2013-10-20 UTC
othiym23 and paulcp joined the channel
bnvk joined the channel
josephboyle, paulcp, bnvk and tantek joined the channel
# 01:53 bret hope the talk went well. Eagerly waiting the recording
# 01:53 bret anyone coming to webrtc camp tomorrow?
paulcp, tpinto and taterbase joined the channel
# 04:56 aaronpk we're doing mapattack tomorrow, then getting on an airplane :)
# 04:57 tantek hopefully we'll see more folks in the channel from realtimeconf?
jernst_ joined the channel
benwerd joined the channel
# 05:02 Loqi benwerd: tantek left you a message 2 days, 6 hours ago: when did you implement sending (and receiving) webmentions? also, what do you think of adding rel=webmention in addition or instead of the .org rel in idno?
# 05:02 Loqi benwerd: tantek left you a message 2 days, 4 hours ago: do you discover&send webmentions synchronously or asynchronously in your posting UI? either way, do you report webmention results back to your posting UI?
ryana joined the channel
# 05:04 benwerd I sent webmentions May 31st, received them sometime in June
# 05:05 benwerd I've just added rel=webmention in addition to .org, adding it now (think it's a good idea)
# 05:06 benwerd and no webmentions are reported back in the UI - yet (I want to change the form's visual context based on anything it discovers in the target)
# 05:08 benwerd tantek (and other SF indiewebbers): got time for a catchup sometime this week maybe?
# 05:09 tantek benwerd - above werd.io permalink is now returning empty content
# 05:09 benwerd tantek - yep, refreshing the whole site from scratch, is down temporarily
# 05:10 benwerd I'm actually doing this one manually, on a relatively slow connection so a minute or two. normally, seconds
# 05:12 benwerd re: catching up, are any of Tuesday, Thursday, Friday good for you?
paulcp joined the channel
squeakytoy2 joined the channel
# 05:23 tantek !tell benwerd - awesome, you're using <link href="http://werd.io/webmention/" rel="http://webmention.org/ webmention"/> which defeats all the regexes
# 05:23 Loqi Ok, I'll tell them that when I see them next
# 05:24 tantek so they have to parse for rel attributes properly
# 05:30 tantek aaronpk - the footer reference link is quite powerful
# 05:32 tantek OMS - love this slide - so true: "You Want to Blog, But..."
taterbase and caseorganic joined the channel
# 05:34 tantek hey caseorganic - just reading your slides! :)
# 05:34 caseorganic tantek: preso got very good response. some people registered their own domain name already
# 05:35 tantek glad you got the Twitter - Jaiku event in there - that was an important one that's been swept under the rug
# 05:35 caseorganic tantek: ha, well, i had blaine cook review my slides and we talked for at least 2 hours while sitting next to each other at dinner yesterday
# 05:36 tantek I remember that happening (I was at Social Web Foo) and I remember blaine talking about it later too
# 05:36 caseorganic tantek: ralph meijer was in the front row for the talk! i didn't recognize him until he pointed to himself
# 05:36 caseorganic tantek: blaine was absolutely beaming about it. he was so excited when it happened
# 05:36 caseorganic tantek: i felt like something was missing but i didn't know what it was
# 05:36 tantek when what happened? was he at realtimeconf in your talk?
# 05:37 caseorganic tantek: ralph was at the talk and in the slide - he pointed to himself on the big screen from the front row. extremely amused
# 05:39 caseorganic tantek: to show just how much more communication can be done over irc vs. email
# 05:40 aaronpk and showing how much more productive IRC discussions are
# 05:40 tantek how did that go over? that's a very mailing list heavy crowd
# 05:41 aaronpk hah as far as I can tell eran hates mailing lists as much as you do
# 05:41 tantek oh boy - there's IndieMark in a slide - guess it's real now :)
# 05:44 tantek caseorganic - short for "Obligatory Neo reference" :)
# 05:46 caseorganic also, first time doing html slides! feels great and they look great
# 05:46 caseorganic looking forward to throwing away ppt and converting my old slides so i can centralize all content on my site
# 05:47 tantek and better than sharecropping on slideshare ;)
# 05:47 aaronpk slideshare is only as good as slideshare is around :)
# 05:49 aaronpk unfortunately the new auto-parse legacy mf1 class names makes for some weird results
# 05:49 aaronpk because of the class names the designers chose for things
# 05:49 tantek really? it shouldn't unless they use a vcard root class name
# 05:50 tantek I mean, unless a they use a uf1 root class name in general
# 05:50 tantek e.g. inside h-entry, only dt-published should work, not dtpublished
# 05:51 aaronpk specifically "author" and "description" were getting added
# 05:51 tantek so inside h-entry, only p-author should work, not "author"
# 05:52 tantek "author" should only work inside "hentry" for backcompat
# 05:53 tantek but that shouldn't matter if there's a "description" inside an h-entry
# 05:54 tantek "description" should only show up inside an "hentry"
# 05:54 aaronpk from what I can tell it is mapping legacy class names to new class names
# 05:55 tantek those mapping should only occur *inside* classic microformats root class names
caseorganic and taterbase joined the channel
taterbase, thatryana, ryana, andreypopp, LauraJ, catsup, caseorganic and melvster joined the channel
ryana, paulcp, bnvk, LauraJ, squeakytoy, barnabywalters, melvster, taterbase and barnabywalters_ joined the channel
bnvk and taterbase joined the channel
# 15:43 barnabywalters heads up for anyone who uses php-mf2: latest version (0.2.0) contains breaking changes, if you installed it with "mf2/mf": "*" now is the time to fix the version or update your app before things start breaking
# 15:44 barnabywalters lots of good stuff in v0.2.0, including support for the new e-* parsing rules (safe and consistent at last)
# 15:45 barnabywalters and a Mf2\parse function which saves you the trouble of making an object when all you want to do is parse mf data
# 15:45 barnabywalters unless any large changes to the parsing spec happen, the API should be fairly static now
# 15:47 barnabywalters aaronpk: I left BC on by default to help root out bad markup, but improved the way classnames are converted
# 15:48 aaronpk ok. same thing happened with "description", did you see my comment?
# 15:49 barnabywalters as instead of just changing everything in bulk, I restrict the scope of changes
# 15:49 barnabywalters so class=description will only be changed to p-description under classic root names which have a description property
# 15:51 aaronpk oh neat, returns "html" and "value" separately now!
# 15:51 barnabywalters I figured I should just bundle all the various breaking changes required into one update
# 15:51 barnabywalters so new e-* parsing rules come along with a free helping of improved API and docs
# 15:55 aaronpk I do feel like I get extra leading and trailing whitespace sometimes, due to indenting of the html tags
# 15:55 barnabywalters php-mf2 fails a bunch of the official mf2 tests because I don’t collapse whitespace
# 15:57 aaronpk shouldn't that just follow HTML rules of whitespace so that it matches what people see?
# 15:57 barnabywalters aaronpk: yeah, that’s my thinking behind <pre> whitespace, but it’s actually defined by CSS
# 15:58 barnabywalters e.g. a lot of your and tantek’s notes would look weird with collapsed whitespace
# 15:58 barnabywalters so I think it should be left up to the consumer whether or not to collapse whitespace
ryana joined the channel
# 16:09 barnabywalters aaronpk: okay with you if I make a bunch of breaking changes to php-mf2-shim?
bnvk joined the channel
# 16:24 aaronpk yeah sure, I'm going to have to update p3k for the new php-mf2 anyway
# 16:25 barnabywalters okay, cool. I’m also considering removing the CSS selector dependency, and making a little function which takes a classname and returns xpath syntax for matching that
# 16:27 aaronpk it's kind of funny the difference between PHP and things like Ruby and Node in that respect
# 16:27 aaronpk people have much fewer issues with dependencies in ruby and node because the package managers are much more mature
# 16:28 aaronpk and there really isn't a culture of "download this code and include it", it's always "add this to your Rubygems file"
# 16:28 barnabywalters I love using composer, but want as many people as possible to be able to use php-mf2
# 16:29 barnabywalters if requiring use of composer stops people like adactio from consuming mf data, it’s not something I want to do
# 16:30 aaronpk even i was really late to the commposer bandwagon, but i like it now
andreypopp joined the channel
# 16:41 spinnerin aaronpk: Given the gem dependency issues I've seen, you're making PHP sound really bad :)
# 16:41 barnabywalters spinnerin: I’ve barely used rubygems, but composer is one of the best dependency management systems I’ve ever used
# 16:43 barnabywalters PHP has a mixed history of code sharing and reuse, mostly ghastly, like PEAR and phpclasses.org :/
# 16:46 spinnerin Ah. It does help to have a common system for handling things.
# 16:47 aaronpk I wasn't saying anything about composer, just saying that a lot of people who use PHP don't even bother using a dependency management system
# 16:48 aaronpk spinnerin: out of curiosity, have you seen any alternative dependency manager other than rubygems? maybe something that works more like npm?
# 16:48 barnabywalters to be fair, if you’re not accustomed to using one it’s a steep learning curve and a lot of work
# 16:49 spinnerin aaronpk: Nope.
# 16:49 spinnerin But the kind of work I've been doing, we'd try to stick to the best common option, since the project is likely to get handed off to another team.
# 16:50 aaronpk I'd be curious to see what it would take to create something closer to npm for ruby
# 16:50 barnabywalters aaronpk: is it just the UI you’d change, or the package registering and dependency resolving infrastructure?
paulcp joined the channel
# 16:51 aaronpk actually I don't mind the rubygems.org site, but needs better version/dependency control stuff like npm has
taterbase, bnvk and tantek joined the channel
# 17:09 tantek barnabywalters - yes, this: "if you’re not accustomed to using one it’s a steep learning curve and a lot of work" (re PHP package managers)
LauraJ joined the channel
josephboyle joined the channel
# 17:11 aaronpk in general, yes. I find it interesting that Ruby and Node.js both grew up with package managers so basically everyone who writes ruby and node are already accustomed to them
# 17:13 barnabywalters man, twitter’s markup is obese. after cleaning, a tweet permalink page weighs in at 2288 lines
# 17:14 tantek and php runs on every cheap webhosting service (ruby and node still require sometimes painful setup)
# 17:16 tantek barnabywalters - you're right about adactio being a good test case for this sort of thing
# 17:16 tantek he's quite representative of the far larger number of web designer/developers who maybe do *a little* bit of backend coding here and there
# 17:17 barnabywalters tantek: yeah, people who don’t want to go through the learning curve of setting up composer but are still super smart and want to consume mf2 data
# 17:17 tantek heck I myself haven't learned about composer yet
# 17:17 tantek and I'd rather be coding on my indieweb site than learning more overhead
# 17:18 aaronpk there's definitely a learning curve to seting up composer, but I'm glad I took the plunge finally. it does make things easier going forward.
# 17:18 tantek the "just include one file" metric that adactio conveyed at IndieWebCampUK is a key bit of information
# 17:19 barnabywalters yes, the benefits outweigh the initial upfront cost quickly, if like using lots of dependencies
# 17:19 Loqi barnabywalters meant to say: yes, the benefits outweigh the initial upfront cost quickly, if you like using lots of dependencies
# 17:19 tantek but maybe most will not be using lots of dependencies
antic joined the channel
# 17:22 tantek well, both Taproot and p3k do much more than Falcon so I can see there's some advantage to that
# 17:22 aaronpk oh I lied, the Twitter client should be there too, but I had the twitter client in there before I was using composer
# 17:23 aaronpk that's all. the rest are my own libraries that are part of p3k
# 17:24 aaronpk i should look at guzzle. right now I just use curl
# 17:24 barnabywalters guzzle is nice. a little more OO than I’d like, but it does a lot of stuff very well
# 17:25 aaronpk actually nvm... it's just these libraries that use curl, and I wouldn't want to make them depend on guzzle :)
thatryana and antic joined the channel
# 17:27 tantek the only external code I'm using in Falcon right now is tmhOAuth
# 17:27 tantek which I've been updating by hand because I don't trust things to not braek
# 17:27 Loqi tantek meant to say: which I've been updating by hand because I don't trust things to not break
# 17:29 tantek so who here is going to IndieWebCampHollywood ?
# 17:30 tantek aaronpk - how long will you be in LA? I might extend my stay
# 17:31 aaronpk looks like I'm driving to Redlands on Monday morning
# 17:31 aaronpk so, just there on Sunday, but saying in LA sunday night
# 17:31 tantek aaronpk - do you have a hotel room with a spare bed for Sunday night?
BjornW joined the channel
# 17:32 aaronpk let me check the reservation, not sure what we ended up with
# 17:34 aaronpk sounds like we could probably fit you there on Sunday night!
# 17:36 tantek and then I may hang out in Claremont for a while
taterbase joined the channel
# 17:37 tantek aaronpk - how long are you going to be down there, e.g. out in Redlands?
# 17:37 aaronpk Redlands Monday and Tuesday at the Esri campus all day, flight leaves Ontario Tuesday night
# 17:41 tantek barnabywalters goes climbing? did not know that.
# 17:46 aaronpk wow, tripit parses out the esri travel emails really well
# 17:51 tantek now if only tripit would do email properly and give you a custom email address to send to like Flickr does for upload by email
# 17:51 aaronpk heh, yeah the generic "plans@tripit.com" thing seems rather easy to "hack"
# 17:55 tantek to forward fake emails to that look like they were being sent to tripit customers, and then pollute their profiles
jernst_ joined the channel
# 17:57 tantek btw - aaronpk - I know you are interested in all things phone-call-like - have you tried the new Google Hangouts iOS app? free phone calls to US phonenumbers.
# 17:57 tantek so now I can make free phonecalls from my iPod to anywhere in the US
# 17:58 tantek no need to signup for google voice, no need to give google your phone #
bnvk, ryana and caseorganic joined the channel
# 18:26 aaronpk oh yeah, would be easy to inject "trip spam" into people's tripit profiles like the google calendar invite hack
tilgovi joined the channel
# 18:49 tantek !tell shaners I've signed up for IndieWebCampHollywood :) Do you have enough data to decide on whether to make it a one or two day event? Page makes it seem like it is still up in the air and some folks may be travel planning. ;)
# 18:49 Loqi Ok, I'll tell them that when I see them next
caseorganic and taterbase joined the channel
# 19:06 tantek and that's another that happens with monoculture
# 19:08 tantek (in terms of actually scaling and succeeding - which doesn't bely the other problems: attack magnet, update maintenance etc.)
caseorganic joined the channel
paulcp joined the channel
Jeena and tilgovi_ joined the channel
# 19:15 bret talking about "why STP in webRTC" at webrtc camp :)
ttepasse joined the channel
# 19:16 tantek jeena - logs available too: indiewebcamp.com/irc/today
bnvk joined the channel
tantek joined the channel
# 19:21 Jeena So what are you guys working on nowadays?
caseorganic joined the channel
# 19:22 tantek I'm working in figuring out how to automatically send webmentions from my site. (Something that many others here are already doing)
tilgovi joined the channel
# 19:23 Jeena ah that was the easier to implement successor to Pingback, wasn't it?
# 19:24 tantek Yes, webmention is a simplification of pingback
# 19:24 bret tantek: I was thinking about taking a PuSH approach. Publish a list of webmentions you want sent then ping the mention server which looks at the list to see if there are any new mentions that need sending
andreypopp joined the channel
# 19:25 aaronpk bret: yea we've talked about that kind of thing for webmention.io before, right?
# 19:25 bret similar to how one would publish a push atom feed
# 19:25 tantek Bret hmm not sure I understand how the server could tell if it was new or not
# 19:26 bret im not sure, how does push know if the atom feed has new stories?
# 19:30 bret learning more about node.js, my self imposed restrictions are starting to feel way to strict... thinking about moving on to a more capable back end
# 19:31 Jeena and my spec was also heavily inspired by Pingback ^^
# 19:35 aaronpk we're doing that mostly with an h-card on the home page
caseorganic joined the channel
# 19:36 Jeena yeah, 2006 was either before or while you were starting with micro formats I think
# 19:36 Jeena and my main concern back then was the lack of HTML parsers for PHP
# 19:37 Jeena there are many of them now and people use many other languages for web programming too so it is a not a issue anymore
# 19:38 aaronpk we're also starting a good collection of microformats parsers which is even easier to deal with now
# 19:40 aaronpk takes the whole HTML part out of the picture as far as your code is concerned
# 19:40 Jeena ah, nice, but it still uses a normal SGML/HTML parser internally?
tilgovi_ joined the channel
taterbase joined the channel
# 19:46 aaronpk it's funny how much people want this to exist and yet the only thing to stick has been gravatar
# 19:48 Jeena And then companies like Github only let you use Gravatars as your avatar, you can't even upload one yourself.
paulcp joined the channel
# 19:49 tantek Aaronpk I'd say indieweb commenting with author avatars working interoperable across several implementations is getting close to "sticking" ;)
# 19:51 Jeena Hm where should my hcard be, on my homepage? I have it on the contact page but I assume there it can't be found automatically
# 19:51 aaronpk but an extremely recent development compared to the timeline of people trying to solve this before
# 20:09 Jeena oh, ah, it is totally different to what I have been using with the vcard class, need to update that
caseorganic joined the channel
jernst_ joined the channel
taterbase, friedcell, abrereton, caseorganic, tantek and benwerd joined the channel
# 22:00 Loqi benwerd: tantek left you a message on 10/19 at 10:23pm: - awesome, you're using <link href="http://werd.io/webmention/" rel="http://webmention.org/ webmention"/> which defeats all the regexes
andreypopp joined the channel
# 22:02 benwerd I actually didn't think to check that. The PHP client code I'm using does also use a regexp instead of parsing header links
paulcp, caseorganic, lukebrooker and ttepasse_ joined the channel
# 22:15 tantek jeena - which img tag, this one? <img src="/wiki/skins/Microformats/images/logo.gif" alt="Microformats Wiki"/>
# 22:16 Jeena <img class="u-photo" src="http://example.org/photo.png" />
taterbase joined the channel
# 23:01 Jeena http://indiewebcamp.com/comment is kind of a neat idea, especially because it can be used for everything (kind of like on Facebook but the parent is just a URL or something) I have to admit that it takes quite some time to get your head around it
taterbase joined the channel
# 23:10 tantek jeena - indeed, the ability to comment on any URL, not just inside a silo, is quite powerful. And like a web interconnected by hyperlinks, a web interconnected by comments takes quite some time to get your head around .
caseorganic joined the channel
paulcp, jernst, andreypopp and taterbase joined the channel