#indiewebcamp 2015-07-09

2015-07-09 UTC
#
aaronpk
oh my, my worker script got stuck trying to verify a webmention and now i'm processing a huge backlog
#
aaronpk
the spam pingbacks I get are pretty ridiculous
snarfed1 joined the channel
#
tantek
wondering if sss.stanford.edu is worth adding to /site-deaths
#
tantek
also wondering how to determine when it went down
#
tantek
via kylewm - the opening rhetorical question here is a great example of lack of selfdogfood: https://twitter.com/monteiro/status/618817474826498048
KevinMarks1 joined the channel
#
@tomcoates
@kevinmarks You should really publish it on your own weblog rather than on some stupid centralized publishing service.
(twitter.com/_/status/618936388092719104)
KevinMarks__ joined the channel
#
tantek
of some relevance to indieweb: http://mashable.com/2015/07/06/why-web-design-dead/ - in particular this assertion: “Symptom 4: Facebook pages as the new small-business homepage”
KevinMarks_ joined the channel
#
KevinMarks1
(see that whole thread - I may have indiesniped Tom)
KevinMarks__ joined the channel
#
kylewm
seems like all the twitter people i like just got mad at each other :(
#
tantek
kylewm - don’t worry, they’re all being jovial about it - not actually mad. (having heard most of them talk, even to each other, in person)
#
tantek
definitely a LOLworthy thread
#
KevinMarks1
tom may have given me my iwc project there
#
KevinMarks1
though that will mean actually templating my site
#
tantek
KevinMarks++ for indiesniping tomcoates of all people
#
Loqi
KevinMarks has 125 karma
#
KevinMarks1
also, about half the thread is now in my blogpost's comments, thanks to bridgy
#
tantek
kylewm: also, context, it’s quite late right now in the UK where adactio is, and that might be making him, particularly eager (not holding back) to reply on a Twitterfight.
KevinMarks_ joined the channel
#
tantek
do we have a place on the wiki where we document indieweb pain points?
#
tantek
more than just itches that is
#
@ade_oshineye
The punishment for self-hosting is eternal -damnation- maintenance.
(twitter.com/_/status/618874062333452288)
#
achangeiscoming.net
edited /2015/Guest_List (+311) "/* Participating */"
(view diff)
#
tantek.com
edited /archival_copy (+178) "services, internet archive,"
(view diff)
#
@bobbie
@kevinmarks @tomcoates @adactio are you using two services started by ev to tell me why a third service started by ev won't last?
(twitter.com/_/status/618943363065090048)
#
tantek
nevermind Odeo, and while not technically started by ev, funded by his incubator Obvious: Branch
#
aaronpk
what a crazy thread
#
@lucian
@davewiner I still prefer a network of #indieweb + RSS. #openweb
(twitter.com/_/status/618948230873661444)
#
GWG
Everyone counting down to IWC?
j12t joined the channel
#
tantek
interesting article / criticism of /Twitter found via that tweet thread: http://www.nytimes.com/2015/07/09/style/twitter-who-do-you-think-you-are.html?_r=0
#
tantek
many of the questions asked could be asked of any indieweb effort as well
#
tantek
GWG - a bit anxiously, but yes :)
#
GWG
I look forward to participating remotely.
lukebrooker joined the channel
#
tantek
GWG, you asked earlier in the channel about implying post kinds right? or am I misremembering?
tilgovi joined the channel
#
GWG
tantek: Not asked per se. It was one of my ideas for a solution to the post formats problem I was trying to come up with a proposal that people might get behind.
#
tantek
I believe implied post kinds is necessary for a variety of use-cases
#
aaronpk
oh man I'm getting pretty excited to turn that SD card into a micropub client
#
aaronpk
it arrives tomorrow, just in time for IWC
#
tantek
I’m wondering if we should set a goal to get 100% SWAT0 across 3 indieweb implementaitons for IWC 2015
#
aaronpk
i could be convinced to do that instead
#
tantek
(which AFAIK no one has ever done)
#
GWG
Post Formats in WordPress is this messy hybrid function that no one understands.
#
aaronpk
except argh my new site isn't ready to go yet
#
tantek
(Statusnet and Cliqset got *close* with two implementations)
#
tantek
(I think there’s a YouTube video that shows it working supposedly)
#
tantek
but it didn’t last, nor did it get any uptake beyond the demo
#
tantek
aaronpk - I’d say getting your new site hackable worthy should be your #1 coding goal in the next 48 hours.
#
tantek
time to take a another look at SWAT0 to see what’s left undefined
#
tantek
what is SWAT0?
#
Loqi
SWAT0 is an abbreviation for The Social Web Acid Test level 0 and is an user-feature interoperability test for social web functionality defined in 2010 https://indiewebcamp.com/SWAT0
#
tantek
How’s this for incentive for achieving 100% reproducible, selfdogfooded 3 person and sites and implementations SWAT0 at IWC 2015:
#
tantek
The first 3 people that get it working *with each other*, we update the /SWAT0 summary with THEIR NAMES and so it shall be set (as long as they maintain that functionality at a minimum on their site).
#
aaronpk
hehe i like it
#
tantek
open to remote attendees too of course, even more impressive if all 2+ geos are involved
#
tantek
right now I think ben_thatmustbeme is the only one that has person-tagging in posts working
#
aaronpk
IIRC only one person needs to be able to create a person-tag post, correct?
#
tantek
so he’s the current leading candidate for (A) here: http://tantek.com/2015/029/t1/swat0-posts-tags-mobile-photo-comment
#
@t
SWAT0 A posts+tags mobile photo of B B photo notified C(follows A) sees it; replies A&B comment notified Prev: http://tantek.com/2015/029/t1/swat0-posts-tags-mobile-photo-comment
(twitter.com/_/status/560921112767242241)
#
tantek
aaronpk, correct
#
tantek
however we currently have zero known examples of indieweb folks getting push notifications of person-mentions https://indiewebcamp.com/person_mention#IndieWeb_Examples
drz3 joined the channel
#
tantek
(the *push* aspect of the notification was very much the intent in step 2)
#
tantek
original intent of SWAT0
#
loqi.me
created /communicator (+51) "prompted by tantek and dfn added by tantek"
(view diff)
#
tantek
aaronpk, since you have notifications working, and close to getting push notifications working I suspect you’re the leading candidate for player B.
#
tantek
and player C, I think the leading candidate may be kylewm - since he has a reader
#
aaronpk
agreed
#
tantek
pretty sure kylewm can post a reply to a post directly from his reader via micropub
#
tantek
which will send a webmention to the photo publishers (e.g. Ben)
j12t joined the channel
#
tantek
however, the notification to aaronpk (person-tagged subject of the photo) about the comment on the photo may depend on Ben supporting salmentions
#
tantek
and for that matter a previously undefined form of salmentions
#
tantek
when someone comments on your post and sends you a webmention, you have to send a webmention onward to anyone person-tagged in the post
#
tantek
then *their* person-mention notification system has to know to receive salmentions, that is not only know about just “someone posted a photo that you’re in” BUT “someone commented on a photo that you’re tagged in"
#
KevinMarks1
when I visited aaron last time, I webmentioned him and his watch buzzed
#
tantek
aaronpk - at that github summary this: “Because this is an update to the contents of the URL, Ben's server re-sends the original Webmentions to indicate there was an update to the receiver.” is now what we call a /salmention sender!
#
tantek
not sure if ben_thatmustbeme’s software supports sending salmentions
#
tantek
aaronpk: everything else you have written up in SWAT0.md aligns exactly with what I said above - so I just came up with the same thing you wrote. Good verification of your analysis.
#
tantek
the only change I would suggest is explicitly calling out the “re-sends the original Webmention to indicate there was an update” as supporting salmention sending
#
aaronpk
i chose the names for that writeup based on who I thought was closest in each category, but that has changed now
KartikPrabhu joined the channel
#
tantek
yes, don’t change the names yet
#
tantek
this also means you’ll have to support *receiving* a salmention
#
aaronpk
how do I receive a salmention?
#
tantek
this in particular in your write-up: “It sees there is a new property, comment, containing an h-cite.” is essentially support for receiving salmentions
#
aaronpk
oh I wrote that didn't I
#
tantek
support for receiving salmentions is nothing more than looking for *more* information when you receive a webmention
#
aaronpk
looks like /salmention needs a #How_To header
#
tantek
and then passing that along, exactly as you wrote it up: "Barnaby Walters commented on Ben Werd's photo of you"
#
tantek
the challenge here is that the only people supporting sending and receiving salmentions are acegiak and kylewm
#
tantek
yes, salmention was the last remaining building block piece missing from completing an indieweb explanation of how SWAT0 works
#
aaronpk
now I want a summary of SWAT0 showing what each person needs to implement to fill the role
#
Loqi
gives aaronpk a summary of SWAT
#
tantek
ok I think I can help with that
#
aaronpk
gives Loqi a summary of SWAT
#
Loqi
grabs the summary of SWAT
#
aaronpk
keep it, you'll need it later
#
tantek.com
edited /SWAT0 (+1315) "resolved the last undefined piece by realizing that's exactly what salmentions (sending/receiving) do - invented/named since the last time we assessed what's in/missing in SWAT0 for indieweb"
(view diff)
#
tantek
and with that, the last UNDEFINED thing in IndieWeb SWAT0 is now defined
#
tantek
we have a solution for how to implement SWAT0 using indieweb protocols, now to see if we can the pieces implemented
mdik joined the channel
#
aaronpk
were you going to add a summary of what each person (role) needs to support?
#
tantek
braindumping that now
#
aaronpk
perfect
#
tantek
waves his hands over the keyboard
#
acegiak
Ok so here's a novel thing I've done:
#
acegiak
Amygdala is the platformer I've developed and I'm preparing for Steam launch so I'm re-examining high scores and achievements
#
acegiak
the current system is that anyone can enter any name they like when submitting high scores and achievements
#
acegiak
because they're good things pretending to be someone else doesn't really hurt them
#
acegiak
BUT some people might want to keep their record "clean"
#
acegiak
so now, on the high scores and achievement pages if the username is a domain name I add a delete button
#
acegiak
and then if you click that it will attempt to auth with indieauth.com and if successful, delete the record
#
acegiak
plan to add twitter auth and email tokens later but indieauth was easy to set up as proof of concept
#
tantek
cannot process, swap space locked with SWAT0 processing
#
acegiak
tantek: I put an indieauth.com powered delete button on my game's high-score board
#
Jeena
something went wrong with your page acegiak https://jeena.net/t/amygdala.png
#
tantek.com
edited /SWAT0 (+1413) "first draft of implementation requirements for each of the SWAT0 players ABC"
(view diff)
#
tantek
so the race is on, who will be first to implement all of A, or all of B, or all of C?
#
acegiak
jeena, can you hit it again so I can watch the error log as it flies past?
#
ben_thatmustbeme
i think i'm pretty close to A
#
tantek
ben_thatmustbeme: yes we think you’re in the lead for A
#
tantek
and aaronpk for B
#
tantek
and kylewm for C
#
ben_thatmustbeme
i don't do salmentions yet
#
tantek
you only have to support half of them
#
tantek
the sending half
#
ben_thatmustbeme
i also need to check that mobilepub can do person tagging on photos
#
tantek
which I think is the easier half
#
ben_thatmustbeme
also, my notifications aren't really correct, i get a notification there was a mention, thats all
#
tantek
really hoping kylewm is going to be able to participate remotely
#
aaronpk
both kylewm and I both fulfill C right now
#
aaronpk
tantek++ this is great
#
Loqi
tantek has 212 karma
#
acegiak
tantek: Interesting. I don't do point two of the C section because I expect the webmention I pass to the reply post to be forwarded upstream
#
ben_thatmustbeme
although we don't have 2d point tagging working at all
#
aaronpk
point tagging isn't important
#
ben_thatmustbeme
that became a mess in micropub as it was an embedded object
#
tantek
point tagging was never specified in SWAT0 and I think the consensus was that it was not necessary nor intended
#
ben_thatmustbeme
okay, headed to bed
#
ben_thatmustbeme
maybe tomorrow morning on the train ride in i'll get started on salmentions
#
ben_thatmustbeme
although, I'm so close to getting micropub chaining working. and i have just been too busy to finish it
#
tantek
acegiak++ for happening to describe, and name the last missing (undefined) piece we needed for SWAT0 indieweb building blocks!
#
Loqi
acegiak has 25 karma
#
acegiak
:P ta!
#
tantek
you had no idea you were creating such an essential building block did you? or maybe you did :)
#
acegiak
I knew it was an important feature that was missing from the experience I was wanting
#
acegiak
that's all
#
KevinMarks
that is the key
#
KevinMarks
seeing and documenting the gap
#
tantek
KevinMarks - I’m good at identifying gaps like that, but the solution to the gaps is sometimes so far away from personal itches / working on, that it’s hard to figure that out
#
tantek
however I can see when pieces fit together. thank you LEGO
#
acegiak
is trying to work out a way to authorise game players /before/ high score submission using indieauth
#
acegiak
I should wait until my meeting with my security friend before letting my mind run too wild
lukebrooker and drz3 joined the channel
#
aaronpk
ben_thatmustbeme: could you do me a favor and /person-tag me in a post?
#
aaronpk
tantek: the SWAT0 summary "aaronpk gets a person mention notification that he's been tagged in a photo" shouldn't that actually say "aaronpk gets a person-tag notification"?
#
tantek
both are true
#
aaronpk
person-tag is a subset of person mention
#
aaronpk
but person mention is not sufficient for SWAT0
#
tantek
a person-tag notification is one form of person mention notification - see list here: https://indiewebcamp.com/person_mention#Why
#
tantek
other way around - person-tag is sufficient for SWAT0
#
tantek
no need to support all of person-mentions
#
aaronpk
i think we are saying the same thing
#
aaronpk
in SWAT0 the specific "was tagged in a photo" is the key, not just "was mentioned in a post"
#
tantek
agreed
#
tantek
fixing
#
tantek.com
edited /SWAT0 (+137) "specify person-tag mentions when the detail is important above and beyond just a person-mention"
(view diff)
#
tantek
aaronpk: fixed
#
aaronparecki.com
edited /SWAT0 (+8) "linky"
(view diff)
#
aaronpk
tantek++
#
Loqi
tantek has 213 karma
#
aaronpk
now reading "B must support"
#
aaronpk
does that mean a notification like "was tagged in a post" is sufficient?
KevinMarks__ joined the channel
#
Loqi
slack/snarfed: hey acegiak, will you eventually add your salmention implementation to the wp webmention or semantic linkbacks plugins?
#
GWG
snarfed, I asked acegiak that as well.
#
GWG
acegiak: I just asked where it might fit the best
#
acegiak
I think the SENDING of salmentions got rolled back in one of my git fuckups
#
acegiak
but if you're using that version of semantic linkbacks you're accepting salmentions
#
acegiak
admittedly the implementation is bona-fide not the right way to do it but it works
#
Loqi
slack/snarfed: awesome thanks!
#
tantek
aaronpk - hence the possible newly raised importance of http://indiewebcamp.com/implied_post_types - so that you can determine that it’s a photo post
#
aaronpk
whoa where'd this come from
#
tantek
it’s quite old with lots of minor iterations over time
#
aaronpk
interestingly this is basically what I am doing in the implementation of my new site
#
tantek
aaronpk: not surprising, see the “Articles about …” ;)
#
aaronpk
jeez that was from 2012?
#
aaronpk
your original 2010 description of SWAT0 did say "B gets notified that they are in a photo"
#
tantek
very well, then apparently Post Type Discovery is also a necessary building block, good thing it’s been gestating for 3+ years as a subsection of the /posts page!
tilgovi joined the channel
#
tantek
happy to leave it as a subsection of the /posts page until someone implements it and documents why
#
aaronpk
SWAT0 would be a good reason why!
#
tantek
nevermind there’s an outstanding action for me to specify it for Social Web WG
#
tantek
maybe I should suggest it as its own Working Draft there
#
tantek
separate from AS2, so it doesn’t need to have all the baggage
#
aaronpk
interestin
#
aaronpk
okay so reading https://indiewebcamp.com/person-tag ... in order to consume a person-tag I need to check if any of the values of "category" for the post are an h-card with my site as the url?
#
tantek
or if any of the links to your site are a u-category h-card
snarfed joined the channel
#
tantek
or maybe you’re thinking from the perspective of using the parsed JSON output?
lukebrooker joined the channel
#
tantek
because the intent was you could just do it as part of webmention target verification
#
aaronpk
my webmention handling code works like: look for <a> tag with href = my site, then parse for mf2 and possibly upgrade it from a generic mention to something more specific
#
aaronpk
so yes, thinking from the perspective of working with the parsed JSON output
#
tantek
ah you’re using the “parse for mf2” as the method to upgrade from generic mention
#
tantek
ok that makes sense
#
aaronpk
yep that was the easiest
gRegorLove, snarfed, yakker, Erkan_Yilmaz, tilgovi, KartikPrabhu and snarfed1 joined the channel
#
aaronpk
wow getting notification text right is subtly hard
#
aaronpk
ah there's a very different pattern for notifications of your home page!
#
aaronpk
"{author} tagged you in a {source post type}" vs "{author} liked your {target post type}"
#
aaronpk
for regular mentions, I want to know more about the thing on my site being mentioned
#
aaronpk
for home page mentions, I want to know more about the post that mentioned me
KevinMarks_ joined the channel
#
aaronpk
so that happens now
#
@mikkelz_za
RT @SaraSoueidan: #TIL about IndieWeb POSSE: Publish (on your) Own Site, Syndicate Elsewhere. http://indiewebcamp.com/POSSE —this is why I love T…
(twitter.com/_/status/619004610523594754)
KevinMarks__, snarfed and tantek joined the channel
#
aaronpk
whee now I can receive person tags
#
aaronpk
and have notification text that says "{author} tagged you in a photo" as well as "{author} liked your post"
#
aaronpk
this should improve my experiences with bridgy greatly :)
#
snarfed
aaronpk++
#
Loqi
aaronpk has 871 karma
#
tantek
aaronpk: re: “getting notification text right is subtly hard” yes! hence the importance of text first design!
#
tantek
or at least a subset thereof?
#
aaronpk
a small subset!
gRegorLove joined the channel
#
tantek
aaronpk - would be really interesting to know what subset
#
tantek
implementer subsetting is one of the best forms of feedback for aspirational specs
tantek, gavinc, KevinMarks_, Zegnat, sanduhrs, Jihaisse, lukebrooker, KevinMarks__, eschnou, glennjones, friedcell, petermolnar, yakker, csarven, keroberos, evalica, mapkyca, sivoais, Sebastien-L, tvn, frzn and adactio joined the channel
#
adactio.com
edited /2015/Brighton (+209) "/* Participating */"
(view diff)
yakker, Sebastien-L, mdik_, lukebrooker, j12t, mlncn and cweiske joined the channel
#
cweiske
kylewm, I shrunk the setup instructions for the tt-rss subtome plugin quite a bit now:
#
cweiske
1. Go to Tiny Tiny RSS preferences.
#
cweiske
that's it
yakker and indie-visitor joined the channel
#
Loqi
Welcome, indie-visitor! Set your nickname by typing /nick yourname
#
cweiske.de
edited /SubToMe (+144) "/* Indieweb Examples */"
(view diff)
yakker joined the channel
yakker and friedcell joined the channel
#
@pypi_updates
django-webmention 0.0.1: A pluggable implementation of webmention for Django projects. https://pypi.python.org/pypi/django-webmention/0.0.1
(twitter.com/_/status/619110252395536384)
yakker joined the channel
#
@artwisanggeni
#python django-webmention 0.0.1: A pluggable implementation of webmention for Django projects. https://pypi.python.org/pypi/django-webmention/0.0.1
(twitter.com/_/status/619114420791218177)
yakker and elf-pavlik joined the channel
#
GWG
Morning, all
yakker joined the channel
#
@PythonJo
PyPI latest- django-webmention added to PyPI https://pypi.python.org/pypi/django-webmention #jo #amman #JPY
(twitter.com/_/status/619116806570033152)
yakker, frzn and j12t joined the channel
#
ben_thatmustbeme
kylewm: if I'm reading salmention description right, why do you poll from the comments on the reply and not from the original?
parzzix, yakker, wolftune, jdcauley and scoates joined the channel
#
kylewm
ben_thatmustbeme, there shouldn't be any polling involved. Just checking for updates on a reply when you receive another webmention from it
#
@Ubikuity
Google just started to offer personalized Gmail address (you@yourdomain.com) for $1 per user per month #OwnYourData https://twitter.com/Ubikuity/status/619139516373278720/photo/1
(twitter.com/_/status/619139516373278720)
#
petermolnar
that has nothing to do with ownyourdata az all :((
#
@petermolnar
@Ubikuity this has nothing to do with #ownyourdata; hell, this is the opposite, not even knowing you're sending to Google's all seeing eyes
(twitter.com/_/status/619139938525822976)
#
GWG
petermolnar: It means Google has gone from free provider to service provider. Google Apps id not something a single user would use for their one email
#
petermolnar
sure; this not #ownyourdata
#
GWG
petermolnar: It is in a data portability dense, but overall no
#
GWG
sense not dense
#
petermolnar
your keyboard caught a cold :) ( 'you're not fine when you say a D at the end of fine' Friends )
#
GWG
I have a cold.
#
petermolnar
get well soon, then
fourtonfish joined the channel
#
GWG
But I make typos more on mobile
#
ben_thatmustbeme
kylewm: so its not "forward the webmention up the tree" its "resend the same webmention as before"
#
ben_thatmustbeme
if B is a reply to A and B receives a reply C, then B send A we WM of target=A, source=B
Sebastien-L joined the channel
#
ben_thatmustbeme
i was thinking it would send WM(t=B,s=C)
#
ben_thatmustbeme
yeah, not polling, was on my phone on the train
#
ben_thatmustbeme
and brain still waking up
#
@jechtom
RT @Ubikuity: Google just started to offer personalized Gmail address (you@yourdomain.com) for $1 per user per month #OwnYourData http://t.…
(twitter.com/_/status/619143629798510592)
#
ben_thatmustbeme
did i just writing Webmention math?
#
kylewm
ben_thatmustbeme, yep that's right. Effectively when you receive a wm from a downstream, you send one upstream. I think that's all that was meant about forwarding up the tree
#
ben_thatmustbeme
wait, that doesn't answer my question though
#
ben_thatmustbeme
what is the exact webmention i am sending
#
ben_thatmustbeme
i reply to aaron, you reply to me
#
ben_thatmustbeme
when i process your reply
#
ben_thatmustbeme
is send to aaron a webmention
#
Loqi
ben_thatmustbeme meant to say: I send to aaron a webmention
#
ben_thatmustbeme
do i send (source=ben, target=aaron) or (source=kyle, target=ben)
#
Zegnat
Question in Math: when B receives WM(t=B,s=C), should B send WM(t=A,s=C) or WM(t=A,s=B)?
#
ben_thatmustbeme
i had assume the latter, as then you can build infinitely down (or a person replying can send the same WM all the way up, the middle doesn't need to support salmentions at all
#
ben_thatmustbeme
Zegnat: close, but i figures (t=B,s=C) not (t=A,s=C)
#
ben_thatmustbeme
can make that even more concise, WM(C->B) or WM(B->A)
#
ben_thatmustbeme
not sure WM(C->A) would work
#
ben_thatmustbeme
well, you can process the in-reply-to from C and then check if B is something you have already parsed or not, so yes, i suppose that could work
#
Zegnat
Alright, trying again, the question marks are discussed here, right?: A = Post; B = Reply to A; C = Reply to B; B receives WM(C->B); B sends WM(?->?) to C.
#
ben_thatmustbeme
a/B sends WM(?->?) to C/B sends WM(?->? to A
#
ben_thatmustbeme
tantek is going to look at this later and be like... WHAT??
#
ben_thatmustbeme
i think all 3 options can work (B->A), (C->B), (C->A)
#
ben_thatmustbeme
(B->A) is certainly simplest for B, but for A it requires processing for edits and for all replies
#
ben_thatmustbeme
C does not need to support Salmentions in that case
Sebastien-L joined the channel
#
ben_thatmustbeme
and A only needs support reprocessing replies and displaying responses to replies (and possibly resplies to replies to replies to ....)
#
Zegnat
I think, reading /Salmentions, it goes: C says it replies to B, WM(C->B), B parses the reply and *includes it on its page*, then B tells A it has been updated by resending WM(B->A). The wiki says that A is supposed to check the first reply and parse “the second reply (which is displayed as a comment on the first reply)”
#
Zegnat
So as described on /Salmentions it only works if B actively displays replies on its post pages
#
ben_thatmustbeme
yes, thats what i'm guessing
#
ben_thatmustbeme
but it does get confusing once you go beyond 2 levels i think
#
kylewm
yes I agree wtih Zegnat, you resend the same original webmention B->A
#
ben_thatmustbeme
when it percolates all the way back up from WM(Z->Y) then B is displaying a complex tree
#
ben_thatmustbeme
that A then has to know what to do with
#
Zegnat
yep, and if K displays replies in a flat style instead of a tree then all levels of comments are reset. Very vulnerable system
#
ben_thatmustbeme
indeed, that can get messy
#
ben_thatmustbeme
for my storage system it would be much easier if i got a WM(C->B)
#
Zegnat
kylewm, you are one of the two only implementations. Have you thought about handling multi-level replies yet?
#
Loqi
[mention] Tino Kremer commented 'Why I write on my own site first 4 min read There are many s...' on a post that linked to http://indiewebcamp.com/ (http://stream.tinokremer.nl/2015/why-i-write-on-my-own-site-first)
#
ben_thatmustbeme
since i already have B stored, i just need to fetch C
#
kylewm
Zegnat: it just shows a flat list, sorted in the order received
#
Zegnat
ben_thatmustbeme: I originally thought that was the idea, to literally “pass along the WM”. So if I receive WM(C->B) I also send WM(C->B) to A
friedcell and tantek joined the channel
#
kylewm
it seems pretty weird to send a webmention to A where target is not on A; though youcould definitely make it work that way, that seems like a much bigger departure from how webmention usually works
#
parzzix
just saw this conversation...holy confused I am.....lol
#
Zegnat
I was actually reading the /Webmention page to see if implementations should even accept non-self targets
#
Zegnat
“The receiver SHOULD check that target is a valid resource belonging to it […]” - might require a spec change to allow non-self targets
#
Zegnat
ie. resending WM(B->A)
sanduhrs joined the channel
#
tantek
reads logs
#
aaronpk
reads logs
#
tantek
kylewm clearly “forward” is open to easy misinterpretation in the context of salmentions, and thus we probably shouldn’t use the term to describe what is happening
#
aaronpk
tantek: currently I determine post type by looking for the h- type and then presence of certain properties, in this type order: event, drink, food, sleep, video, photo
#
aaronpk
i need to add 'article' next, but that is more complicated because it involves normalizing and comparing the "name" vs "content" properties
#
tantek
kylewm is correct re: “it seems pretty weird to send a webmention to A where target is not on A”
#
tantek
that breaks existing webmention, and adds unnecessary complexity to webmention processing
#
kylewm
tantek: ben_thatmustbeme: does it say "forward" somewhere on the wiki?
#
tantek
kylewm checking
#
tantek
(I just saw it in IRC)
#
tantek
nor in any combination with “webmention” on any page.
#
aaronpk
haha "tantek is going to look at this later and be like... WHAT??"
#
Loqi
awesome
#
Zegnat
What is SWAT0?
#
Loqi
SWAT0 is an abbreviation for The Social Web Acid Test level 0 and is an user-feature interoperability test for social web functionality defined in 2010 https://indiewebcamp.com/SWAT0
#
Zegnat
Thanks Loqi
#
tantek
aaronpk: it’s ok, I’ve developed a complexity filter ;)
#
Loqi
you're welcome
skalpin joined the channel
#
tantek
!tell acegiak: re: http://indiewebcamp.com/irc/2015-07-08/line/1436409419389 “forward” appears to be open to easy misinterpretation [1] in the context of salmentions, and thus we probably shouldn’t use the term to describe what is happening. [1] http://indiewebcamp.com/irc/2015-07-09/line/1436450175162
#
Loqi
Ok, I'll tell them that when I see them next
#
tantek
kylewm: reference found ^^^
chalettu joined the channel
#
kylewm
gotcha
KevinMarks_ joined the channel
#
ben_thatmustbeme
so we are OK with 1 person doing flat responses instead of threaded messing everything up?
#
aaronpk
i don't think it matters if someone displays flat responses vs threaded
#
aaronpk
won't the algorithm work regardless?
#
ben_thatmustbeme
well, yes, but all those response when processed up, loses all threading
#
ben_thatmustbeme
unless you look at every reply and process that way
#
aaronpk
if each comment had an in-reply-to property when displayed as a comment you could put the threading back together if you want
scoates joined the channel
#
aaronpk
personally i don't ever want to display threaded comments, I think the fact that most systems have moved away from that is telling
#
aaronpk
but I would like to show/link to the context for each comment because that's helpful when discussions get really long
#
Zegnat
Facebook is moving towards threaded comments instead of away from them
#
aaronpk
only one level deep
#
ben_thatmustbeme
thats fine when you are only doing replies to your post, but now we are taking replies from further down
#
ben_thatmustbeme
it could get confusing
#
Zegnat
aaronpk, true, only 1 level
#
aaronpk
anyway point is, displaying threaded vs flat vs some hybrid is a personal preference, and isn't actually affected by how salmention works
#
Zegnat
as long as in-reply-to is included display doesn’t really matter, otherwise salmention is affected because the receiver is supposed to get the whole comment thread from the sender. So the way the sender includes comments on their site has a direct effect. That’s what ben_thatmustbeme was afraid for as weakness, if I understood right
#
aaronpk
but you can always go fetch the original
#
Zegnat
that would require you to “reindex” every single comment in the thread every time you get an update mention, because you don’t know which one triggered the update
#
aaronpk
i don't think so
#
yellowrabbitapps.co.uk
edited /2015/Brighton (+177) "/* Participating */"
(view diff)
#
aaronpk
if I get a webmention from your page, I would just check the contents of that page to see which comment is new
#
aaronpk
then I could go fetch that comment's URL to find the canonical data for it including in-reply-to
chreekat joined the channel
#
Zegnat
Basically treating the comment list on a post only as a list of permalinks?
#
tantek
either approach is reasonable
#
tantek
build what’s easier for you to get working and see what happens!
tilgovi joined the channel
#
Zegnat
I could definitely see that working, because then all the actual data you are using will be from the individual reply pages
#
tantek
yup. tradeoff data integrity vs. # of HTTP requests etc.
#
tantek
best solved by experimentation :)
#
aaronpk
my personal preference would be to go fetch the comment from the source, since that means the whole page would also get archived by my system
#
tantek
hey anyone else use a legacy feed reader and see <div xmlns="http://www.w3.org/1999/xhtml">CSS… in the title of my recent blog post?
#
ben_thatmustbeme
i would fetch from source as well
#
ben_thatmustbeme
s/fetch/prefer to fetch/
#
tantek
I just got an email bug report from a Feedly user
#
Loqi
ben_thatmustbeme meant to say: i would prefer to fetch from source as well
#
tantek
so I replied with a proof of my feed being correct / valid per spec and validator and asked them to file a bug report against Feedly :)
#
Zegnat
I don’t see any problems in my feed reader, Tantek
#
tantek
Zegnat cool :)
#
Zegnat
tantek: http://i.imgur.com/MwtiL3k.jpg , looks the way I expected it to
#
@kevinmarks
“The Awl makes a virtue out of its narrow audience when talking to advertisers, calling their readers "indielectuals" http://www.theverge.com/2015/7/9/8908279/the-awl-profile-choire-sicha-john-herrman-matt-buchanan
(twitter.com/_/status/619156883899703296)
snarfed, j12t, j12t_, elf-pavlik, shiflett, Sebastien-L and KevinMarks__ joined the channel
#
KevinMarks__
Facebook's reply threading is a way of not sending notifications
#
KevinMarks__
If I reply to your fb post normally, and you respond in the same way, I get notified
#
KevinMarks__
If you reply directly to me, I get notified
#
KevinMarks__
But if you reply directly to someone else's reply, I don't
friedcell joined the channel
#
chreekat
KevinMarks__: I dislike that about reddit, too. Esp. on reddit, where a reply-to-a-reply-to-me does not notify me.
#
chreekat
I like threading, but i like explicitly opting out of conversations more
#
ben_thatmustbeme
on G+its the other extreme
#
KevinMarks__
So if you want to emulate their threading, you would selectively send comments triggered webmentions
#
ben_thatmustbeme
if i reply to Kevin's post and then someone else replies to Kevin's post, i get notified
#
ben_thatmustbeme
which is why they added a mute button for that
#
ben_thatmustbeme
G+ is more like "if i reply to a post then i am actually subscribing to all updates on that post"
#
ben_thatmustbeme
at least it was, i don't know if they have fixed that recently, but i stopped using it mostly. mainly because there is no write API
snarfed, tvn, Sebastien-L, niversen, mlncn, jciv, gRegorLove, hless, tilgovi and KevinMarks_ joined the channel
#
KevinMarks_
Interesting from slack extending like to all emoji http://slackhq.com/post/123561085920/reactions
#
KevinMarks_
Given that my face piles have ♥ and ♻ already thanks to webmention.herokuapp.com
j12t joined the channel
#
Zegnat
That is an interesting move by Slack. It does finally allow for 👎 to be used
cmhobbs and tilgovi joined the channel
#
KevinMarks_
Also 🚂 and 📛
#
Loqi
slack/kylewm: this feature seems... unnecessary :stuck_out_tongue:
#
Loqi
slack/aaronpk: Doesn't really translate to IRC unfortunately
friedcell joined the channel
#
aaronpk
I kind of like the idea of putting emoji reactions on my own post, and allowing people to click them
#
KevinMarks_
And custom emoji, so zoidberg and left shark show up
#
aaronpk
i could allow people to click them after they sign in with indieauth
#
KevinMarks_
One answer to star versus heart is all the stars and hearts
#
kylewm
-1 to a fixed set of emoji though, sounds very buzzfeed-y (trashy, lol, wow, etc.)
#
kylewm
i noticed that nytimes article about twitter took an anti-"favorite", pro-"like" stance
#
aaronpk
ooh I could let people click from the list after signing in, but if you wanted to add a new emoji to the list, you'd have to write a reply post and send me a webmention :)
#
aaronpk
and of course accept votes as webmentions too
#
KevinMarks_
Is the webmention a u-like-of?
#
aaronpk
hm, not really
#
gRegorLove
I'm pretty pro-heart, pro-favorite, anti-like
#
Jeena
So which time does it start tomorrow?
#
Jeena
(in Portland)
#
aaronpk
the pre-party in portland is 4:30-6:30
#
KevinMarks_
So, how do we mark up an emoji reaction?
#
aaronpk
hm i should probably send out an email summary of the schedule and stuff, since i have a bunch of email addresses from the ti.to registrations
#
bret
Zegnat: modern irc clients support emoji
#
bret
👍 👻
#
kylewm
man, Woodwind has started taking a really long time to load, and I don't know why
#
KevinMarks_
Use chrome's profiler?
#
kylewm
oh it's almost all database calls on the backend
#
kylewm
i just can't figure out why they're so slow all of a sudden
#
rhiaro
wonders vaguely about people sending rectangles to each other
#
rhiaro
's irc client is obviously less sophisticated than others
#
KevinMarks_
What's the backend?
#
KevinMarks_
hasn't looked at wordwind source
snarfed joined the channel
#
KevinMarks_
⬜ for rhiaro
#
kylewm
KevinMarks_: it's Flask with SQLAlchemy connected to Postgres
#
rhiaro
<3 []
#
csarven
kylewm If it was SPARQL, I would ;)
#
aaronpk
i'm sure there are some guides tha would help
KevinMarks_ and fourtonfish joined the channel
#
csarven
It is just merging two tables on some columns
#
aaronpk
i'd just go make sure you have indexes on everything you're using to join and in where clauses
#
bret
kylewm: you might try implementing new relic's trial to see if you get hints
#
kylewm
++ everyone for not giving me a hard time for /database-antipattern
#
ben_thatmustbeme
index by your IDs you are joining on, and by anything you have in where clause and ORDER BY
#
aaronpk
ironically this is actually a case where a relational database is not always the best solution :P
#
aaronpk
read about how twitter constructs peoples' home timelines and you'll see what i mean
#
Zegnat
bret: I know IRC support emoji, I wrote the thumbs down here on IRC. But people have been asking for a thumb-down next to thumb-up on Facebook for years, that’S what I was getting at
#
aaronpk
also this is where memcache came from, made by bradfitz for livejournal
#
bret
ohhh lol
#
KevinMarks_
Woodwind was responding well for me, so may be a function of who you're subscribed to
#
csarven
emoji--
#
Loqi
emoji has -1 karma
#
csarven
--emoji--
#
csarven
Strong dislike on the idea of the OS taking over text chars
#
csarven
goes outside for fresh air
#
aaronpk
taking over? the OS is always responsible for displaying characters...
#
Zegnat
The OS doesn’t even impose emoji, the Unicode Consortium does
#
kylewm
adding a couple of indices seems to have helped a bit
#
kylewm
KevinMarks_: oh that's interesting
#
kylewm
considers periodically shuffling off old entries into an archive table...
#
KevinMarks_
Slowest for me was 60ms
#
KevinMarks_
the 2 stages of database slowness:
#
KevinMarks_
1st your data stops fitting into ram, table scans slow down 3 orders of magnitude
#
KevinMarks_
You add indexes, continue
#
KevinMarks_
Later, your indexes stop fitting into ram, you slow down another 3 orders of magnitude
benwerd joined the channel
#
gRegorLove
Out of curiosity, how many feeds are you subscribed to kylewm?
#
KevinMarks_
Hm, so is there any way I could handle salmentions when my webmentions are run through js insertion
friedcell joined the channel
#
KevinMarks_
Would webmention.herokuapp.com have to synthesise them?
#
KevinMarks_
Adding memcache is a way of dedicating ram on other machines to stop you churning the database's cache
#
kylewm
gRegorLove: 97 (!)
#
gRegorLove
I'm up to 30.
#
gRegorLove
I need to sift through my old OPML file and set up a 'following' page. Then it will go way up.
#
gregorlove.com
edited /User:Gregorlove.com (+69) "/* Interests */"
(view diff)
j12t joined the channel
#
KevinMarks_
So, should I update to Yosemite? Any actual gains worth the "where did my dev tools go?"
#
aaronpk
"where did my dev tools go" is solved by having documentation on how you installed them in the first place and just doing that again
KevinMarks joined the channel
#
KevinMarks_
So, that will work for next time if I document it this time
#
aaronpk
worth it
chreekat and jciv joined the channel
#
aaronpk
wow great writeup
#
snarfed
thanks!
#
snarfed
(sorry, that anchor is behaving badly, not sure why)
#
aaronpk
omg mine keeps adding the "-2" to the name too! thanks!
#
KevinMarks
I assume the discoveryd bug is gone with .4 ?
#
snarfed
i stopped seeing the bug, but no clue why
#
aaronpk
i'm on 10.10.2 and still see it
#
KevinMarks
10.10.4 remobves discoveryd for mdnsresponder
#
KevinMarks
which was what Iw as waiting for before upgrading
#
snarfed
pre .4, try that —no-namechange plist change, might work
#
aaronpk
oh am i that far out of date?
#
snarfed
rip bonjour
#
snarfed
was so ahead of its time back in the day
#
KevinMarks
no, it's back
#
snarfed
oh! i misread that slug
#
KevinMarks
and still open source
#
KevinMarks
though maybe not the iOS build
#
snarfed
ios open source lol
#
KevinMarks
well, the darwin/freebsd core is, kinda
#
snarfed
any fragmention users or developers here? (JonathanNeal KartikPrabhu)
#
snarfed
hey JonathanNeal!
#
JonathanNeal
HEY snarfed
#
snarfed
my site uses fragmention.js, and i'm debugging why this behaves oddly: https://snarfed.org/2013-03-02_mac_os_x#Yosemite
#
snarfed
i think it's because fragmention.js detects and handles existing ids, but not <a name>s…?
#
snarfed
cool, thanks. let me know if you want me to file an issue
#
JonathanNeal
What is the expected result? I was taken right to the Yosemite section on the page.
#
snarfed
JonathanNeal: oh! i wasn't. i'm first taken to the Yosemite link in the TOC. if i click on the url bar and press enter again, it then takes me to the section
eschnou joined the channel
#
snarfed
firefox 39 on mac os x
#
snarfed
i expected it'd immediately take me to the section because of the <a name>
#
GWG
snarfed: Do you use post formats in WordPress?
#
snarfed
GWG: no clue
#
snarfed
probably not
#
JonathanNeal
snarfed: would you be willing to test something for me?
#
snarfed
JonathanNeal: sure!
#
JonathanNeal
snarfed: I’d like to know if this behavior works the same if the element is not an <a>. Other elements to try would be a heading element and an input element.
#
GWG
snarfed: I am thinking of trying to get involved with their revamping.
#
GWG
Looking for Indieweb feedback
#
snarfed
JonathanNeal: got it. want me to change it to anything in particular? p?
#
snarfed
GWG: gl!
#
JonathanNeal
I you can setup a page with that. I will do all the browser testing ASAP (today) and push the change ASAP (today).
#
snarfed
oh definitely not urgent
#
snarfed
updating the page
#
JonathanNeal
Sure, <h3 name=“fragment”>, <p name=“fragment”>, and <input name=“fragment”>, somewhere that would make the page need to scroll down.
#
JonathanNeal
And I will see which browsers scroll.
#
JonathanNeal
Hopefully, all browsers behave the same, and this isn’t a special Firefox behavior that only has to do with <a name>
#
KevinMarks
chrome and safari went to the Yosemite header for me correctly
#
GWG
snarfed: gI?
#
snarfed
good luck
#
GWG
Oh, I thought that was an I. Oops.
#
JonathanNeal
Ping me when that test is ready and I’ll have at it, snarfed
#
GWG
snarfed: None of my attempts to get more Indieweb stuff into WordPress have met with interest.
#
snarfed
JonathanNeal: thanks! will do
#
snarfed
it's updated, but my site is now unhappy, so…meh
chreekat, elf-pavlik, KartikPrabhu, Pierre-O and eschnou joined the channel
#
csarven
aaronpk Zegnat What I intended to say was that when crappy OSes match a string with their internal emojis representations, it overrides the "text" with the graphical view of that text. For example, if you are on Mac OS X, you will probably see the character ✊ (raised fist) graphically as opposed to a glyph being rendered by a font available in the system.
#
Zegnat
ah, gotcha csarven
#
KevinMarks
slack lets you pick which emoji set to use
JonPincus joined the channel
#
csarven
emojis are not consistent across systems. Not anymore than a default set of font families. I highly prefer to rely on a particular font family to deliver that (which will be fairly consistent across systems, if the family is availabe), then some gimmick taking over.
#
JonPincus
KevinMarks and tantek, remember me demoing TapestryMaker's emoji reactions to you at HWC last summer? good to see the silos catching up :)
#
JonPincus
it really does make a huge difference -- especially on mobile
#
Zegnat
Unicode Consortium is trying to fix some emoji irregularities though, right? redefining dancer I believe and getting strict re orientation rules (so e.g. gun always aims to the same side)
#
Zegnat
So consistency should increase
JonPincus joined the channel
#
snarfed
JonathanNeal: https://snarfed.org/2013-03-02_mac_os_x now has <p name>, <h3 name>, and <input name>
#
JonathanNeal
Awesome, what are those names?
tantek joined the channel
#
snarfed
oh, all Yosemite
#
snarfed
sorry, did you want something else?
#
JonathanNeal
Array.prototype.map.call(document.body.querySelectorAll('[name]'), function (element) { return element.name });
#
KevinMarks
is tryign to use flexbox on semantic elements fruitless?
#
snarfed
JonathanNeal: iterating on this will probably be faster with a local html file :P
#
KartikPrabhu
I thought "name" attribute was considered obsolete
jciv joined the channel
#
snarfed
KartikPrabhu: sure, may be, but that's no reason to break backward compatibility
#
tantek
it is
#
KartikPrabhu
snarfed: fair argument
#
snarfed
i have lots of old html. i'd like to use new code. the argument here is…i'm out of luck unless i migrate all of my old html?!
#
tantek
except on forms
#
tantek
er, inputs.
#
snarfed
tantek don't you usually argue to make things easy for publishers?
#
tantek
snarfed, not disagreeing with you.
#
KartikPrabhu
snarfed: yes. I agree that fragmention.js might want to take into account nam
#
snarfed
oh sorry, i think i misread the "it is"
#
snarfed
right. sorry. ignore me.
#
tantek
I mistimed the it is ;) it was in-reply-to KartikPrabhu
#
snarfed
lol yes. drawback of online text comms :P
#
KartikPrabhu
tantek: that is how I read it :P
#
JonathanNeal
Name is deprecated, but should still be supported. Is that the consensus?
#
gRegorLove
kylewm: Looking for a link I think you shared recently about updating startssl certificates
#
tantek
JonathanNeal: it’s trickier than that.
#
tantek
“name” is a legit attribute on input elements, e.g. for grouping radio buttons
#
KartikPrabhu
JonathanNeal: for fragmention.js I think supporting name would be a good thing for back-compat reasaons as snarfed said
#
KartikPrabhu
tantek: oh! yeah. name on inputs might mess somethings up
#
tantek
but <a name> and <area name> is obsolete, yet worth supporting for backcompat as snarfed points out
#
tantek
however, <randomElement name> MUST not be supported
#
tantek
e.g. <p name> is nonsense
#
KartikPrabhu
fragmention.js at the moment does not look at tag-names
#
tantek
then it should
#
KartikPrabhu
JonathanNeal might have to code a DOM parser into fragmention.js :P
eschnou joined the channel
#
JonathanNeal
That’s a bit trickier, isn’t it. So <input name=“fragment”> should not be scrolled to, is that correct?
#
JonathanNeal
Is that the behavior in Firefox?
frzn_ joined the channel
#
tantek
JonathanNeal: correct
#
tantek
then only name attrs that create fragments to scroll to are on <a> and <area>
#
JonathanNeal
That’s still very doable, but a bit more tricky.
#
tantek
in practice you could likely get away with just <a name> support
#
KartikPrabhu
if one adds <a name> suport then <area name> might not be much of a stretch
#
JonathanNeal
And no matter what, we scroll to id=“fragment”, even if it is on an input or such?
cmhobbs joined the channel
#
KartikPrabhu
JonathanNeal: i think so
#
KartikPrabhu
anything that counts as a fragment should be scrolled to so any id, and <a name> <area name>
#
tantek
yes, id attribute is global
#
bear
puts on his Indieweb Ops hat...
#
bear
if you run servers you will want to check if you need a new version of OpenSSL - see https://www.openssl.org/news/vulnerabilities.html for details
#
tantek
oh is it that time of the quarter?
#
bear
if you run on a mac and use homebrew you can do this via: brew update && brew upgrade openssl
#
bear
yep :/
#
kylewm
gRegorLove: I'm not seeing anything particularly relevant in my browser history, though I did generate a new StartSSL cert recently...
#
kylewm
bear: this SSL vulnerability doesn't have as good of branding as the last few
#
bear
yea, it lacks marketing flare
#
snarfed
bear: iirc this only affects openssl versions since june or so right?
#
bear
yes, specific versions
#
snarfed
i think literally only affects versions that are ~1mo old
#
snarfed
so probably very very few of us (probably none) admin ssl stacks with one of those versions
#
bear
1.0.2c, 1.0.2b, 1.0.1o, 1.0.1n
#
snarfed
yup, dates on those are very recent
#
snarfed
anyway
#
bear
if you run older ubuntu or redhat you will have to check
#
kylewm
ubuntu 14.04 has 1.0.1f from Jan 2014
#
bear
I'm posting because with Ubuntu end-of-lifing their older TLS a lot of personal servers are now risky
#
bear
s/TLS/LTS/
#
Loqi
bear meant to say: I'm posting because with Ubuntu end-of-lifing their older LTS a lot of personal servers are now risky
#
snarfed
1.0.2b/c are from june, 1.0.1n/o are from at least march or april
#
bear
yea, everyone should run: openssl version
#
bear
and check their versions
mlncn and lukebrooker joined the channel
#
gRegorLove
kylewm: I think it might have been https://kuix.de/blog/index.php?entry=SSL/TLS-servers,-SHA-1/SHA-256-and-StartSSL.com-certificates I might just be starting from scratch anyway, since my personal certificate expired. Sent them an email to merge my accounts.
lukebroo_ joined the channel
#
gRegorLove
Yeah, they told me since all my certs are expired, just continue with the new account.
lukebroo_ joined the channel
parzzix joined the channel
#
gRegorLove
Ooh https://sslmate.com/ looks interesting, if you don't mind $16/year
#
kylewm
I want to say KartikPrabhu switched to sslmate (in a fit of pique at startssl)
#
gRegorLove
What is sslmate?
#
aaronpk
once I realized that startssl only uses the public key in the CSR, I just leave a CSR on my server and use that every time I need a new cert
#
loqi.me
created /sslmate (+58) "prompted by gRegorLove https://indiewebcamp.com/irc/2015-07-09/line/1436474426731 and dfn added by gRegorLove"
(view diff)
KartikPrabhu joined the channel
#
KartikPrabhu
kylewm: gRegorLove: hmm yes I did
benwerd joined the channel
#
kartikprabhu.com
edited /sslmate (+110) "add me"
(view diff)
jciv, friedcell and KartikPrabhu joined the channel
#
JonathanNeal
How do I label these elements? anchorable?
#
KartikPrabhu
that is not a bad name
#
snarfed
JonathanNeal: ooh thanks!
#
snarfed
yay, works great!
#
Loqi
woot
tantek joined the channel
#
KartikPrabhu
is so confused by github interface.
#
KartikPrabhu
how do I update my fork of fragmentions!
#
aaronpk
afaik you have to do it from command line or git app
#
bret
KartikPrabhu: what do you mean by update?
#
bret
two options, merge new changes in, or rebase
#
KartikPrabhu
bret: I have a fork. JonathanNeal just made changes to the original. I want to update my fork with those changes
#
aaronpk
git pull origin master; git push kartik master
#
snarfed
aaronpk++
#
KartikPrabhu
:| why does github have a GUI then!?
#
Loqi
aaronpk has 873 karma
#
bret
¯\_(ツ)_/¯
#
KartikPrabhu
goes back to science, which is easier
#
snarfed
git is a funny combination of brutally simple internals and brutally complicated UI
#
snarfed
github helps the UI part a bit, but only a bit
#
JonathanNeal
KartikPrabhu: are you comfortable in the terminal?
#
bret
merges and stuff on gh are confusing
#
aaronpk
I suspect github will add that feature soon, it's kind of weird it's still missing
#
KartikPrabhu
JonathanNeal: I can do it through the command line but I was hoping to do it quickly with some clicks. Also I don't have the fork locally
#
KartikPrabhu
oh I do have it locally. ok command line here we go
#
JonathanNeal
I can show you what to do.
#
KartikPrabhu
JonathanNeal: no I got it on the command line. thanks :)
#
KartikPrabhu
alright done :)
#
KevinMarks
you can do it with the UI but you end up making a pull request from master to you
#
KartikPrabhu
err yeah that's why i got confused. I tried it on my fork and it was "sending a PR to the original"
#
KartikPrabhu
anyway add upstream-pull-automerge-push on command line did it. a few lines really
#
KartikPrabhu
GUI is hard
mlncn, snarfed, lukebrooker, j12t_ and tantek joined the channel
#
gRegorLove
Arg. I'm still getting "outdated security settings" warning in Chrome after updating my StartSSL to SHA256
#
gRegorLove
Oh, I think the Startcom root cert in the trust store is still SHA1 for some reason.
#
tantek
I think I’m going to manually copy (incorporate into the post) a few indieweb replies to my recent post but not the silo comments (which I’ll wait til I have code that process the webmention queue).
#
tantek
Am I the only one that does that? Manually copies comments into a post?
#
tantek
what is a post footer?
#
KartikPrabhu
I have done that with blogs that don't support webmention
#
snarfed
tantek: traditional publishers do that sometimes, right?
#
rhiaro
I manually add replies to my database when I particularly want them to show up, as I'm not receiving webmentions or scraping them from webmention.io yet
#
rhiaro
via SPARQL query :)
#
loqi.me
created /post_footer (+34) "prompted by tantek https://indiewebcamp.com/irc/2015-07-09/line/1436482207012 and dfn added by tantek"
(view diff)
#
tantek
funny thing is, even just manually copying the comments into a post forces you to solve the /comments_presentation problem
#
rhiaro
automatically showing mentions on posts is my goal for this weekend I think
#
tantek
rhiaro: awesome! Adactio can likely help you with that if you run into any trouble.
mlncn joined the channel
#
tantek
just realizing my articles don’t explicitly link to syndicated copies
#
tantek
I mean, not automatically at least, despite the data being in storage
#
gRegorLove
Can any TLS gurus help me out with the "outdated security settings" on https://gregorlove.com/ The intermediate cert I set up includes the root SHA2 cert and class1 SHA2 cert but https://www.ssllabs.com/ssltest/analyze.html?d=gregorlove.com still shows SHA1 for cert path #2.
KartikPrabhu joined the channel
#
kylewm
gRegorLove: I'm not an expert, but I think you're good ... mine reports RSA1 on Path #2 as well, but it says it's coming from "the trust store" which i don't think you have any control over, and sha^13.com is happy with it
#
tantek.com
edited /comments-presentation (+187) "/* Tantek */ experiments with incorporating comments into a post"
(view diff)
#
gRegorLove
I think it's a locally cached version of the intermediate cert. Reading up more here https://code.google.com/p/chromium/issues/detail?id=473105
#
tantek
KartikPrabhu: could you add your manual comments copying examples to a section for yourself in here: https://indiewebcamp.com/comments-presentation#IndieWeb_Examples ?
#
gRegorLove
kylewm: Do you get any TLS warning visiting my site?
#
kylewm
gRegorLove: nope, all good
#
gRegorLove
Good. I wonder how many Windows users have the Startcom intermediate cert cached, though.
#
kartikprabhu.com
edited /comments-presentation (+216) "/* Manually Display */ add me"
(view diff)
#
kylewm
worked in Chrome and IE too (which don't share a cert store with Firefox afaik)
#
gRegorLove
Cool, thanks.
#
gRegorLove
If I understand correctly, Startcom's sha1 and sha2 intermediate both used the same fingerprint, and if you visited a site that had an incomplete cert chain, the sha1 was probably cached in Windows. Seems to be the case with me.
#
gRegorLove
And that sha1 intermediate is valid through 2017
#
gRegorLove
Frustrating
elf-pavlik and gRegorLove joined the channel
#
@t
@kevinmarks @grorgwork +1 <style scoped> — strong #indieweb use-case: custom post style: http://tantek.com/2015/185/t1/coded-per-post-styling-style-scoped (ttk.me t4c51)
(twitter.com/_/status/619293310549626881)
#
tantek.com
edited /app (+149) "/* The problem with apps */ deep linking in apps sucks too, with citation of course"
(view diff)
#
tantek.com
edited /feedly (+138) "issues, Fails To Handle Atom XHTML titles"
(view diff)
#
KevinMarks
is learning subtle things about CSS selectors