpelli
submission to eg. HN is done by both authors and users
#
voxpelli
while curation is crowdsourced through voting
tvn joined the channel
#
voxpelli
stuff can be submitted by anyone to IndieNews so in a way it's like HN there, just without the voting part
#
tantek
voxpelli: there's no curation of Planets other than the subscription list, which is typically done by hand by the owner / admin(s) of the Planet
#
tantek
there's no *content* curation that is, on planets
#
voxpelli
tantek: well, in the case of eg. Drupal Planet, the curation is done by the blogger tagging it's post with tag the planet is subscribed to
#
voxpelli
(and if one tags a post erroneously so that it ends up on Drupal Planet, then one will get many many angry people telling one about it)
mattronix joined the channel
#
Guest96033
Not heard of Planets
#
tantek
voxpelli: that is kind of funny actually (the many angry people in response to an erroneous tag)
#
voxpelli
tantek: there's some very strict guidelines: https://www.drupal.org/planet/guidelines :)
#
tantek
Mozilla planet was more about community than topic, so aggregates Mozillians blog posts in general, most which tend to be relevant to web technologies
#
aaronpk
tantek: have you documented why your home page is an h-card with a child h-feed? (vs the other options of top-level h-feed, etc)
nitot joined the channel
#
tantek
because my site is me, ergo top level h-card
#
tantek
sign of community challenges, when "Content that doesn't fit Planet Drupal:" has to explicitly include " "I Love Drupal!" posts that don’t provide valuable, actionable content to readers. "
#
aaronpk
you also have other pages that contain lists of posts, such as http://tantek.com/2015/08, which has no h-feed markup, nor a top-level h-card. is that intentional?
#
Loqi
[Tantek Çelik] Surprised that Wikipedia has many wolf parables&fables, yet lacks the Cherokee Two Wolves story: https://en.wikipedia.org/w/index.php?search...
#
tantek
(I'm curious how many times that happened before someone decided an explicit rule was necessary, and that having such a "shutdown positivity" rule was not worse than the content showing up)
#
tantek
aaronpk: those are my "in progress" /archives
#
tantek
what is an archive?
#
Loqi
archive in the context of the indieweb refers to date-grouped (often monthly) sets of posts (AKA personal historical archives, a common form of navigation), but can sometimes mean archival copy, a copy of a web page made (often by someone other than the author) at a particular point in time https://indiewebcamp.com/archive
#
aaronpk
do you (or are you planning to) have tag feeds? like https://aaronparecki.com/tag/indieweb
#
tantek
re: "tag feeds", only as a side-effect of : https://indiewebcamp.com/Falcon#linked_hashtags
#
tantek.com
edited /Falcon (+18) "/* linked hashtags */ AKA tag feeds"
(view diff)
#
aaronpk
a tag feed shouldn't have a top-level h-card like your home page, correct? it is primarily a feed, so the top-level object should be an h-feed?
#
tantek
I think of it primarily as a search result
#
tantek
never really thought of it as a feed
#
aaronpk
interesting
#
aaronpk
but you could imagine adding a tag feed as the input to a planet
#
tantek
"feed" is a dated plumbing-biased term IMO
#
tantek
from a user's perspective "search" makes more sense
#
tantek
and in fact, Twitter uses that wording - # hashtag results are part of *search*
#
tantek
what is a tag feed?
#
Loqi
It looks like we don't have a page for "tag feed" yet. Would you like to create it? https://indiewebcamp.com/s/10FP
#
aaronpk
okay, so from the user's perspective, https://aaronparecki.com/tag/indieweb is a search result?
#
tantek
yes! just like https://twitter.com/hashtag/indieweb is a search result
#
voxpelli
I think "search" gives too much of a Google vibe – I'm trying to stay away from calling features of a web sites search as far as possible to avoid giving the wrong impression
#
Loqi
[Aaron Parecki] Not so subtle. #uber
#
voxpelli
I actively try to use terms like "filters" and probably "feed" and such rather than search
#
aaronpk
(you may notice the summary from Loqi on these is leaving something to be desired, which is where I'm actually going with this line of questions)
#
tantek
other attempts in the past have used things like "channel"
#
tantek
aaronpk: I figured :)
#
aaronpk
(I call those channels internally!)
#
tantek
it's all about serving Loqi
#
Loqi
is done
#
tantek.com
edited /planet (+54) "WP"
(view diff)
#
Loqi
[indieweb] "Do Facebook Instant Articles Support The Open Web... or Facebook's Walled Garden?" by Dan York http://feedproxy.google.com/~r/DisruptiveConversations/~3/3v_6mljmFNY/do-facebook-instant-articles-support-the-open-web-or-facebooks-walled-garden.html
#
tantek
what is Instant Article?
#
Loqi
It looks like we don't have a page for "Instant Article" yet. Would you like to create it? https://indiewebcamp.com/s/10FQ
#
tantek
what is Instant Articles?
#
tantek
what is a snowflake?
#
Loqi
snowflake in the context of the indieweb, is typically used to refer in a derogatory way to "snowflake APIs", APIs that are (often silo) website or service provider-specific (unique like snowflakes) rather than an open standard https://indiewebcamp.com/snowflake
#
tantek.com
edited /Instant_Articles (+81) "fix dfn, move docs link, see also"
(view diff)
j12t, wolftune, tantek and thehighfiveghost joined the channel
#
aaronpk
so if http://tantek.com represents a person, as does http://twitter.com/t, you might expect XRay to say that it's a "profile"?
#
Loqi
Tantek Çelik
#
tantek
no, because "profile" is silo-framing, as in you are *just* a profile in *their* silo
#
aaronpk
what's the right term then?
#
tantek
an h-card by default is a *person*, unless the name==org, in which case it is a organization or venue (may need more logic)
#
aaronpk
well that's a thought
#
tantek
pretty sure that's on the classic hCard docs
#
aaronpk
I could have it return type: "person" or "organization" or "venue"
#
Loqi
hCard 1.0
#
Loqi
hCard 1.0
#
tantek
Loqi, could you parse the fragment indicated subtree instead?
#
Loqi
is done
#
aaronpk
but that is not a bad idea
#
tantek
would also help with old-style blog permalinks
#
aaronpk
finding the tree of that would be a bit of work
#
tantek
not at all, first id="" attribute with it
#
aaronpk
i think it would be all the HTML after the element with that name/id up until the next named element or the end of the document
#
aaronpk
in practice, most of the occurrences of that don't actually encapsulate the HTML inside that elmeent
#
tantek
it encapsulates enough for this use case :)
#
tantek
(Loqi providing summaries)
#
Loqi
is done
#
aaronpk
most of the ones i've seen are just <a name="..."></a> with stuff after it but not inside it
#
tantek
apparently Loqi is serious about being done
#
tantek
aaronpk: which is why you don't bother with name=
#
tantek
I was very specific. first id="" attribute with it :)
#
tantek
here's another test case for you on that very topic: http://tantek.com/log/2002/11.html#L20021128t1352
#
aaronpk
anyway...
#
tantek
should return:
#
tantek
<A>norexic </A>nchors
#
aaronpk
it sounds like you're suggesting dropping the concept of a feed as an entity itself
#
tantek
feeds are dying/dead as user level entities, having been superseded by silo UX long ago.
#
aaronpk
and tantek.com is a person, and has a list of posts on the page
#
tantek
recent posts, recent articles, upcoming events
#
tantek
I suppose I could add a rel=feed that linked to the fragement identifier of *just* the primary list of posts on the page
#
tantek
I wonder if that would help with discovery of posts for readers
#
aaronpk
I would rather the mf2 data structure have all the information it needs to know where the primary list of posts is, rather than relying on external information from the HTML
#
tantek
user adds a URL to a reader to "follow", reader goes to discover the list of posts to from that URL, looks for a top level h-feed, then top level h-entry(s), then rel=feed which if it goes to a local frag ID, then parses that subtree for an h-feed, then h-entry(s)
#
tantek
interesting, so you'd like to actively recommend *against* rel-based discovery of content?
#
tantek
(not questioning, but rather seeking clarification)
#
aaronpk
it makes it harder to implement when you have to handle both the parsed microformats and also raw HTML
#
tantek
because that has larger ramifications (we could consider that path)
#
tantek
this is why the mf2 JSON included rels to start with
#
aaronpk
for me, the whole benefit of microformats is that I don't have to think of it as HTML anymore
#
tantek
so all you needed was the mf2 JSON (the "parsed microformats")
#
tantek
ah, even if you have the rel values, if they are URLs to things in the HTML, you'd have to look at the HTML again?
#
tantek
*interesting* this is implying that rel values in the mf2 JSON to local things in the page should somehow also provide references to those parsed microformats in the mf2 JSON!
mattronix joined the channel
#
aaronpk
basically when I look at tantek.com I see this: https://aaronparecki.com/uploads/Screen-Shot-2016-02-29-08-26-59.png and expect to be able to find everything I need from that
#
tantek
do we parse the id="" attribute on a root microformats element as an implied "p-id" or even a fully qualified "u-id" on that microformat?
#
tantek
because then you could do the look-up from a rel URL value into the list of microformats and find the one that was referenced
#
aaronpk
huh, that's a thought
#
tantek
we've had other suggestions I think for things like the lang="" attribute
#
aaronpk
probably would need to be fully qualified u-id since it is likely the same id string would exist on other pages
#
tantek
yes that makes more sense
#
ben_thatmustbeme
that would certainly be useful information though
#
tantek
in that case, should we overload by implying "uid" if none is specified?
#
tantek
(rather than making up a new property "id")?
#
tantek
or keep it as "id" for HTML attr-name consistency (and avoiding collision/confusion with "uid")
#
aaronpk
my instinct is to not use the uid property, since that has specific semantics of being a globally unique identifier
#
tantek
OR prefix it as an HTML attribute "html-id"
#
tantek
yes, I agree with your instinct
#
tantek
where's the issue about specifying "lang" or natural languages for microformats?
#
tantek
someone here (voxpelli ?) filed it somewhere and I can't find it
#
Loqi
microformats2-parsing-issues
#
Loqi
microformats2 parsing brainstorming
#
tantek
hah at least I remembered the author!
#
ben_thatmustbeme
KevinMarks was interested in that too i though
#
ben_thatmustbeme
don't much like html-id as an attribute, that would seem even more confusing
#
ben_thatmustbeme
uid would be fine so long as its the full URL wouldn't it?
#
ben_thatmustbeme
and so long as the html is correctly using ID only once per page
#
tantek
ben_thatmustbeme: that's awfully optimistic about fragment identifiers ;)
#
ben_thatmustbeme
yeah, i know :(
#
tantek
(in practice not very globally unique, nor canonical, both of which implied by "uid")
#
ben_thatmustbeme
well there are easy tests to see if the ID is re-used on the page right?
#
tantek
but that's the lesser of the two important constraints / semantics I noted above
#
ben_thatmustbeme
doesn't safe-guard you against future errors though
#
tantek
(canonical is more important)
#
ben_thatmustbeme
is uid supposed to be the cannonical only? (i never fully understood that with h-cite for example)
#
tantek
aaronpk, ben_thatmustbeme, voxpelli updated: http://microformats.org/wiki/microformats2-parsing-brainstorming#Parse_language_information please add more comments there on the proposal(s) !
#
Loqi
microformats2 parsing brainstorming
#
tantek
ben_thatmustbeme: yes, uid implies asserted canonical URL
#
tantek
which also implies globally unique
#
ben_thatmustbeme
where would one put the syndicated copy ID?
#
tantek
"u-syndication" :P
#
ben_thatmustbeme
the syndicated copy should have a u-syndication to itself as the only identifier?
#
tantek
huh? no - why does a copy need an id at all beyond its URL?
[kevinmarks] joined the channel
#
@CoralineAda
@_danilo I got so fed up with that that I just created a simple blog off my own site. It just wasn’t worth it.
(twitter.com/_/status/704340629745115136)
#
tantek
if you're looking to de-dup, you should use the *original* canoncial URL UI
shiflett joined the channel
#
tantek
s/URL UI/URL and uid
#
Loqi
tantek meant to say: if you're looking to de-dup, you should use the *original* canoncial URL and uid
#
tantek
hence why syndicated copies should link back to originals
#
tantek
they don't need their own "id" ever
snarfed joined the channel
#
ben_thatmustbeme
post <famous person> lied when he claimed i said <some quote>! Here is the original <canonical> and here is his misquote of me <copy>
#
ben_thatmustbeme
as an example
#
ben_thatmustbeme
if the syndicated copy has a fragment id that is unique to the page, you have a definite unique url / id for that syndicated copy
#
tantek
that use of <canonical> is just a link
#
tantek
it can't assert anything canonical about it since it's an *external* link
peacekeep3r joined the channel
#
tantek
only the destination of that link can assert whether it is canonical or not
#
ben_thatmustbeme
sorry, yeah, i was shorthanding that as <link to canonical>
KartikPrabhu and loic_m joined the channel
#
tantek
ben_thatmustbeme: no I understand. I'm saying that from an outside perspective, it's just a link
#
tantek
you don't get to assert what is canonical on someone else's site
#
tantek
only they do
#
tantek
so it's pointless to design markup to capture that
#
tantek
since it's not trustworthy / trustable information in the first place
#
tantek
it's *just* a link
#
[kevinmarks]
With html the id duplication problem is solved by going to the first in the document, but mf2 parsing doesn't preserve document order, so may add ambiguity
#
tantek
anyone coming along that wants to see if it is canonical has to go *get* that link and check it out to see what it asserts about itself
#
ben_thatmustbeme
yes, i know that, how do you refer to the data of that version though, thats still content that is parsable and referencable, just not trustworthy information
#
[kevinmarks]
You can assert that the current page is non-canonical with a rel=canonical link
#
tantek
nah, it's pointless to refer to it since it's not trustworthy information
#
tantek
all you can refer to is what is trustable, which is that it's a link, nothing more
#
tantek
KevinMarks: the dup id problem can be solved at parse time, since the document *is* parsed in document order even if the parsed result isn't guaranteed to be so (yet the algorithm pretty sure specifies that order)
#
ben_thatmustbeme
<div class="h-cite"><a class="u-url u-uid" href="twitter.com/t/00000000101010101">permalink</a><span class="e-content">I believe in creationism<span></div>
#
ben_thatmustbeme
If i post this, how do you comment on it and call me out as a liar?
#
ben_thatmustbeme
any attempt to parse it says that it doesn't exist as the url is gone
#
aaronpk
I don't understand the problem? That post would have a URL on your site, so someone would just reply to that URL
#
ben_thatmustbeme
just because its not validated as having being correct (verifying the source) does not mean it isn't something useful
#
ben_thatmustbeme
aaronpk: who said it has a permalink?
#
ben_thatmustbeme
its just in a feed
#
aaronpk
now you're just being silly
#
ben_thatmustbeme
not so much of how you comment on it, but how do you cite it
#
[kevinmarks]
Page with id was the original permalink model before posts got standalone pages
#
[kevinmarks]
Tantek: so that is an extra constraint for parsers, that first id wins?
#
ben_thatmustbeme
oh, i hadn't though of two IDs, that gets even messier
#
ben_thatmustbeme
what i was trying to get at is that a syndicated copy should have and ID of itself and and ID of what it is claiming to be a copy of, whatever that ID of the syndicated copy is, thats what you should be parsing the id from html into
#
tantek
KevinMarks: yes, mf2 parsers should only look at first instance of any "id" attribute hence doing de-duping at parse time for id
#
KartikPrabhu
ben_thatmustbeme: that example makes little sense, in that if you wanted to lie you could lie in the markup too ;)
#
ben_thatmustbeme
i am doing just that, thats what i'm saying
#
tantek
agreed - no need to bother with such abusive use-cases
#
tantek
because they are ignorable
#
tantek
best practice is to not link to them
#
KartikPrabhu
right and there is no way to fix it by making rules as a "liar" can break any rules you make up
#
ben_thatmustbeme
i went down this path to figure out what a field should be for parsing an html id into
#
ben_thatmustbeme
and because i do think there is use in being able to reference what is effectively a link object,
#
aaronpk
ben_thatmustbeme: there isn't even an html ID in your example, i'm not sure what you're getting at
#
ben_thatmustbeme
probably a bad example
gRegorLove joined the channel
#
tantek
designing first or primarily by abusive use-cases is not a good methodology. do the opposite instead
#
gRegorLove
Good morning, indiewebcamp.
#
tantek
design by productive/positive use-cases first
#
aaronpk
this got way off track of what I was originally asking about anyway
#
ben_thatmustbeme
forget the abusive case and it works the same as accidental
#
tantek
aaronpk: what you are asking about originally is essentially a /link-preview use-case question
#
tantek
aaronpk: could you add your use-case to /link-preview, perhaps in the Brainstorming section?
#
aaronpk
sort of, but I also want to be able to reliably find entries on the page for consuming in a reader
#
tantek
well that's another use-case, add that to /reader
#
tantek
point is, capture your use-cases individually as they may have different or overlapping solutions
#
tantek
but unless we capture them, they get lost, then history repeats itself
#
aaronpk
hm there is already an "autodiscovery of h-feeds" in the reader brainstorming
#
KartikPrabhu
aaronpk: that was added primarily by me IIRC
#
KartikPrabhu
feel free to disagree/tweak
#
aaronpk
but i thought we were trying to get away from the plumbing-centric view of feeds
#
tantek
aaronpk: we are, hence update /reader accordingly!
#
tantek
with exactly what you said
#
tantek
"to be able to reliably find entries on the page for consuming in a reader"
misa__ joined the channel
#
tantek
^^^ note no use of "feed" in that use-case
#
tantek
welcome misa__ !
#
misa__
hello!
thehighfiveghost joined the channel
#
tantek
misa__: you said earlier in #microformats "indieweb seems very right for me, I can comprehend most of its concepts"
#
misa__
yes
#
tantek
glad to hear it - what is your personal site, and what are your personal human-facing goals for it?
#
KartikPrabhu
regarding rel based vs mf2 based discovery. parsed mf2 does have the rel attributes too
#
tantek
(KartikPrabhu keep catching up on logs - I noted that already ;) )
#
KartikPrabhu
very long discussion ;)
#
ben_thatmustbeme
KartikPrabhu: save youself some time and ignore my whole convo section
#
KartikPrabhu
ben_thatmustbeme: not to worry I am good at speedy parsing of relevant content ;)
#
misa__
well I'm working on a website related to cooking and gastronomy, it's going to be kind of a blog/recipe/discussion website, still working on the format
#
ben_thatmustbeme
aaronpk: also noticed your comment about JF2 ealier, what was it you were referring to specifically? the "refs:" ?
#
aaronpk
not sure, which comment?
#
KartikPrabhu
misa__: nice. do you already have it up at a URL?
#
misa__
i registered the domain, gastronomades.gr
#
misa__
it's going up these days
#
ben_thatmustbeme
<aaronpk> ben_thatmustbeme: what do you think about that method for returning nested objects in jf2?
#
misa__
jekyll/github-pages style
#
aaronpk
ah yes!
#
tantek
ben_thatmustbeme: I think jf2 discussion, since it's generic to microformats in general, could be done in #microformats
#
KartikPrabhu
misa__: good to hear. Is there something particular you wanted on the site that lead you to indieweb?
#
ben_thatmustbeme
good point, moving over
#
kylewm
gRegorLove: already have a PR in to Known for it
tantek joined the channel
#
KartikPrabhu
alright gotta run
#
KartikPrabhu
misa__: feel free to ask questions in here and good luck with your site :)
#
kylewm
gRegorLove: actually it was already merged, werd.io probably hasn't updated yet
#
gRegorLove
Just noticed it on your posts too.
#
misa__
yes well, the concept of POSSE is something that i'm interested in just i'm not sure how to implement it on a static website.
[kevinmarks] joined the channel
#
[kevinmarks]
Bridgy publish is one way
#
gRegorLove
What is Bridgy?
#
Loqi
Bridgy is an open source project and proxy that implements backfeed as a service https://indiewebcamp.com/bridgy
#
gRegorLove
Huh, surprisingly few Bridgy Publish mentions on /Bridgy.
#
misa__
that looks great!
#
gRegorLove
Could probably use a sub-section, at least.
#
gRegorLove
What is Bridgy?
#
Loqi
Bridgy is an open source project and proxy that implements backfeed and POSSE as a service https://indiewebcamp.com/bridgy
#
loqi.me
created /Bridgy_Publish (+19) "prompted by gRegorLove and dfn added by gRegorLove"
(view diff)
snarfed joined the channel
#
aaronparecki.com
edited /reader (+305) "/* Autodiscovery of h-feeds */"
(view diff)
#
gRegorLove
Hm, the two sub-sections under /Bridgy#Why seem the same? "Send silo interactions" is actually describing backfeed.
#
tantek
KevinMarks: can you add that to /silo-quits ?
thehighfiveghost joined the channel
#
snarfed
gRegorLove: feel free to revise!
#
Loqi
bear has 107 karma
#
snarfed
...but oh man is that going to get confused with http://news.indiewebcamp.com/ :P
#
snarfed
(heh, sorry, i see that was already discussed)
#
bear
yea, I figure there will be some adjustments :)
j12t joined the channel
#
aaronpk
maybe we should have explicitly stated goals with them that are different
#
aaronpk
e.g. news.indiewebcamp.com is intended to be explicit submissions, not aggregated like a planet
#
bear
I feel that planet isn't the right term for indieweb.news to be honest - but I also have not staked out any firm position
#
snarfed
i was actually just thinking about the names/domains, even before i knew what they did (and definitely before i got anywhere near reading text goals anywhere)
#
bear
if it's determined that there is too much overlap (or folks find it confusing) then I'll point indieweb.news to news.indiewebcamp.com
#
gRegorLove
What is indieweb.news?
#
Loqi
It looks like we don't have a page for "indieweb.news" yet. Would you like to create it? https://indiewebcamp.com/s/10FT
#
bear
I haven't defined it yet gRegorLove precisely because right now it's just a way for me to explore some python code for displaying different posts, mentions, comments, etc
#
gRegorLove
tanteking fail.
#
bear
one of my thoughts was that it would be a self-refreshing list of items someone is interested in, with the default being a list of activity from the indieweb community
#
bear
not just posts from indieweb community members
#
bear
(sheesh I hope that makes sense)
j12t joined the channel
Calli joined the channel
#
@michielbdejong
RT @fakebaldur Anybody who is interested in the Open Web but isn't paying attention to https://indiewebcamp.com/ isn't, actually, paying attention.
(twitter.com/_/status/704375066553933824)
#
tantek
we need to help him upgrade his classic microformats to mf2 to answer the issues at the end of his post
#
tantek
is wondering where he got the old vCard copy/pasta
#
aaronparecki.com
edited /Posts_about_the_IndieWeb (+129) "/* 2016 */ add barrymcgee.co.uk post"
(view diff)
#
tantek
oh there's a tweet about it - maybe we should encourage both Barry and Calum to join us here beyond just twitter? https://twitter.com/calum_ryan/status/704370659871301634
#
@ChrisAldrich
@benwerd Let us know if you need a side of #indiepublishing to go with your #indieweb! @boffosockobooks
(twitter.com/_/status/704376941424447488)
#
Loqi
[indieweb] "@benwerd Let us know if you need a side of #indiepublishing to go with your #indieweb! @boffosockobooks" by Chris Aldrich http://stream.boffosocko.com/2016/benwerd-let-us-know-if-you-need-a-side-of
#
@LukasRosenstock
RT @fakebaldur Anybody who is interested in the Open Web but isn't paying attention to https://indiewebcamp.com/ isn't, actually, paying attention.
(twitter.com/_/status/704377672055574529)
snarfed joined the channel
#
@jgarber
@nicolaschevobbe Congrats on the launch and welcome to the #indieweb! /cc @SaraSoueidan
(twitter.com/_/status/704383400103309313)
#
aaronpk
that's a great thread
#
@SaraSoueidan
I wish developers/designers wrote more articles on their own blogs. I’d rather share links to those blogs than to articles on Medium, tbh.
(twitter.com/_/status/702063641877725184)
calumryan joined the channel
#
@WendyandCharles
ReadersGazette: BLOG Indie Author Answers by Jim Heskett http://www.jimheskett.com/indieauth Get help writing your book #bookbloggers 90
(twitter.com/_/status/704123801609568256)
ikomi joined the channel
#
calumryan
tantek: I’m here :) I’ll try and get Barry to join in next time I see him and suggest this
#
ikomi
hello, I was referred here by wolftune from snowdrift.coop, I have a question regarding html meta tags and was wondering if I could ask for opinions here
#
tantek
hey calumryan !
#
tantek
any idea how Barry got the idea of using old vcard microformats?
#
ikomi
I'm helping out at snowdrift.coop (a site focused on free/libre/open works) and looking into adding meta tags for better search/social visibility. Specifically considering opengraph tags and twitter cards. I noticed the IndieWebCamp website has opengraph tags, is this a "standard" and have you noticed any improvements in visibility and such?
#
tantek
welcome ikomi !
#
ikomi
tantek: ty :) hope you're all having a good week so far
#
tantek
so far so good
#
tantek
in general html meta tags are to be avoided for all the usual invisible metadata bad/rot problems
peacekeep3r joined the channel
#
ikomi
tantek: good :) is that just for keyword tag or even description tags?
#
tantek
thanks for pointing out that we have some on indiewebcamp.com - we're investigating how that happened
#
tantek
ikomi - pretty much all meta
#
tantek
except meta charset
#
tantek
it's one more invisible place you forget to update / change when the *visible* content changes
#
tantek
better to just markup the page with visible microformats and be done
#
calumryan
tantek: I’m not sure how Barry used the old vcard, certainly not from my talk which didn’t go into this much detail
#
ikomi
tantek: microformats as in http://microformats.org/ ?
#
tantek
ikomi yes!
KartikPrabhu joined the channel
#
tantek
and #microformats for discussion
snarfed joined the channel
#
tantek
ikomi: in general, for anything indieweb you can ask "what is" here
#
tantek
or what are
#
tantek
what are microformats?
#
Loqi
microformats are extensions to HTML for marking up people, organizations, events, locations, blog posts, products, reviews, resumes, recipes etc https://indiewebcamp.com/microformats
#
ikomi
we have a meta description tag for the 1-liner that appears in search engine results (though if we didn't add it some search engines would just take the body text for their description), so looking if anything else would be useful
#
ikomi
tantek: okay, ty for the recommendation on microformats :)
#
tantek
ikomi: exactly - taking the body text for description is *much* better because then you can just make sure your body text is up to date!
#
ikomi
tantek: true. that works best if the first few lines of the body text contain the relevant keywords or does it not matter?
#
tantek
search engine keyword heuristics have become too complex to bother trying to jam/pack etc.
#
tantek
just make sure your content has relevant prose with good valid semantic HTML markup
#
tantek
they look for that in general
#
@SamatJain
RT @fakebaldur Anybody who is interested in the Open Web but isn't paying attention to https://indiewebcamp.com/ isn't, actually, paying attention.
(twitter.com/_/status/704390598652928000)
#
ikomi
tantek: ty … and further to the body text question, is taking the body text common behaviour for all major search engines? I see a lot of large sites (index page) have no description in search results
#
snarfed
kylewm++ again for the huge mf2py optimization in https://github.com/tommorris/mf2py/pull/61
#
Loqi
kylewm has 300 karma
#
snarfed
early data suggests that it's noticeably reducing both bridgy's costs and its backlog, from 30-180m down to 0!
#
tantek
ikomi who knows why sites lack description, it might be because of the js;dr problem
#
tommorris
yep very good work kylewm
#
tommorris
kylewm++
#
Loqi
kylewm has 301 karma
#
gRegorLove
What is metadata?
#
Loqi
Metadata, specifically "descriptive metadata", is an umbrella term for all the different ways data (content) can be marked up or organized to facilitate machine-readability https://indiewebcamp.com/metadata
#
gRegorLove
What is invisible metadata?
#
Loqi
For reference: WS-Deathstar [1][2][3][4][5], e.g.: https://indiewebcamp.com/invisible_metadata
#
ikomi
tantek: okay. sorry, one more question, is the current version of microformats w3c valid? just looking at rel=tag for links, the copyright says they intend to submit to standards body, is it still wip?
#
tantek
ah - that should be updated, because rel=tag *was* submitted and incorporated into HTML5
#
tantek
ikomi what URL did you find that at?
#
ikomi
http://microformats.org/wiki/rel-tag from the front page under rel microformats
#
@thornlow_brenda
RT @ReadersGazette BLOG Indie Author Answers by Jim Heskett http://www.jimheskett.com/indieauth Get help writing your book #bookbloggers 90
(twitter.com/_/status/704121931965067268)
#
@ReadersGazette
BLOG Indie Author Answers by Jim Heskett http://www.jimheskett.com/indieauth Get help writing your book #bookbloggers 90
(twitter.com/_/status/704121416011087872)
#
Loqi
rel="tag"
#
tantek
ikomi: all microformats are w3c valid BTW - because microformats are designed to make use of existing valid W3C standards
#
ikomi
*wiki front page
#
ikomi
tantek: great, ty … so microformats is a significant improvement worth incorporating, but not a "must"? will let the group know and see about incorporating it. ty again!
#
tantek
ikomi: what do you mean by "a must"?
#
tantek
everything on the web is "optional" :0
#
ikomi
tantek: bad wording, must as in, essential for better search/social network sharing visibility
#
ikomi
*important for
#
bear
kylewm++ for mf2py optimizations and for mf2util
#
tantek
yes microformats are important for better search / social network / indieweb sharing visibility
#
Loqi
kylewm has 302 karma
#
tantek
ikomi: thank you for the question about rel-tag, I have updated the specification accordingly to note that rel-tag has been incorporated into HTML5: http://microformats.org/wiki/rel-tag
#
Loqi
rel="tag"
#
ikomi
tantek: does it have any effect on accessibility as well or do readers prefer entirely different specifications?
#
tantek
ikomi some readers use microformats. for *screen* readers you have to check each one individually as they all differ, however they do tend to make use of ARIA attributes.
#
ikomi
tantek: and np (sorry for all the questions!)
#
tantek
ikomi questions appreciated! :)
#
ikomi
tantek: got it, ty again!
#
miklb
ikomi: think of it this way, there are only positives to using microformats now, and going forward
#
ikomi
miklb: nods, good to hear it's already in html5, initially wasn't sure if I should suggest implementing if the spec is wip and may change
#
ikomi
the specs looks pretty useful
#
ikomi
ty again, might be back again but regardless, have a good day :)
#
loqi.me
created /résumé (+132) "prompted by gRegorLove and dfn added by tantek"
(view diff)
#
tantek.com
edited /résumé (+41) "citation"
(view diff)
tantek and friedcell joined the channel
#
gregorlove.com
created /resumé (+376) "stub"
(view diff)
#
gregorlove.com
created /resume (+21) "redirect until we need the verb "resume""
(view diff)
#
gRegorLove
Anyone post their resumé on their site? Add to /resumé#IndieWeb Examples
#
petermolnar
^^^ rhiaro ?
ttepasse joined the channel
#
aaronparecki.com
edited /resumé (+35) "add my example"
(view diff)
#
rhiaro
rhiaro.co.uk/tag/cv it's just tagged blog posts because I refuse to be bound by resumé rules, does it count?
#
Loqi
rhiaro: tantek left you a message on 1/19 at 1:15pm: do you know when the first Edinburgh HWC was? And how many were held last year? Can you edit https://indiewebcamp.com/2015-review#Homebrew_Website_Club with that info? http://indiewebcamp.com/irc/2016-01-19/line/1453238124010
#
Loqi
rhiaro: tantek left you a message on 2/16 at 12:35pm: do you keep a list of all the books you own? do you publish that list anywhere? what do you call it? personal library? http://indiewebcamp.com/irc/2016-02-16/line/1455654923934
#
rhiaro
Oh man it's been a while since I typed in this channel
#
tantek
rhiaro: it counts as a good counter-example - definitely worth adding as such
#
rhiaro
quietly goes back into hiding
#
rhiaro
tantek: re books, no I don't keep a list
#
gRegorLove
Come back! We have cookies.
#
rhiaro
Are they vegan cookies?
#
tantek
just /read posts then?
#
tantek
notes to self to start trying vegan cookie recipes
#
rhiaro
Yeah I did a couple of read posts, mostly papers rather than books
#
rhiaro
And tantek re HWC Edinburgh, bother tbrb or kongaloosh for that information :)
#
tantek
tbrb: ^^^
#
tbrb
I'll grab a look tomorrow :)
#
tantek
tbrb in particular: do you know when the first Edinburgh HWC was? And how many were held last year? Can you edit https://indiewebcamp.com/2015-review#Homebrew_Website_Club with that info?
#
tbrb
I'll dig into the records tomorrow
#
petermolnar
it having the (u-in-reply-to|u-like-of|etc) within the e-content part valid or should that be outside of e-content?
#
petermolnar
because in case of a reply, it's kind of makes sense as part of the content
#
aaronpk
it's fine to have it inside the content
#
aaronpk
as long as you don't mind the text of the link being rendered when the comment is displayed
tantek and yoowho joined the channel
#
yoowho
i want to thank the san francisco indie web meet folks last week for their feedback helping me with creating a new homepage look http://moshecohen.net/
#
tantek
welcome yoowho and great to meet you last week!
#
tantek
other folks you met last week: benwerd, KevinMarks, erinjo
#
yoowho
thanks tantek!
squeakytoy joined the channel
#
tantek
yoowho: there's people here from around the world - feel free to ask any indieweb related questions you like, at any time, 24/7 :)
#
@d0gada
@malcolmblaney Most #IndieWeb silos like #Twitter and #Facebook protects timelines with #OAuth tokens and your… (https://dogada.org/e/118t685U13912)
(twitter.com/_/status/704409191956385793)
#
tantek
nice resizing image yoowho !
sparverius and shiflett joined the channel
#
Loqi
[bridgy] Dmytro V. Dogadailo replied '@malcolmblaney Most #IndieWeb silos like #Twitter and #Facebook protects timelin...' to a tweet that linked to http://indiewebcamp.com/indie-config (https://twitter.com/d0gada/status/704409191956385793)
begriffs and arthurspooner joined the channel
#
Loqi
[indieweb] "Alaska Cruise Log Day 3: Juneau" http://tantek.com/2015/221/b1/alaska-cruise-log-juneau
#
tantek
how did that happen?
#
petermolnar
"Today, we’re launching a new feature for Svbtle members: it’s a promise that the Svbtle service and your published content will remain available on the web forever."
#
aaronpk
tantek: there's some mentions of "indieweb" at that page
#
aaronpk
not sure why it only now popped up though
snarfed1 joined the channel
#
kylewm
wow, that Svbtle announcement is cool. i think the most forcefully i've seen that sort of statement made
#
aaronpk
It's nice, but I'm not sure how much I can believe it
#
aaronpk
They did acknowledge "If Svbtle, its parent corporation, or any affiliated entities are acquired, purchased, or taken control of by a third party in a way that would negatively impact the service, we will do everything in our power to ensure that the above two promises endure as part of any legal agreements into which we enter."
#
aaronpk
but it's hard to actually control what happens in that situation
#
kylewm
hmm, but then fine print: c. "Forever" means until the HTML "web" is no longer generally accessible or until Svbtle becomes financially insolvent (which we intend to avoid "
#
kylewm
I intentionally cropped out some context there :P but it's a pretty big hedge
snarfed joined the channel
#
Loqi
[indieweb] "so you are actually POSSEing video right to twitter ? by the look of your post is that even possible ? #amazed #indieweb" by Rick Mendes http://www.rmendes.net/2016/so-you-are-actually-posseing-video-right-to-twitter-by
#
aaronpk
more details from dcurtis about svbtle...
#
aaronpk
in particular, "...we would work very closely with [archive.org] to archive all data. And perhaps even give them the domain name."
#
aaronpk
and "I'm working on a trust that will be created in the event of my death (or the company's death) that can keep the servers running for a very long time to come, even with no paying customers."
KartikPrabhu joined the channel
#
aaronpk
so he looks like he's pretty serious about it, and it's not just a marketing gimmick to try to get people to use svbtle instead of medium
#
Loqi
[indieweb] "Comment on by D.V.D." by D.V.D. https://petermolnar.eu/o37tvb/#comment-11569
#
KartikPrabhu
dcurtis writes pretty good stuff and seems reasonable
j12t and snarfed joined the channel
#
KevinMarks
looks like gnusocial is losing the spaces between tags on tantek's posts https://quitter.no/notice/1157444
#
Loqi
[Tantek] Tantek (tantekcom)'s status on Monday, 29-Feb-2016 22:35:00 CET
#
tantek
KevinMarks: that would be bad HTML parsing on their part, dropping line-breaks instead of turning them into spaces
#
tantek
Perhaps show them our work on white-space handling
#
KevinMarks
well, once we fix it
#
tantek
are they using php-mf2?
#
aaronpk
not sure, but they'd be using the same thing php-mf2 used
#
tantek
do we track webmention mentions here? https://twitter.com/t/status/704428208133832704
#
gRegorLove
What is Svbtle?
#
Loqi
It looks like we don't have a page for "Svbtle" yet. Would you like to create it? https://indiewebcamp.com/s/10FW
#
tantek
Twitter sucks at auto-detecting even the most obvious patterns of spam: the same exact thing tweeted in rapid succession by multiple accounts: https://twitter.com/RikMende/status/704425458037706752
#
tantek
Begun, the Bot Wars have.
snarfed1 joined the channel
#
snarfed1
does known not support video posse yet? i know benwerd has lots of experience and opinions on video handling, maybe he's waiting until he feels like he can do it right?
#
Loqi
[indieweb] "Just reading the features of this beast of the #indieweb #Falcon @falcon https://indiewebcamp.com/Falcon #amazed @t Video POSSE to tw,fb is crazy!!!" by Rick Mendes http://www.rmendes.net/2016/just-reading-the-features-of-this-beast-of-the-indieweb
#
gRegorLove
Feel free to expand with info on the promise ^
[kevinmarks] joined the channel
#
[kevinmarks]
snarfed: ben doesn't support video hosting on known, only audio
#
Loqi
[indieweb] "Alright I just had a #Eureka moment!!! #indieweb" by Rick Mendes http://www.rmendes.net/2016/alright-i-just-had-a-eureka-moment-indieweb
#
snarfed
he is definitely enthusiastic :P
KartikPrabhu joined the channel
#
tantek
this is why it's important to push forward with what you want on your site, as hacky as it may be. show a broader set of people what is possible.
tantek joined the channel
#
@RikMende
RT @fakebaldur Anybody who is interested in the Open Web but isn't paying attention to https://indiewebcamp.com/ isn't, actually, paying attention.
(twitter.com/_/status/704440605099859970)
#
gregorlove.com
edited /site-deaths (+48) "archive Mailbox"
(view diff)
jciv joined the channel
#
kylewm
KevinMarks: tantek: gnu social isn't using php-mf2 -- those status updates come from t's atom feed
#
snarfed
aaronpk: just fyi, looks like something of yours sent bridgy 1-5qps of GET https://brid-gy.appspot.com/comment/twitter/t/704421094451589120/704424081131634688 around 3pm pst
#
Loqi
[Kevin Fox] @t I have such lament of missing out. LOMO.
#
snarfed
client IP 2600:3c01::f03c:91ff:fe6e:e59f
#
snarfed
totally ok, just fyi
lukebrooker and KartikPrabhu joined the channel
#
aaronpk
snarfed: for how long?
#
snarfed
maybe just 5m or so
#
snarfed
100s-1000s of reqs though :P
#
aaronpk
that's likely webmention.io trying to verify the webmention from bridgy