#indieweb 2016-10-25

2016-10-25 UTC
snarfed, korayal, funwhilelost, KevinMarks_, KevinMarks and misuba joined the channel
#
GWG
!tell snarfed Found a problem that will require a rewrite of Post Kinds.
#
Loqi
Ok, I'll tell them that when I see them next
tantek, wolftune, KevinMarks and snarfed joined the channel
#
snarfed
GWG: ruh roh. problem with wp-micropub?
#
Loqi
snarfed: GWG left you a message 46 minutes ago: Found a problem that will require a rewrite of Post Kinds.
#
snarfed
or post kinds?
#
GWG
snarfed: You are storing single properties as arrays, consistent with the Micropub recommendation.
#
snarfed
aha yes
KevinMarks joined the channel
#
GWG
I wasn't.
#
GWG
Because I never set the UI to look for multi-replies and such
#
snarfed
got it
#
GWG
But that should be fixable.
#
snarfed
sounds like CRUD worked though, including auth etc?
#
GWG
Yes
#
snarfed
yay good
#
Loqi
does a happy dance!
#
GWG
snarfed: I was leading toward JF2 for storage, as opposed to MF2. But I can adapt.
#
snarfed
GWG: if you want, you can use the before_micropub filter to change the post meta to any format you want
#
GWG
snarfed: Compatibility is an issue though.
#
snarfed
GWG: eh post kinds isn't expected to work with posts made before it was installed, right? this would be the same
#
snarfed
but up to you
#
GWG
snarfed: Just trying to think what is good for everyone.
#
GWG
I want to support multiple in-reply-tos, even if the UI isn't up for it right now, so probably better off converting to array
#
GWG
I can still do an is_array to check for the old way
KevinMarks_ joined the channel
#
snarfed
released wordpress-micropub v1.0! https://wordpress.org/plugins/micropub/
#
GWG
snarfed++
#
Loqi
snarfed has 243 karma (239 in this channel)
snarfed joined the channel
#
miklb
is indeweb.org down for anyone else? Trying to load the /2016/LA page
#
snarfed
miklb: fine for me
#
KartikPrabhu
works here
#
KartikPrabhu
what's the "Pivotal" logo on /2016/LA ?
KevinMarks, Pierre-O, sl007, loicm, tantek, KevinMarks_, smcgrego1, jihaisse, misuba, DanC_, pindonga, moredhel, friedcell, sarek and julianf joined the channel
misuba, Hory_, mlncn, julianf and wolftune joined the channel
#
rhiaro
I tried protecting my tweets, but it lasted two days before the allure of twitter's 'impression' stats chart on the side became too strong
#
rhiaro
Guess I need to install some spyware on my site instead
snarfed joined the channel
#
@jkphl
Just got the banner for upcoming @indiewebcamp's w/ the new logo design \o/ Will bring it to Berlin in 2 weeks. :) https://twitter.com/jkphl/status/790922338288164864/photo/1
(twitter.com/_/status/790922338288164864)
julianf joined the channel
#
julianf
Yay wordpress-micropub v1.0! Seems like an important milestone.
#
aaronpk
yeah that's awesome!
#
aaronpk
snarfed++
#
Loqi
snarfed has 244 karma (240 in this channel)
#
julianf
In Quill -- that's you aaronpk, isn't it? -- I'm a bit in the dark about what format the 'content' field accepts. Plain text? HTML? Markdown? Whatever the micropub endpoint accepts?
#
julianf
I skimmed through the Micropub spec, and http://microformats.org/wiki/h-entry , but still not clear.
#
Loqi
[Tantek Çelik] h-entry is a simple, open format for episodic or datestamped content on the web. h-entry is often used with content intended to be syndicated, e.g. blog posts. h-entry is one of several open microformat standards suitable for embedding data in HTML/H...
#
aaronpk
julianf: hm sorry it's not clear. the "content" field is from h-entry. in microformats, it can be either text or HTML, buf if it's html then the value of content will be {"html":"<b>foo</b>","value":"foo"} whereas when it's plain text it's just a string
#
aaronpk
so in micropub, it's plain text by default, and there's optionally a way to post HTML (if the endpoint supports it)
#
Loqi
[Aaron Parecki] Micropub
#
aaronpk
the simple quill editor only posts the plaintext format, but there's also an HTML editor in Quill which will post that {"html":"..."} object
#
julianf
Thanks. s/by default/in its lowest common denominator/ ?
#
snarfed
PSA: i'm not doing anything with the domain freedom.io right now, and it's up for renewal soon. anyone want it for a project?
#
aaronpk
well, more like the UI of clients assumes plain text so they won't do anything fancy with it
ncollig joined the channel
#
aaronpk
some peoples' servers accept things like markdown or other variations of plaintext, and of course if you type markdown in quill's plaintext box and your server is expecting it, then it works fine.
#
julianf
Huh? Not sure what "client" you refer to. In my case the writing client software was Quill, and the reading client is WordPress's HTML rendition of the post on my blog.
#
aaronpk
the micropub client, so in this case quill
#
julianf
quill.p3k.io/new is a micropub UI for composing plain text content, and quill.p3k.io/editor is a micropub UI for composing rich text content, and other UIs can do plain or rich content as they choose, right?
#
aaronpk
correct
shiflett joined the channel
#
julianf
Thanks.
#
aaronpk
micropub provides a way for the client to tell the server whether the content is plain text or HTML
#
julianf
But there's no way to negotiate or check if other markup is supported (or what subset of HTML), I take it.
#
aaronpk
correct
Pierre-O joined the channel
#
julianf
I'm just trying to understand the boundaries of micropub. OK, that's good to know.
#
julianf
That's enough about that for now, I think.
ncollig joined the channel
#
julianf
likes your news self-hosting article.
#
aaronpk
not mine, i just posted it to indienews :)
#
julianf
Oh, oops.
#
julianf
Ben Borges' article.
#
julianf
http://www.databoxproject.uk/ looks closely aligned with IndieBox / UBOS. https://app.mozillafestival.org/#_session-950 this weekend at MozFest in London.
#
julianf
I'm thinking of going.
#
julianf
They and IndieWeb should talk to each other if they're not already.
ncollig joined the channel
#
julianf
A 3-yr research project with money and big sponsors, and an intention to build an "open-source personal networked device ... that collates, curates, and mediates access to an individual’s personal data ...".
#
julianf
I wonder what the key focuses will be.
#
petermolnar
that sounds hilarously vague
#
petermolnar
I wanted to go to mozfest, but I'm still not convinced it would worth it; looking at the sessions now
calumryan joined the channel
#
calumryan
petermolnar: I'm going to some of MozFest for a few sessions, haven't decided which yet
jmelesky, snarfed, StatelessCat, friedcell, tantek and wolftune joined the channel
#
tantek
good morning #indieweb!
#
aaronpk
morning!
funwhilelost joined the channel
#
bear
morning everyone, this is your friendly security ursine with a request that everyone on iOS update the the latest 10.1 version that dropped yesterday -- some important security fixes are in it
#
tantek
as compared to 10.0 or 9.x?
#
aaronpk
wow that's quite the list
ncollig and snarfed joined the channel
#
bear
that's compared to 10.0
#
snarfed
one critical fixed exploit can get you owned remotely by just rendering a jpeg in the browser
#
tantek
good, another reason I'm glad I didn't waste time upgrading to 10 :(
#
snarfed
we can try to avoid the whether to upgrade software debate...but tantek it's likely that exploit was in 9 too
#
aaronpk
i avoided the heartbleed bug on one of my super old servers because the bug hadn't been written yet in the version of openssl it had ?
#
tantek
snarfed, yup, definitely a YMMV thing. In my experience, best not to to upgrade major (integer) versions unless there is a very specific strong positive reason you need to. Because they always break things you were depending on.
#
tantek
the minor versions (decimal) versions however are always a good set of bug fixes, and I've never seen a decimal version upgrade break anything
#
tantek
that's iOS versions
#
tantek
OSX you have to shift the decimal point one to the right for actual semantics since the "10." has no meaning
#
tantek
or rather is just "marketing" "brand" etc.
#
snarfed
yup, it's a tradeoff of feature changes vs risk of getting hacked, losing accounts (/money), unknowingly letting your phone be part of a botnet like the one that took down Dyn the other day...
hs0ucy joined the channel
#
snarfed
security is definitely a tragedy of the commons thing, like vaccinating. there's real value in helping herd immunity.
#
snarfed
but i'm ok with disagreeing :P
#
tantek
snarfed, re: security, the decimal versions tend to fix those bugs
#
tantek
without actual citations, I will dispute that the major (integer) versions actually make security differences
#
tantek
also re: the one that took down Dyn, that was cited as being IoT devices like security cameras, I didn't see any reference to phones in any of the articles I read
#
snarfed
oh agreed. the problem is that afaik there's no 9.x that includes these fixes for 9, since apple doesn't do long term support for older ios versions
#
tantek
snarfed, there's no evidence these bugs were in 9. major security bugs often get introduce in integer version upgrades (another reason to avoid x.0 versions)
#
snarfed
(re Dyn, yes, hence the word "like." there definitely have been definitely phone updates.)
#
snarfed
tantek: there's no evidence that it wasn't in 9 either :P since apple doesn't say. other ios exploits have been reverse engineered and found in old major versions, though, so there's definitely precedent.
#
snarfed
(no, i don't have a link handy :P)_
#
tantek
snarfed, the evidence is that major versions add security bugs, e.g. the example aaronpk gave
#
tantek
so absent counter evidence I'm sticking with that
#
snarfed
oh definitely! bugs are added and found in all versions, new and old
#
snarfed
the openssl exploits are a great example of bugs found in almost decades-old releases
#
snarfed
both have happened in ios. we don't know when this specific one was introduced.
#
snarfed
your stance is definitely important for security ppl to understand though. lots of people (maybe most) prioritize functionality over security. we have to figure out how to work with that and still keep software secure. we mostly don't know how to do that yet.
wolftune and shiflett joined the channel
#
tantek
snarfed, indeed it's a challenging set of trade-offs for security people
#
tantek
I think a conservative security person stance here would be to insist on decimal security updates to software for the hardware lifetime of those devices.
#
bear
it's made harder when you have some items that can be updated on their own schedule and others that are part of a bundle (like all core iOS apps/features)
gRegorLove joined the channel
#
gRegorLove
Good morning, indieweb
#
snarfed
hey gRegorLove
#
gRegorLove
Hi snarfed
snarfed, mlncn, funwhilelost, mlncn_, shiflett and KevinMarks joined the channel
#
Loqi
[indieweb] "Yes, surely there has to be an easier way to do all of this, and surely more than two of us see the…" by ChrisAldrich on 2016-10-25 https://medium.com/@chrisaldrich/yes-surely-there-has-to-be-an-easier-way-to-do-all-of-this-and-surely-more-than-two-of-us-see-the-120d62247d13?source=rss-f9e3747f5480------2
Pierre-O and wolftune joined the channel
#
KartikPrabhu
!tell ChrisAldrich your HTML on the medium POSSE seems broken
#
Loqi
Ok, I'll tell them that when I see them next
snarfed, j12t, renoirb and KevinMarks_ joined the channel
#
tantek
ok so I think I'm pretty convinced of the workability of two brainstorms I braindumped into IRC yesterday
mlncn joined the channel
#
tantek
1. how to distinguish where a post was published *from* (e.g. Twitter tweet locations) vs. where a post is *about* (e.g. Instagram photo locations) https://chat.indieweb.org/2016-10-24#t1477341844757000
ncollig joined the channel
#
tantek
2. a proposal for how to minimally markup a checkin post such that the "check in" user intent is captured, with sufficient information, re-using existing formats / properties and only adding one property to h-entry, p-checkin: https://chat.indieweb.org/2016-10-24#t1477349895913000
#
tantek
so unless someone finds any gross flaws in the reasoning of those two brainstorms, I'm going to write them on the wiki as a next step. then once they're there, will attempt some prototyping of sorts.
#
Loqi
[indieweb] "Comment on Notes, Highlights, and Marginalia: From E-books to Online by Chris Aldrich" by Chris Aldrich on 2016-10-25 http://boffosocko.com/2016/10/24/notes-highlights-and-marginalia/#comment-32464
#
KartikPrabhu
tantek: one question, would it be better to use a more "Generic" name than "*-checkin", so that it might be reused in other things. For example "p-name" is generic enough to be used in "h-card" and also in "h-entry". Avoids proliferation of specialised properties
#
KartikPrabhu
I don't have a suggesstion though
#
tantek
KartikPrabhu: indeed proliferation of specialized properties is a reasonable concern, and ought to be minimized.
#
tantek
however, the "might be reused" is a horrible methodology to design things as it results in architecture astronomy
#
tantek
whereas if there were other documented (similar?) use-cases, then it may make sense to find more common / minimal solutions
#
KartikPrabhu
right. trying to think of examples similar to a checkin but failing
#
aaronpk
there's the whole "checking in" to media. watching and listening to things.
#
KartikPrabhu
maybe it is a special case that needs its own property
#
aaronpk
i don't really like the term "checkin" for referring to those, and feel it's mostly forced on that use case
#
aaronpk
also the actual apps that people made for those don't call them "checkins" either
#
aaronpk
heh https://trakt.tv/ calls it "scrobbling"
#
KartikPrabhu
right, there seems to many things like that "read", "watched", "listened"
#
tantek
aaronpk: agreed with the "forced on that use case" e.g. checking into beers (Untappd)
#
tantek
KartikPrabhu: yes - the general area of media consumption, whether passive or semi-passive seems like it could use additional brainstorming
snarfed joined the channel
#
tantek
as someone who posts /read /jam posts, I'm using plain text + emoji until there's a specific use-case that may require something more
#
tantek
(checkins do require something more since they have multiple pieces of information that need to be conveyed in a way that is otherwise ambiguous, e.g. photos/notes with location info)
#
tantek
one possible interesting case of the "checking into media" thing is the checking into a live broadcast case, where it is an actual "event" as it were, with a specific time (even start / end times) and a virtual "location" (whatever channel / frequency / URL the broadcast is occuring on)
#
aaronpk
oh speaking of which, some checkins on foursquare are checkins to events too!
#
tantek
interesting - in that case, would a u-checkin that was a URL to the h-event be sufficient (and work?)
#
tantek
the *existing* h-event
#
aaronpk
my most common example of that is checking in to movies at a theater
#
aaronpk
they do it for concerts too
#
tantek
or is there a need to distinguish when the event started, and when you actually got there (checked into it)
#
aaronpk
the venue manager can create events and when people check in at specific times, the app prompts the user to choose the event to check in to also
#
tantek
same with me with movies
#
tantek
aaronpk, on the indieweb, anyone can create events, and anyone could check into them :)
#
tantek
aaronpk your thoughts on start time of the event vs time of actual checkin? does it matter?
#
tantek
looks into parent events
wolftune and mlncn joined the channel
#
tantek
looks like iCalendar has a property RELATED-TO for linking a VEVENT to another VEVENT that has a default relationship type of PARENT
#
tantek
so... to answer your question aaronpk, if we wanted to capture ALL the information, e.g. what time you checked into the movie as distinct from when the movie *started*, your p-checkin h-event would have your actual checkin time and movie theater venue, then also have a u-* (TBD) property that would link to the actual Event (movie instance) you were checking into
renoirb joined the channel
#
tantek
normally we try to re-use terms from iCalendar, however "related-to" is pretty useless, especially since the actual meaning is captured in the "reltype" param of the "related-to" property
#
tantek
which starts to smell a lot like microformats.org/wiki/existing-rel-values
#
tantek
which are by design global, rather than scoped to a specific property like "related-to" (even if generically named, is scoped to iCalendar)
#
tantek
so it would be reasonable to introduce a new rel value like "parent-event"
#
tantek
(since there's a specific use case for it)
#
tantek
except that we now (and have for a while) frown upon mixing use of 'rel' and 'class' in a microformat (because it introduces sufficient confusion of when to use which attribute as to result in much more authoring confusing and bad data)
#
tantek
(same reason we switched from e.g. rel=in-reply-to to u-in-reply-to)
#
tantek
(and then didn't even bother with rel values for likes reposts etc. and use u-like-of etc.)
#
gRegorLove
I don't think most 4sq events are associated with a specific time, though. Maybe a date or date range at best.
#
tantek
gRegorLove: they are, it's just not shown to you
#
gRegorLove
Oops, missed some scrollback.
#
bear
isn't this one place where nesting of items works?
#
tantek
likely in the admin interface for venue admins
#
tantek
bear - yes, nesting is not the question, but rather, how to express the meaning of that nesting
#
Loqi
[indieweb] "? Want to read: Personal Archiving: Preserving Our Digital Heritage by Donald T. Hawkins" by Chris Aldrich on 2016-10-25 http://boffosocko.com/2016/10/25/%f0%9f%94%96-want-to-read-personal-archiving-preserving-our-digital-heritage-by-donald-t-hawkins/
#
tantek
so instead of rel=parent-event, we could propose a u-parent-event property for h-event
#
bear
oh, I thought the implied model like VCALENDAR uses would work
#
gRegorLove
movie times, even? What's the point if there's no UX for it?
#
aaronpk
setting the times there enables that event to appear in the app during those times
#
tantek
which then raises the question, can an event have a parent that is not just another event?
#
tantek
would u-parent be sufficient as a property for h-event to refer to a containing h-event
#
gRegorLove
For movies I've definitely seen options that are not showing at that time, which is what made me think it's a date or date range granularity
#
aaronpk
our indiewebcamp events could be considered to be under the umbrella of the indieweb organization
#
aaronpk
gRegorLove: they probably just set the time wrong
#
tantek
aaronpk: that's the p-organizer
#
aaronpk
how about an event series?
#
tantek
"parent" is a bit of a misnomer from the iCalendar spec unfortunately
#
aaronpk
is that also just an event?
#
tantek
it's not really a "parent" in the "creator of" sense
#
aaronpk
like a film festival
#
tantek
aaronpk - repeating events (series) are a whole different rathole
#
tantek
iCalendar would call those related-to, with reltype of SIBLING
#
aaronpk
no i mean like "this showing is part of this film festival"
#
bear
I would suggest avoiding repeating anythings
#
aaronpk
or like XOXO weekend, there were a bunch of events under XOXO, including our indieweb meetup, etc
wolftune joined the channel
#
tantek
"as part of"
#
tantek
containment appears to be the relationship that is being expressed most commonly
#
tantek
from the smaller (shorter) event to the larger one
#
tantek
so that's the naming challenge, come up with a term better than "parent" for, thing thing that this thing is a part of
#
tantek
"container" is too generic
#
tantek
"parent" is a misnomer, you're not part of your parent (human)
#
aaronpk
"part of"
#
tantek
"part-of" could work
#
tantek
"during"
#
tantek
though that sounds like a weaker tie
#
tantek
aaronpk - XOXO is a good example. the indieweb meetup was officially in the schedule thus "part of"
snarfed joined the channel
#
tantek
however, if there were informal meetups, then those may have been more like "during", like a small group lunch
#
tantek
rather than considered "part of" the larger XOXO event
#
tantek
back to the movie example, your checkin "event" is not technically "part of" the movie
wolftune joined the channel
#
tantek
your checkin "event" is more like it's *at* the movie
#
tantek
which then sounds like a temporal interpretation of p-location
#
tantek
thus perhaps a checkin to an event (which has its own venue) could look like
#
tantek
h-entry / p-checkin h-event / dt-start (checkin time), p-location h-event / dt-start (movie start time), p-location h-card (venue)
#
tantek
or presumably if the event was already posted somewhere (like a movie or concert)
renoirb joined the channel
#
tantek
h-entry / p-checkin h-event / dt-start (checkin time), u-location (URL to the actual event, movie, concert etc.)
#
tantek
so then the question is, is that too loose a use of p-location / u-location, or does it completely make sense that a "location" could be a spacetime location, and thus an h-event?
wolftune joined the channel
#
tantek
ok with that extended use question of p-location to the channel, off to do a few other things - discuss amongst yourselves!
KevinMarks, wolftune, mlncn, friedcell, snarfed, KevinMarks_, renoirb, shiflett and pindonga joined the channel
#
Loqi
[indieweb] ""Kevin writes a plea on Ev’s blog" - I often still think of Medium this way too. Though isn't Backchannel technically a Conde Nast joint" by Chris Aldrich on 2016-10-25 http://stream.boffosocko.com/2016/kevin-writes-a-plea-on-evs-blog---i-often