2014-01-18 UTC
# 00:01 tantek that lack of type forces programmers to canonocalize on ISO8601 long form in Z which is both human unfriendly and loses information (timezone etc.)
# 00:02 bear just like mf2 - you can use json in a way that doesn't lose data
# 00:03 tantek or rather - can you point to any examples of anyone using JSON in a way that doesn't lose data in a human-unfriendly fashion?
# 00:03 bear all of the examples are from internal data conversation tools that go from SQL -> JSON or XML -> JSON
# 00:04 bear because of the lack of data types, everything is stored as JSON strings
# 00:07 tantek bear - then you likely are losing the SQL type information in your JSON
# 00:08 bear it wasn't but that's only because each field value had a corresponding meta data value of where it came from and how it was stored there
caseorganic joined the channel
# 00:08 bear I found with data conversion code, to always store the schema along with the data when transporting
# 00:08 bear but again, probably out of band to IndieWeb discussions
# 00:09 tantek whereas with uf2 - we've minimized that to infrequent occurences
# 00:09 tantek with a handful of dt- u- e- prefixed properties
# 00:10 bear sorry - I must be in a really off mood as I'm seeing everything thru a curmudgeon filter
# 00:11 bear I actually mutter that about what most of you all say
# 00:11 bear (being probably the older person in the room)
# 00:12 bear is 50 and has been in computers since 1979
# 00:13 tantek is not 50 but has also been coding on computers since 1979 (even if it was BASIC on a timeshare).
# 00:13 tantek notes that length of time coding on computers does not really correlate to much, except for having spent that time.
# 00:14 bear but age doesn't matter - i've learned to try and listen to everyone and then ask questions
# 00:14 bear my older habits get in the way some times
# 00:17 bear the authorship update earlier has me chewing over some ideas and now i'm going to go add them to the python code I am using
# 00:18 bear all of my earlier computers where other people's that bought them but couldn't figure them out - they hired me to fix, setup and program them
# 00:18 KevinMarks I really need to write up my post about mapreduce mindset too
# 00:19 tantek aaronpk - can we get Loqi to play music in the channel?
# 00:19 bear it's amazing how pervasive of a mindset earlier tool use is
pfenwick, li3n3 and snarfed joined the channel
# 01:58 iangreenleaf Hi all! Do you know of / maintain a list of data liberation options anywhere?
# 01:59 iangreenleaf I.e. "How to get your __[type of data]__ out of __[cloud service]__"?
# 01:59 tantek we tend to call them "silos" rather than "cloud services" (If I get your meaning)
# 01:59 tantek and such information (how to export your data from) is available on the wiki page for each
# 02:00 iangreenleaf yeah, I noticed the Twitter and Tumblr pages - is there much beyond that?
# 02:02 iangreenleaf nifty, thanks. so if I wanted to contribute, the way to do it would be to edit/create separate pages for each silo?
# 02:03 iangreenleaf I started thinking (as I wrested my 360 movie ratings out of Netflix) that we should have a good list of all this stuff
# 02:03 iangreenleaf especially since it's hard to figure out what is currently working, what has broken since first being released, etc
# 02:05 tantek yes it would be good to create a Netflix page - please go for it
# 02:07 iangreenleaf awesome, I'll work on contributing to a couple tonight - off the top of my head I can offer some info for Netflix, Facebook, Fitocracy, Garmin...
# 02:09 tantek you might ping aaronpk about Fitocracy and Garmin - I know aaronpk has done a lot of work with personal metrics services - and I'm sure he'd be interested in whatever you document for those.
# 02:10 iangreenleaf BTW, what's the word on "unofficial" export solutions? legit to add to the wiki or not?
# 02:10 iangreenleaf for example, I worked up a bookmarklet to export photos from others I was tagged in on Facebook a while back
KevinMarks2 joined the channel
# 02:23 tantek iangreenleaf - 3rd party tools for exporting data would be great to add to the wiki
# 02:25 Jeena no webmention on it yet and no automatic POSSE
# 02:26 tantek start with manual POSSEing! I (and most others) did.
# 02:26 tantek it also helps you figure out how you want your code to automatically POSSE for you
# 02:26 tantek and heck, I'm not even sending automatic webmentions yet myself - I manually curl them
# 02:27 tantek I'm preferring to document my thinking / figuring out of such features publicly in the hopes that such thinking helps others.
# 02:27 Jeena I have all the code for doing it automatically implemented because I use it for posts already, so it should be quite easy to add tomorrow
# 02:34 snarfed iangreenleaf: i'll take a note to go update export pages on the wiki too
caseorganic joined the channel
# 03:01 bret In terms of UX vs Security, threema.ch sets the bar for encrypted messaging. Been using it the past week and its been great
# 03:01 bret its unfortunately phone app only though :(
# 03:03 iangreenleaf snarfed: that looks super cool! are you working through the official APIs?
# 03:04 iangreenleaf !t snarfed that looks super cool! are you working through the official APIs?
# 03:06 iangreenleaf !tell snarfed freedom.io looks super cool! I'm curious if you're going through the official APIs, and if so, how you're handling the restrictive terms
# 03:06 Loqi Ok, I'll tell them that when I see them next
scor and j12t joined the channel
snarfed joined the channel
# 04:12 Loqi snarfed: iangreenleaf left you a message 1 hour, 5 minutes ago: freedom.io looks super cool! I'm curious if you're going through the official APIs, and if so, how you're handling the restrictive terms
caseorganic and hallettj joined the channel
snarfed joined the channel
# 05:25 snarfed in practice, Twitter is the most aggressive by a wide margin
# 05:27 iangreenleaf do you think, if it got wildly popular, that you could stay on their good sides?
# 05:27 snarfed my workaround for the last one was to have users generate their own api keys :/
# 05:30 snarfed feel free to Kiko into the github repo if you're interested!
# 05:30 Loqi snarfed meant to say: feel free to jump into the github repo if you're interested!
# 05:32 iangreenleaf yeah, I might poke around, POSSE is of great interest to me so always cool to see people building things
j12t joined the channel
KevinMarks2 joined the channel
caseorganic, schalkneethling, KevinMarks2, anselm, friedcell, CheckDavid and voxpelli joined the channel
squeakytoy, hallettj, pfenwick, XgF, igalic, brianloveswords, zaal, acegiak and hober joined the channel
# 10:32 Jeena good morning! Is it possible to POSSE my notes automatically to Facebook? Is there a API for that?
Zegnat, pasevin and dvirsky joined the channel
eschnou, Kopfstein and glennjones joined the channel
glennjones joined the channel
# 13:19 Jeena Wohoo now my mentions can POSSE to Twitter and send+receive webmentions!
eschnou, glennjones, friedcell, bnvk, pfenwick and snarfed joined the channel
Acidnerd and eschnou joined the channel
pasevin joined the channel
# 14:56 Jeena my blog has from the start been a rails app so I just upgraded first from rails 2 to 3 and now to 4 and added the indieweb stuff on top of everything. Back then I only had Pingback which later on stopped working
# 14:58 Jeena For now I added a atom feed for my notes too because there are no h-entry feed readers that I know of
LauraJ joined the channel
# 15:08 Jeena Hm now I also need to implement the special in_reply_to for Twitter so they can connect them in the right way
friedcell and scor joined the channel
# 15:13 Jeena wow, that was easy, and it is done, not I finally can answer people on twitter via my own notes :D
eschnou joined the channel
# 15:29 neuro` Jeena: BTW, what's your blog engine? I read it was Rails based
# 15:29 Jeena yeah it is just a plain rails app I wrote, nothing special
# 15:30 Jeena I tried to use blog engines but never got them working like I wanted them to work so I gave up and wrote my own blog software.
# 15:31 neuro` Oh OK. I've been maintaining a Rails publishing engine for 8 years now, that has notes and Twitter replies, that's why I was asking
# 15:31 Jeena just small stuff like a subtitle can be really complicated to add to a blog engine
# 15:32 Jeena and then this blog is a playground for me too, I try crazy stuff on a real audience sometimes just to check their reactions :-p
# 15:33 neuro` used to power blog.rubyonrails.org in 2004/2005
# 15:34 Jeena back then in 2004 you used rails already?
# 15:34 neuro` I started Ruby in 2002, and Rails when they released 0.4 iirc.
voxpelli joined the channel
# 15:35 Jeena in 2002 I bought my first PC and got a modem for it :D
# 15:35 Jeena before that I had an Amiga without internet between 1992 and 1995
# 15:36 Jeena yeah I was kind of late to the whole computer thingy
# 15:38 Jeena but it was my first programming project so it is spagetti code all the way
eschnou joined the channel
# 15:42 neuro` One day, a French guy was bored during the summer holliday, so he started to learn PHP and coded a spagetti blog engine. IIRC, it was first called "cafelog", then "b2".
# 15:43 neuro` B2 popularity grew quickly (was in 2001 or 2002), and after a while, that guy moved to something else.
# 15:43 neuro` Then, another guy forked the project, and named his fork "Wordpress"
# 15:44 neuro` So you see what can happen to PHP4 blog projects :)
acegiak, Jeena, glennjones, scor, snarfed, LauraJ and barnabywalters joined the channel
heath joined the channel
# 16:39 Loqi Ok, I'll tell them that when I see them next
bnvk joined the channel
# 16:40 Loqi bnvk: barnabywalters left you a message 1 minute ago: RE feedreader mockups hidden: indeed, my openphoto instance broke and I haven’t figured out how to fix it yet :( More up-to-date feedreader mockups at http://waterpigs.co.yk/intertubes
# 16:41 bnvk barnabywalters: afternoon- what are you up to?
# 16:45 Jeena cool! so how do you get new data from the websites, are you spidering those websites every couple of minutes or something?
# 16:46 barnabywalters nothing as fancy as that yet — I’m just requesting them live at the moment, which is why it’s quite slow
# 16:47 Jeena are you requesting the index pages or something else?
# 16:47 barnabywalters I’m just trying to fix the authentication problem so you can see the config UI
eschnou and tantek joined the channel
# 16:49 Jeena I'm also confused about where to point #tags in my notes
# 16:50 barnabywalters Jeena: point them to pages on your own site which aggregate content tagged with that tag
# 16:51 Jeena I see, ok I will do that in the future, right now there are just too few posts ^^
# 16:54 Jeena but that is just read only for me right now I assume?
# 16:54 Jeena and it is your own instance of this feedreader?
# 16:55 barnabywalters Jeena: yep, this is my instance of a flow-based-programming environment built for processing microformats data
# 16:55 Jeena it kind of looks like Yahoo pipes, I always liked their intuitive UI
# 16:56 Jeena ah and everything is HTML, I wonder how you do the connections
# 16:58 barnabywalters which handles all the connections. you just specify which HTML elements to connect, and it draws and updates the SVG
# 16:59 Jeena neat, even though I kind of would have expected it to be rendered in a WebGL canvas
# 17:01 barnabywalters of course! but in this case the important thing is the flow based programming UI, not the nitty gritty of how it’s implemented, so I used the tools which allow me to focus on that
# 17:03 tantek Jeena - congrats on getting your notes to POSSE to Twitter!
# 17:03 tantek you mentioned replying to people from your site - that's definitely a good thing to implement!
# 17:04 tantek it requires a little more work of course, to set the in-reply-to-status-id for proper threading
# 17:04 Jeena thanks, it wasn't that complicated, I found out that you can get a oauth_token and secret from the dev-page on twitter so you don't have to deal with the oauth process
# 17:05 Jeena I saved twitter_consumer_key, twitter_consumer_secret, oauth_token and oauth_token_secret in my code
# 17:05 tantek barnabywalters - I'm guessing secret URL on his site ;)
# 17:05 Jeena in_reply_to_status_id = in_reply_to.split("/").last if in_reply_to.starts_with? "https://twitter.com/"
# 17:06 Jeena {status: text, in_reply_to_status_id: in_reply_to_status_id }
# 17:06 tantek what happens if you have more than one in-reply-to URL?
# 17:07 Jeena I obviously make a ton of assumptions but that is ok, I'm the only one who uses it
# 17:07 tantek I think we are all doing that (making a ton of assumptions about usage) :)
# 17:07 Jeena tantek I only can have one twitter in-reply-to but I can have multiple others
# 17:07 Jeena I just go through the links in my html and do a webmention to every of them
# 17:09 Jeena Does the Twitter API even allow more then one in_reply_to_status_id?
# 17:09 tantek but if you have more than one in a post, you have to pick one somehow to POSSE
# 17:11 Jeena If I need more I have to put them into the text
# 17:12 Jeena and those only work with webmentions, not with twitter, but ok for now I think
KartikPrabhu joined the channel
# 17:19 Jeena oh but let me change the URL, because next time I make a screenshot this url will show something else ;)
# 17:20 tantek Jeena - or feel free to upload it (perhaps with an ISOdate in the name) to indiewebcamp.com
CheckDavid joined the channel
# 17:52 Jeena hmmmm, perhaps I should write some user-js so if I click on a reply button in twitter it opens my website and populates everything so I don't need to do it manually ...
# 17:55 Jeena I see you guys have already done everything I come up with :p
# 17:58 barnabywalters Jeena: thanks :) hope you find it useful — any feedback/pull requests very welcome, I haven’t really worked on it much for a while
snarfed joined the channel
# 18:12 Jeena hm how do I get to the options? firefox opens a new tab with chrome://webnactions_kango/content/options.html which obviously doesn't work ;)
tantek joined the channel
j12t, KartikPrabhu, iangreenleaf and iangreenleaf1 joined the channel
# 18:26 Jeena oh man, this is almost perfect, the only thing I would need is the twitter username as a variable
# 18:27 Jeena but I can get it from the URL instead as a forkaround for now
iangreenleaf joined the channel
# 18:29 barnabywalters Jeena: yeah I did make an earlier version with some silo-specific things like that but decided instead to make something which just works with URLs
bnvk, snarfed, friedcell and voxpelli joined the channel
# 18:58 voxpelli Displaying WebMentions – sort on h-entry publish date or on webmention receive time?
KevinMarks2, bnvk, barnabywalters and barnabywalters_ joined the channel
# 19:28 voxpelli snarfed: yeah, makes sense with Brid.gy in mind – dt-published it is then
# 19:29 voxpelli Just a bit worried that it will be hard to track new entries when they can appear in the middle of a comment thread
# 19:30 snarfed if you're ordering by time, then a reply's time is dt-published if it provides it
# 19:31 snarfed for direct replies, that will usually be the same as the received time, but for indirect cases like bridgy, it's not
# 19:31 snarfed bridgy often sends webmentions for replies (likes etc) that were created days or weeks or months earlier
pasevin joined the channel
# 19:32 snarfed you can do nested threading too if you want. an h-entry can have multiple u-in-reply-to links, e.g. one to the top level post and one to another reply
schalkneethling and snarfed joined the channel
# 19:52 snarfed Jeena: late to the game, but you can definitely posse to facebook automatically w/their api
# 19:52 snarfed there's plenty of code that does that automatically, e.g. wordpress plugins and more
# 19:53 Jeena Ok, nice, but I need to do it half manually, but perhaps there is a rails plugin too
pasevin, KartikPrabhu, bnvk_, barnabywalters and bnvk joined the channel
# 20:24 iangreenleaf aaronpk: Have you done any POSSE stuff with fitness tracking data? I see you've written about it.
# 20:32 aaronpk iangreenleaf: I haven't done POSSE with the data, but I am tracking a few things on my own site
# 20:33 aaronpk and technically the sleep and weight there is PESOS, but the pushups are original data
j12t and scor joined the channel
# 20:39 iangreenleaf cool cool. you just making up your own data formats? I don't know of any particularly standard ones for most metrics
# 20:41 aaronpk and based off of the format from the device I get the data
snarfed, schalkneethling and voxpelli joined the channel
melvster, pdurbin, ozten, icco and caseorganic joined the channel
# 22:09 bear KartikPrabhu - that is my planning ahead. I *think* that reply mentions will be handled/flagged for further processing (at least I did when I wrote that :)
icco joined the channel
# 22:22 bear KartikPrabhu - and i'm also all for removing it if it's not used - keeping code around "just because" is, for me, a old bad habit - so please do call me on it so I can unlearn it
tantek joined the channel
# 22:23 KartikPrabhu bear: I think when sending mentions it does not matter if it is a reply or not. But it would be important while parsing a received mention.
# 22:24 KartikPrabhu and if I am parsing a mention I received then I wouldn't be searching for all 'a's anyway. just the one that was mentioned
# 22:24 KartikPrabhu that's why I was thinking that the reply check is not necessary in this one
# 22:28 bear that's a good point about the code - please do submit a PR with that change (and even a note as to why) - I would love to have your ideas and usage notes in the code
# 22:28 bear if not a PR then an issue with the above - so I can make the change later
# 22:29 KartikPrabhu i was modifying quite some bits in my forked version. I'll update it on Github with notes of changes. I'll let you know
pfenwick, icco, bnvk and hallettj joined the channel
# 23:21 KartikPrabhu Q: For webmention endpoint discovery is it correct to only look in the document head and not the entire document?
# 23:23 bear jeena - you mean in the HTML server response header?
# 23:23 aaronpk you should probably just parse the whole HTML document for the rel values
# 23:23 aaronpk for example if you run a page through php-mf2 you get back an array of rels regardless of where in the HTML they occur
# 23:23 bear hmm, didn't think it would be returned in the header
# 23:23 Jeena I think it is more complicated to stop at the end of head then just let a parser parse the whole thing
# 23:24 tantek bear - there are examples of sites that do that
# 23:24 bear guess I have some code to change/improve then
# 23:24 tantek listed on the webmention page (that use HTTP LINK header for webmention discovery)
# 23:24 tantek there's a bunch of open source to parse them too - in various languages
# 23:24 KartikPrabhu Jeena: Beautiful Soup in python can parse any fragment. and it is apparently foster to do only the required parts
# 23:25 aaronpk yeah the time saved by parsing only the HTML <head> instead of th ewhole doc is going to be exponentially less than the time spent making the HTTP connection to actually send the webmention
# 23:26 bear ah, and I guess I didn't make it down to the bottom of the page where Alternatives are listed
icco joined the channel
# 23:27 Jeena btw. parsing the header is really easy, I always thought it is complicated because you can have multiple links in one Link: header etc.
scor joined the channel
# 23:32 bear yea, not too worried about implementing it or changing what parts to look for it - just being cranky-grumpy about something that I just need to stop being cranky-grumpy and edit the wiki about :)
CheckDavid joined the channel
# 23:41 Jeena so with me we're 9 people using webmentions? That is not many, but I still like it.
# 23:43 aaronpk i think there's more. that list is probably not complete anymore
pfenwick joined the channel
icco joined the channel
# 23:51 Jeena yeah I just saw, this german tweet was done with this plugin, need to have a look at it
# 23:52 bear KartikPrabhu - great changes! I would love to see this as PR and I would love to add you to the project contributor's list
# 23:53 Jeena I wonder about https://www.brid.gy/ I'm not quite sure I understand it correctly, what does it actually do? How does it know which of the links I post are posts to my website which have webmentions enabled?
# 23:53 KartikPrabhu bear: I'm new to git, so will have to figure out how to do PRs :P. But will do so for these changes...
# 23:54 bear KartikPrabhu - I appreciate you taking the time to improve the code. if you have made the change to your fork on github, the web ui for the commit should have a pull request button that is now visible
muesli joined the channel
# 23:59 KartikPrabhu bear: I want to use ronkyuu instead of writing the same code again, so I am adding things that I would use. :) also did the pull request thing. Did it work?