#dev 2018-08-27

2018-08-27 UTC
renem, snarfed, koddsson, strugee, KartikPrabhu, [asteres], wagle, cweiske, AngeloGladding, plindner[m], barpthewire, grantcodes[m], eli_oat[m], jjuran, zoglesby, myfreeweb[m], loicm, gRegorLove_, gRegorLove__, gRegorLove, TripFandango, jeremych_, [kevinmarks], barpthewire1, trip_, deathrow1|absnt, jgmac1106 and miklb joined the channel; trip_ left the channel
#
GWG
sighs
#
GWG
Too much to do
KevinMarks, eli_oat, [kevinmarks], jgmac1106, sidp and trip_ joined the channel; kline left the channel
#
aaronparecki.com
created /WebManifest (+29) "Redirected page to [[Web App Manifest]]"
(view diff)
#
aaronparecki.com
edited /Web_App_Manifest (+1) "/* How to use for IndieAuth */ fix rel=redirect_url"
(view diff)
kants, snarfed, j12t and jgmac1106 joined the channel
#
aaronparecki.com
edited /Web_App_Manifest (+495) "/* IndieWeb Usage */"
(view diff)
#
aaronpk
Zegnat: there you go, 3 new examples ^^
#
Zegnat
Oh my!
#
Zegnat
I’ll update the survey in a bit :D
AngeloGladding joined the channel
#
petermolnar
jet another json sidefile, wheeee
#
Zegnat
At least this one has a reputable standard, petermolnar. And several people are already using it so their homepages can be added to phone home screens (e.g. adactio)
#
petermolnar
I know. It's only the yes, add one more sidecar file problem
#
petermolnar
the same ~~could~~ should be achieved by pulling the site meta headers
#
Zegnat
rel-icon and meta application name will probably be fallbacks people implement, so you don’t have to have the WAM file. Also h-app for those who use an mf2 parser. Options abound.
#
Zegnat
Borrowing the WAM standard does make a lot of sense for the spec though
#
aaronpk
especially since i'm super worried that without a somewhat sane JSON format, i'll get a bunch of pushback to use the openid connect discovery document which is way worse than this
#
Zegnat
OpenID Connect Discovery uses .well-known?
#
aaronpk
I believe so
#
aaronpk
"OpenID Providers supporting Discovery MUST make a JSON document available at the path formed by concatenating the string /.well-known/openid-configuration to the Issuer"
#
aaronpk
some implementations have an issuer URL that contains a path, so this led to some implementations making the document available at https://example.com/issuer1/.well-known/openid-configuration
#
Zegnat
Yeah, I was just reading that
#
Zegnat
I love how they do /.well-known/ after non-hostnames, just appending it to paths. That’s not how this usually works, right?
#
aaronpk
yeah no not at all
#
Zegnat
Then I remember correctly and don’t have to go check the RFC again :P
#
Zegnat
This doesn’t feel like something that should be propogated in other standards... do you know if other OAuth extensions may already be using it?
#
aaronpk
I don't know of any, but most of what i'm tracking is oauth and openid connect, not some of the other fringe stuff
#
Zegnat
Gotcha
snarfed and wagle joined the channel
#
aaronparecki.com
edited /Web_App_Manifest (+577) "add ownyourgram, indielogin.com, teacup"
(view diff)
#
ludovicchabant
hey here, does anybody have clever ways to handle incoming WebMentions with static websites? I'm currently implementing that quickly stores the WM data in a json file, and a cron job that takes this data, redistributes it to each appropriate page, and re-generates the website if there was anything new.
#
ludovicchabant
s/implementing that/implementing something that/
jgmac1106_, tantek__ and snarfed joined the channel
#
[kevinmarks]
there's webmention.herokuapp.com and webmentio.io
snarfed joined the channel
#
[kevinmarks]
er webmention.io
#
[kevinmarks]
they both accept WM and have a js api
[eddie] and trip_ joined the channel
#
ludovicchabant
No but I mean: what to do after you’ve gotten the mention
#
ludovicchabant
Like, WordPress has a plug-in that will show stuff under the article, but static websites are... static
#
ludovicchabant
So I was wondering if there was any cool stuff people did with their static website
#
ludovicchabant
Otherwise I’ll just keep going with my current implementation idea — it should be fine
#
aaronpk
your approach sounds like a DIY version of how people use webmention.io with static sites
#
aaronpk
some people use webmention.io to collect all the mentions, then rebuild their site periodically by pulling the data out of the API
#
gRegorLove
What is Morris?
#
Loqi
Morris is a self-hosted PHP service for storing Webmentions from webmention.io in a JSON format that static sites can use to render them without querying webmention.io on each build https://indieweb.org/Morris
#
gRegorLove
another option
KevinMarks_ joined the channel
#
@megarush1024
↩️ It's not a client in that sense. It's a service you use to connect a website to syndicate content from a website supporting microformats 2, and then to receive incoming webmentions from platforms like Twitter.
(twitter.com/_/status/1034137602708975616)
[tantek] joined the channel
#
@megarush1024
↩️ Bridgy syndicates those as tweets based on content type, and then, when people reply or like or retweet on Twitter, those interactions come back as webmentions which are displayed as comments under the original.
(twitter.com/_/status/1034138021464031232)
jackjamieson and [kevinmarks] joined the channel
#
@megarush1024
Interesting. Twitter keeps automatically revoking authorization when I attempt to authorize Bridgy to receive and/or publish. Is anyone else seeing this? #indieweb
(twitter.com/_/status/1034115278685327360)
KevinMarks joined the channel
#
[tantek]
Hmm snarfed are you seeing any weirdness like that ^^^
[jon], wagle, trip_ and barpthewire joined the channel
#
GWG
Hi all
#
GWG
I am trying to switch from regex
#
GWG
I took my original regex from WordPress and it seems ill suited for parsing HTML
#
[tantek]
Right, I believe the current approach is to scrape inside HTML *comments* with regex
#
GWG
I am trying to find a good explanation of better ways to parse HTML
#
GWG
For my own use
#
Zegnat
GWG: use an html parser?
#
GWG
Yes, I need a reference guide
#
Loqi
[Masterminds] html5-php: An HTML5 parser and serializer for PHP.
#
GWG
I already include it for php-mf2 in one project
#
GWG
Semantic Linkbacks doesn't though
#
Zegnat
It handles modern html slightly better than php's own DOMDocument.
#
GWG
I should just read the php mf2 code
#
Zegnat
Happy to help if you are wondering about something. Don't really know any primers...
#
Zegnat
mf2 mostly uses XPath to find elements that match specific requirements
#
GWG
Which is what I want to switch to
#
GWG
I did it in Indieauth to get icons and properties
#
GWG
So, I want to switch from regex in my other project
#
Zegnat
Sounds like a plan!
#
GWG
I think I can cut a lot of code
#
GWG
Wonder why the original code in WordPress didn't
#
GWG
I checked, and they use xpath elsewhere
#
GWG
Is there a performance issue?
#
Zegnat
Shouldn't be... Compatibility reasons?
#
GWG
That's why I checked
#
GWG
They include SimplePie, which uses it
#
GWG
I searched their repo
#
Zegnat
Then I'm not sure. Running regex on a big enough file can run into performance problems too.
#
GWG
I could track down the developer, but the code was removed from WordPress anyway
#
GWG
So, I assume that I have nothing to worry about except better results
#
GWG
I am enhancing my ogp to mf2 parser
#
GWG
If there is no mf2 then it tries to find information elsewhere to fill in the blanks
#
GWG
I also have this consistency issue pfefferle pointed out where sometimes I store as mf2, sometimes jf2 and often forget the type
iasai joined the channel
#
ludovicchabant
aaronpk: ok thanks!
#
ludovicchabant
gRegorLove: thanks too :)
[keithjgrant] joined the channel
#
tantek__
what is an SSD?
#
Loqi
It looks like we don't have a page for "SSD" yet. Would you like to create it? (Or just say "SSD is ____", a sentence describing the term)
#
tantek__
hmm well in case someone is running their own indieweb server with SSD drives: https://blog.okmeter.io/real-world-ssd-wearout-a3396a35c663
chrisaldrich and [chrisaldrich] joined the channel
#
tantek__
wow this is another great expression of avoiding architecture astronautics: https://hacksoflife.blogspot.com/2018/08/solve-less-general-problems.html
#
tantek__
architecture astronomy << 2018-08-11 The Hacks of Life: [https://hacksoflife.blogspot.com/2018/08/solve-less-general-problems.html Solve Less General Problems]
[eddie] joined the channel
#
tantek__
anyone here running their IndieWeb site on an Arduino? I know someone at IWC SF recently demonstrated their site running on one! They had to configure and reboot to change display output resolution, and of course rebooting took seconds. It was amazing. But in case you're an Arduino user: https://praeclarum.org/2018/08/27/oops-i-wrote-a-c-compiler.html
#
aaronpk
an arduino?? not a raspberry pi?
#
tantek__
ok now I'm not sure
#
tantek__
maybe it was a pi. will have to watch the demo video
#
tantek__
which reminds me
#
gRegorLove
that was a pi iirc
AngeloGladding, Guest92, [kevinmarks], TripFandango and [cleverdevil] joined the channel
#
[cleverdevil]
FYI, I implemented the beginnings of "?q=source" for Known's Micropub implementation.
#
[cleverdevil]
No paging or querying, yet.
#
[cleverdevil]
I may try and get to that this weekend, if I have some time (its a busy week).
#
@DataG
Webmentions: Enabling Better Communication on the Internet | A List Apart http://bit.ly/2LsmrK0
(twitter.com/_/status/1034203162616328193)
[kevinmarks] joined the channel
#
[kevinmarks]
Heh at the problem space there. DV is really easy to edit as every frame is the exact same number of bytes. You can just use cat from the command line.
#
[kevinmarks]
Though ironically we had the opposite problem ta apples
#
[kevinmarks]
*at Apple
#
[kevinmarks]
imovie was written without using QuickTime and only worked on DV over firewire. We had the abstraction, and it was a good one, which incorporated DV, but they didn't use it because they were too stealth to ask our advice.
#
[kevinmarks]
Then when they were told to support other video formats from digital cameras they had to rewrite the whole thing.
#
aaronpk
finding myself want to search within channels in my reader for specific keywords
DenSchub, petermolnar_ and ichoquo0Aigh9ie joined the channel
#
@miklb
Are you an #indieweb #WordPress user? We could use your help testing the latest version of micropub and Indieauth plugins. Drop by https://chat.indieweb.org/wordpress for details or ping me here. Thanks!
(twitter.com/_/status/1034222308175400961)
#
vilhalmer
aaronpk: the pgp verify button on indielogin refuses to enable itself, want me to file a bug?
#
aaronpk
it will enable itself when the text changes to include "BEGIN PGP SIGNED MESSAGE"
[jgmac1106] joined the channel
#
aaronpk
also make sure you allow javascript
#
vilhalmer
hmm, both of those things should have been true
#
vilhalmer
I wonder what in my browser is breaking it
#
aaronpk
it's still working fine for me in chrome
#
aaronpk
what browser are you using?
#
vilhalmer
also chrome
#
vilhalmer
I'll play extension whackamole :)
AngeloGladding, KevinMarks and TripFandango joined the channel
#
vilhalmer
oh, it has to be "BEGIN PGP SIGNATURE"
#
vilhalmer
maybe I'm using the wrong flags on gpg
#
vilhalmer
ah yeah was using `-s` instead of `-b`
#
vilhalmer
either validates, though
#
aaronpk
my gpg results in "BEGIN PGP SIGNED MESSAGE" and that was recognized by indielogin
#
vilhalmer
huh, I even hard-reloaded to make sure I didn't have stale code
#
aaronpk
indielogin is looking for "BEGIN PGP SIGNATURE" but what it really needs is a PGP SIGNED MESSAGE that also includes the signature
#
aaronpk
you should be giving it a string that looks like this https://media.aaronpk.com/Screen-Shot-2018-08-27-16-57-59-DV0K3ocdo2.jpg
#
Loqi
totally