2013-11-21 UTC
# 00:00 tantek so that's interesting - why does Loqi respect the whitespace in a tweet when I paste the tweet URL in IRC, but NOT when Loqi gets the tweet via search API?
# 00:29 bnvk is there a page on IWC for travel data?
tantek and scor joined the channel
mathpunk and smus joined the channel
# 01:45 bnvk bret: wat, you don't like that aesthetic of web design? Dude, it's totally retro :P
# 01:47 bnvk Too many steps? You don't like sys admin-ing for hours?
# 01:47 bret Email servers are wicked complicated to use, right now in the present
# 01:47 bret i have better things to do, like complain about it on irc
# 01:48 bret please tell me mailpile will cut this down to the same ammount of work as running wordpress
# 01:49 snarfed hell yes, agreed. setting up a mail server is an especially big sysadmin task, and then running it is especially big too
# 01:49 snarfed deliverability, spam filtering, triaging when you get onto an RBL, etc
# 01:50 snarfed not to mention figuring out and implementing *all* of the different DKIM variants since every recipient SMTP server wants a different one
earplugs1, b0bg0d and earplugs joined the channel
# 02:07 tantek @jf and kevinmarks have made it to the Homebrew Website Club meeting already
# 02:08 tantek (and anyone else in the bay area that can make it to @MozSF)
paulcp, KevinMarks, CheckDavid and paulcp_ joined the channel
# 02:15 tantek bret - how about you co-organize it with aaronpk (he wants to do one in PDX also)
# 02:16 tantek you could aim for doing it the same night as the one in SF
# 02:16 tantek ping your MozPDX friends about using their space
# 02:16 tantek we already have 2 way video working from space to space
# 02:17 bret that would be the coolest computer club evaaarrr
# 02:17 bret I'm going to have to defer to aaronpk, I dont know that many people at mozpdx, despite being there twice
jonnybarnes joined the channel
# 02:17 bret KevinMarks moz has this insane video conference system
# 02:19 bret I could ping dietrich ayala, but I just met him breifly the other day. Wish I talked with him some more
scor joined the channel
# 02:22 bnvk bret: ah sorry, got distracted tinkering
# 02:23 bnvk Mailpile is not actually going to be a server- at least for the 1.0 version
# 02:23 bnvk we do have schemes to bundle a stable, wordpress ease of install mail server, but not right off the bat
# 02:24 bret We could start a webRTC room for remote HWC if people want
# 02:26 bnvk and tantek: I'm gonna push back on "deadend" due to being non tinkerable- there are loads of things people depend on daily that are no longer "tinkerable"
# 02:26 tantek bnvk - such things usually are ripe for getting replaced by things that *are* tinkerable, and eventually *are replaced*
# 02:27 tantek so if you identify such things, please speak up, so we can replace them :D
# 02:28 KevinMarks forgets old video conf testing rule: don't join from 2 cameras in the same palce
# 02:29 bnvk tantek: is hardware exempt some your theory?
# 02:30 tantek the easiest non-tinkerable hardware stuff to disrupt is getting disrupted first
# 02:31 tantek it's only a matter of time before all non-tinkerable hardware is disrupted by tinkerable hardware, based on where makerbot etc. are going
paulcp joined the channel
# 02:31 tantek but yes, that will take longer than software/servers
# 02:34 bnvk so is the root of it something along the lines of- humans have innate desire to tinker with their tools and things in their environment, thus things with high tinkerability will prove to be better in an evolutionary sort of way?
# 02:36 tantek tinkerable things are able to out evolve non-tinkerable things
# 02:37 bnvk ok, I'll go along with that, has a nice energy to it
# 02:38 bnvk and things that are "non tinkerable" or below the threshold basically get abandoned?
snarfed joined the channel
jonnybarnes joined the channel
# 02:44 bret bnvk sorry to peace out like that, other responsibilities are calling
# 02:49 bnvk KevinMarks: hrm, I'm the only one there
# 02:52 bnvk tantek: is there no dedicated page on the wiki for Travel data?
# 02:52 bret webrtc has to broadcast to all clients connected
# 02:53 bnvk yah- i've been modeling mine after his
# 02:53 bnvk I've got a few different things tho, I wanted to add
# 02:53 bnvk and that page looks more like a stub from the Hollywood hacki
# 02:58 bret KevinMarks it wasnt for me, i dont think talky works on tablets that well
# 02:58 bnvk I dunno, I was there, but you were just a black box Kevin
# 02:58 bret i can't tune in right now, but maybe bnvk and others do. Maybe google hangouts work work better
scor joined the channel
smus joined the channel
# 03:12 bear bret, talky.io works on tablets that have the latest chrome or firefox. It in general only does well for 1-3 users
# 03:13 bear hmm, I don't think we have tried ever to use talky.io as a broadcast tool
# 03:14 bear it can, you just have to have a rebroadcaster on the server side - in browser things like talky.io would melt under that load
smus joined the channel
# 03:26 Loqi [@kevinmarks] #indieweb @snarfed_org: when indieweb we focus on sharing to people first
# 03:26 bret dietrich you are here :) I was at unhosted the other day
# 03:30 bret I would want to coordinate with aaronpk, but would you care to sponsor such an event at mozPDX?
# 03:31 bret bear good to know about talky, also howdy (was at webRTC camp!) Maybe a chrome update was needed in this case
# 03:32 bret I should have rephrase my statement above to a maybe* talky isnt working...
# 03:32 Loqi [@kevinmarks] #indieweb @jaylovespie: with path I have to narrate my own story
# 03:34 bear oh, no worries, we are constantly working on it and testing - so yea, it will fail a lot for a lot of stuff :)
# 03:34 aaronpk bnvk: I haven't started a wiki page on travel plans but that's a great idea!
# 03:35 bnvk aaronpk: yah, i'll jot down all the things that occurred to me while cobbling together my page today :)
bnvk and smus joined the channel
# 03:39 bret aaronpk, what do you think about doing a HWC:PDX?
# 03:39 aaronpk I don't think it's a problem on my end cause I was using talky.io fine from here with esripdx people yesterday
# 03:39 bret KevinMarks firefox beta and updated chrome should work
# 03:44 aaronpk bear: Chrome 31. but it was working fine for me from here yesterday.
# 03:44 jp not working for me either on Chrome 31
# 03:45 bear we are trying to track down a chrome 31 change to the error handling currently - sounds like you are hitting
# 03:46 aaronpk oh weird, but nothing changed on my computer between yesterday and today
# 03:46 bear IIRC it's on the browser who started the talky session, not joins
# 03:47 bear so you have a video peer session, but zero data being sent to that pipe
# 03:47 bear right now we are seeing Opera 18 have the strongest quality with fewer errors
# 03:53 mathpunk what the-- aaronpk
# 03:53 bret aaronpk not sure if you saw the logs, what do you think about an event at mozpdx at the same time as the hwc in sf?
# 03:53 mathpunk time zones are fucked up... you must have slept, and then become awake again
# 03:53 mathpunk how is that possible
etymancer joined the channel
# 03:54 aaronpk mathpunk: lol yes, I went to sleep around midnight
# 03:54 aaronpk I'm at +0400 which is 12 hours ahead of PDX. it's a really confusing timezone shift
smus joined the channel
# 03:55 bret also, mathpunk welcome! I read your introduction yesterday
# 03:55 aaronpk also confusing is I'm getting on a plane tonight at 2am and landing in PDX Friday morning. I only have one night of sleep ahead of me before I land but you guys have 2.
# 03:55 mathpunk here is etymancer. he made my vps serve terrifying, dark-night-of-the-soul index cards to our mothers
# 03:56 mathpunk i have an introduction? jeeesus.
# 03:56 mathpunk i've just watched @mike_FTW raging at designers and i'm amped for sitemaking
# 03:57 mathpunk indiewebcamp: the friendliest panopticon
# 03:57 bret mathpunk i started my website with the goal of NOT learning html
# 03:57 etymancer mathpunk: so yeah, I'm here, but I'm also broccoli and stuff
# 03:57 bret but then I started reading about html5 and just kept reading and reading and reading
# 03:58 mathpunk broccoli > most
# 03:59 etymancer bret: new in the channel, aye
# 03:59 tantek and the first Homebrew Website Club meeting is now officially closed! Thanks everyone for coming and in particular KevinMarks for tweeting up a storm. :)
# 03:59 mathpunk ugh, spaces sorry
# 04:00 etymancer import urllib;urllib.quote(that)
# 04:00 mathpunk I asked @Eric_WVGG (aka D. Shwanky) to send me whatever archive of my last wp site was appropriate and I'll figure it out later. I want to start fresh
# 04:00 mathpunk "https://dl.dropboxusercontent.com/u/1717994/photos/Screenshot from 2013-11-20 10:59:32.png"
# 04:00 aaronpk etymancer: so there's apparently no web server on the new VPS if I found the IP address of the vps properly
# 04:01 etymancer yeah, no, it's pretty bare
# 04:01 aaronpk ideally we can get mathpunk set up with a minimal environment to start, just something to post HTML files in
# 04:02 etymancer I have a script for that
# 04:02 mathpunk (and log in to the COOL WIKI)
# 04:03 mathpunk I have beer and nicotine and rage-- now what?
# 04:03 etymancer so, this 66.33.209.151 - this is unfamiliar
# 04:03 mathpunk (where the wp site I mentioned is)
# 04:04 etymancer so, you want to set up a new subdomain and point it at the VPS?
# 04:04 aaronpk it sounds like mathpunk wants to just ditch the current site completely
brianloveswords joined the channel
# 04:04 mathpunk let's say, yes. I want to be building a thing-a-day(yeahright) portfolio, and have a subd for our app when it's more appy
# 04:05 etymancer ok, and what currently comes up when you hit mathpunk.net, do you want that to still be available or just burn it?
# 04:05 mathpunk burn it all, it's neglected
# 04:05 mathpunk i mean, i want the data-- I assume that's not lost if we just repoint the thing
# 04:06 etymancer love it with fire to death
# 04:06 mathpunk burn it to save it
# 04:06 mathpunk loqi: u r the best
# 04:06 mathpunk aaronpk: where're those loqi greatest hits?
# 04:07 etymancer hm, oughtta be a way to archive it? I'd hesitate to switch over the DNS just yet since it's a virtual host. just hitting the IP won't get you the contents
# 04:07 etymancer sorry, I'm not very dreamhost- or wordpress-savvy
# 04:08 mathpunk lemme quote Eric_WVGG
# 04:08 mathpunk "re: the old site, it’s on my crusty old server and I don’t know where many of the passwords were. Do you have any object to me changing the passwords now, backing up the old data, and potentially breaking access to the old site? This won’t prevent you from redirecting the site to the new home."
# 04:09 aaronpk mathpunk: maybe easiest is to do a wordpress export from the admin interface? then at least you can import into a new wordpress later if you want
b0bg0d joined the channel
# 04:10 etymancer yeah, that's what I was thinking. or if eric's got it backed up, that'll do too
# 04:10 mathpunk i can get in,doing so
# 04:11 mathpunk wow look at all these inspiring comments, apparently some actual people with not great english are v. impressed with my writrings and laysout
smus_ joined the channel
# 04:12 mathpunk "This format, which we call WordPress eXtended RSS or WXR, will contain your posts, pages, comments, custom fields, categories, and tags."
# 04:12 mathpunk I have downloaded that thing, and surely that will suffice
# 04:13 bear mathpunk - just be aware that uploaded photos will not be in the export data tarball
# 04:13 mathpunk bear: thanks, pretty sure I've uploaded none, or if I did they're flickr'd
# 04:14 bear it caught me by surprise, but fortunately I didn't burn the original before I had done my wp -> markdown conversion
# 04:15 mathpunk REMAIN CALM BURN EVERYTHING
# 04:15 etymancer ok, so. for the nonce I'm thinking I'll just set up nginx
# 04:16 mathpunk wouldn't it be easier for me to compile clojurescript to node?
# 04:16 etymancer apache seems a bit much for this
# 04:16 etymancer node can go behind nginx, but the latter can also handle static files
# 04:16 mathpunk i was trolling :(
# 04:16 etymancer which is what I thought was the first step?
# 04:16 etymancer I wondered, but low on cycles
# 04:17 bear nginx proxy pass thru to nodejs is very straight forward
# 04:17 etymancer 404 humor not found
# 04:17 mathpunk a brother can't troll anymore
# 04:17 mathpunk fsck,I've missed irc
# 04:17 bear sorry - i've been doing this for so long it's just far easier for me to ignore troll attempts
# 04:18 mathpunk the last time i was srsly on irc it was to roleplay vampires so--
# 04:18 mathpunk you guys have been in the shit
# 04:18 mathpunk i'll try and be good
# 04:18 mathpunk I would like to know what nginx is. I am guessing it is telling python, yo start a web server on port X and serve whatever you find in Y directory
tantek joined the channel
# 04:19 aaronpk mathpunk: nginx and apache are about the same, they are web servers that can both a) serve static HTML files and b) send requests to something else behind the scenes like php or node or python or other web servers
# 04:19 etymancer nginx is yeah
# 04:19 etymancer except for not ancient ;)
# 04:20 etymancer there is that
# 04:20 bear if you talk to node and go folks, they think nginx is long in the tooth and will tell you to just put the go code public facing
# 04:20 etymancer nothing could go wrong with that
# 04:21 etymancer hm, what the heck, I'll install them both and see which I can get working first
# 04:21 Loqi [@phoenix2life] RT @kevinmarks: #indieweb @jaylovespie: with path I have to narrate my own story
# 04:21 etymancer this is the multiest-tasking I've been in some time. hope I don't destroy anything
# 04:23 mathpunk etymancer: I trust you and it's not like there's much to preserve. just so long as I know what's happening and I can add new experiments
# 04:23 etymancer mathpunk: I'm more worried about the tofurkey
# 04:24 aaronpk once you get a web server like nginx or apache set up, you can do lots of experiments by using subdomains
# 04:24 mathpunk is worried about the tofurky
snarfed joined the channel
# 04:32 etymancer mathpunk: ok, it's "ready"
# 04:32 etymancer need to switch the DNS over
# 04:32 etymancer you have vr.org credentials, yes?
# 04:33 etymancer ok, so, log in there and look for something that looks nameservery :)
# 04:34 mathpunk hm.... although I didn't realize Seattle was in the north of China...
# 04:35 mathpunk mathpunk.net | A | 66.33.209.151
# 04:35 mathpunk change to livre's ip ya?
# 04:36 etymancer cool. due to the magic of internets, that will take "some time"
# 04:36 etymancer meanwhile, let's set up a new subdomain also cnamed to mathpunk.net
# 04:37 etymancer so you can actually see it
# 04:37 mathpunk what's this A/CNAME/TXT/MX?
# 04:37 etymancer say, test.mathpunk.net or something
# 04:37 etymancer I will explain this
# 04:37 mathpunk to the same ip?
# 04:37 mathpunk no port or nothin?
# 04:37 etymancer no ports in dns
# 04:38 mathpunk I've got to complain about this double-naming nonsense
# 04:38 mathpunk i told it test.mathpunk.net
snarfed1 joined the channel
# 04:38 mathpunk but it saves as test.mathpunk.net.mathpunk.net
# 04:38 mathpunk anyway, that's edited
# 04:38 mathpunk dunno, I just edited and corrected it
# 04:39 mathpunk oooh... will try next subd
# 04:40 etymancer point yer browser at test.mathpunk.net for a personalized message (tm)
# 04:40 mathpunk OMG WOW SERVER NOT FOUND
# 04:41 mathpunk it's what i always wanted
# 04:41 etymancer it behaves appropriately for me
# 04:41 mathpunk magic of internets
# 04:41 mathpunk take some time?
# 04:41 aaronpk ooh you have a CNAME to the IP address. make the CNAME record point to mathpunk.net instead
# 04:41 mathpunk maybe it takes time for the Seattle, China bits to get to New Hampshire
# 04:42 bear a CNAME is an altername name for an existing A record
# 04:43 mathpunk etymancer: so you mean.... the A record points to the IP, the CNAME test points to mathpunk.net instead of the digits?
# 04:43 jp i need to learn about dns lol
# 04:43 mathpunk oh, just like the www.mathpunk.net below
# 04:43 mathpunk Server not found
# 04:44 bear CNAMES are used to allow different subdomain names to point to the same server (aka A record) entry
# 04:44 mathpunk and the server knows it's serving different subdomains? the directory structure does?
# 04:44 mathpunk it looks, from this console, like it's all pointing at the same plac
# 04:44 bear that answer depends on what is receiving the socket connection
# 04:45 aaronpk you can tell nginx to do different things based on what name was used to get there
# 04:45 bear httpd's (aka nginx) know because the caller (the browser) sends a HOST header record
# 04:45 etymancer sorry trying to avoid dinner failr
# 04:45 mathpunk a POST or GET request includes a header. This header includes information that tells nginx what the browser is looking for.
# 04:46 mathpunk dinner > website
# 04:46 mathpunk do not starve my wife b/c i was impressed by Case's realtimeweb talk!!
# 04:47 etymancer well, if you insist
# 04:47 etymancer back in a bit then
# 04:48 etymancer also, setup so far is barest minimum to results
# 04:48 etymancer refinement will come
# 04:48 aaronpk mathpunk: so if you have access to that server you should be able to change "yo" into something that has your name on it
# 04:49 jp do you guys think its important to have your own virtual server with your web server stuff on it or is it ok to buy just web hosting?
# 04:50 mathpunk I have now made a computer being dreamed of by a computer I have never seen send the word "butts" across the country
# 04:50 jp i'd like the flexibility of having my own virtual server but i dont feel comfortable using linux
# 04:50 mathpunk I am as unto a god
# 04:51 aaronpk mathpunk: and now you just made the word "butts" fly across the world and land in a small screen in dubai
# 04:52 mathpunk afk, have 2 weep w/happiness
# 04:54 mathpunk update: the word "butts" apparently needs RESPONSIVE DESIGN 4 MOBILE
# 04:54 mathpunk which seems a bit much
# 05:01 mathpunk btw etymancer suggested I change the page to something with my name on it
# 05:02 mathpunk see, it's subtle
snarfed joined the channel
# 05:02 mathpunk these butts have my name on them OH HI
# 05:03 etymancer girl, are you these butts, 'cause wait
# 05:04 mathpunk Girl, are you these butts, because you are a conspicuous hill or mountain, especially one that attracts attention by its isolation, or serves as a land mark
# 05:05 etymancer so yeah, the mobile thing is probably just something goofy
# 05:06 mathpunk i was just trying to say "butts" to L
# 05:06 etymancer oh wait, no. send her the www.test.mathpunk.net link
# 05:06 etymancer that works mobileyly
# 05:08 mathpunk Gentlemen, we have butts
# 05:08 mathpunk Confirmed, we have butts
# 05:08 etymancer btw that was aaronpk's suggestion.
# 05:09 etymancer we are at t+butts
# 05:09 mathpunk comes close to spraying beer on the lab equipment
# 05:09 etymancer so, whew. all disasters averted
# 05:10 etymancer now to clean the kitchen. later to clean the virt
# 05:10 etymancer in between to clean the brain with sleep
# 05:11 dietrich bret: yeah next time let me know in advance, and if the space is open i'd gladly be the host for any pdx'ers that want to come
# 05:11 etymancer mathpunk: pm me later / demain for more
# 05:12 mathpunk etymancer: word. it's a site!
# 05:12 mathpunk thanks!!!!!!!!!!!!!!!!!!
# 05:12 etymancer my pleasure :)
# 05:12 etymancer also aaronpk: thanks for the help
smus joined the channel
# 05:15 mathpunk hooooooookay. next: rel=me
# 05:16 aaronpk i have to go in a minute, presentation's about to start
# 05:17 mathpunk if yours, "break a leg", else "enjoy!"
# 05:20 mathpunk INDIEAUTH'd!!!!
# 05:23 bret dietrich: way cool! I'll/someone will ping you once we work out the details
tilgovi and snarfed joined the channel
obensource joined the channel
# 06:01 bret Did anyone take a group photo at IWC:H or HWC#1?
snarfed joined the channel
snarfed and LauraJ joined the channel
# 07:13 snarfed bret: we took a couple at HWC #1 tonight. i think tantek and/or kevinmarks have them
tantek joined the channel
Jihaisse and tantek_ joined the channel
andreypopp and smus joined the channel
smus, friedcell, earplugs, brianloveswords, barnabywalters, melvster, tpinto and abrereton joined the channel
andreypopp and pfefferle joined the channel
KevinMarks and bnvk joined the channel
scor joined the channel
scor, bnvk and friedcell joined the channel
smus, bnvk, brianloveswords, mathpunk and snarfed joined the channel
_6a68 joined the channel
ozten joined the channel
tantek and friedcell joined the channel
# 16:31 snarfed progress report: webmentions from facebook are streaming into my inbox as we speak
b0bg0d, andreypopp, barnabywalters, friedcell, squeakytoy, mathpunk and ryana joined the channel
# 17:19 mathpunk Previously, on mathpunk.net....
# 17:20 mathpunk I now have an nginx running on a vps I "own" (rent), and it serves static web pages.
# 17:20 mathpunk Ergo, I have successfully indieauth'd, and then jumped up and down with the joy.
# 17:21 mathpunk I'm not sure what my next indieweb project ought to be, so I'm just going to start curling example web pages I like into a folder on my server for learning/forking.
# 17:22 mathpunk I read a little about the h microformat and I'm not entirely sure I get it, but my current Forever Project is trying to get the web to be more like stacks of shufflable index cards
# 17:22 mathpunk and I think that's probably relevant-- notes vs posts vs articles etc.
# 17:24 barnabywalters You should add some contact information to your homepage, marked up with h-card
# 17:25 mathpunk barnabywalters: Thanks! Couldn't done it w/o the channel
# 17:27 barnabywalters Just getting a name, URL and photo/logo covers most usage (e.g. Putting a name+face next to comments, posts and such things)
# 17:28 mathpunk I don't understand. It looks like, from the minimal example, that "h-card" is a css class. But what uses the class?
# 17:28 tantek barnabywalters that's a good point. We should point that out on the why page
# 17:29 barnabywalters Microformats are open vocabularies of classnames for common things like contact info and posts
# 17:29 mathpunk a selector then? I mean the thing that says <HTMLTAG class="foo">
# 17:30 bear "class" is nothing but additional data for an html tag
# 17:30 bear css uses that data for it's job, but it's not css that defines what goes in a class
# 17:31 mathpunk I've just heard the phrase "css selector" a lot, and one seems to select by tag type, class, or id.
# 17:32 bear asking questions like that means that someone later reading the log will learn from your question
# 17:32 bear and allows others to get a feel if the documentation is working
# 17:33 mathpunk what I mean though, is.... just adding the h-card class to a tag doesn't do anything in and of itself, right?
# 17:33 barnabywalters bear: In practise no one actually reads the logs :) hence why we solidify things onto the wiki
# 17:33 bear oh, then i'm definitely an odd case - I read more logs than wikis
# 17:33 barnabywalters mathpunk: There are open source parsers for various languages which, given HTML with microformats classnames, give you nice data structures to play with
# 17:34 bear any one class item is not important, but taken together when used within the html structure allows logical chunks of metadata to be parsed
# 17:35 bear <div class="h-card"> signals to a microformat parser that this div scope will hold nifty info that it should consider as an h-card
# 17:36 mathpunk when you're talking about these parsers, then, I speculate you're talking about sites in the indieweb ecosystem, that aren't necessarily my own?
# 17:36 barnabywalters Effectively, class=h-card tells the parser "this element represents a person/contactable thing"
# 17:36 mathpunk i.e., the "rel=me" trick
# 17:37 bear visit IndieWebify.me and look at some of the validation sections
# 17:37 bear that is an example of microformat usage and parsing
# 17:37 bear and also a handy way to double check your new site for progress
barnabywalters and snarfed joined the channel
# 17:41 jp tantek: you were mentioned in this book im reading which led me to this irc - which books do you recommend reading for html5/css/js and new web technologies
ozten and benwerd joined the channel
# 17:46 jp reading my way through introducing html5 second edition by bruce lawson and remy sharp
# 17:53 jp heard about whatwg from this book havent had a chance to check out the spec though - but bookmarked it
# 17:53 tantek barnabywalters - the WHATWG spec is a bit tl;dr
# 17:54 tantek There are a lot of HTML+CSS+JS book/references now
# 17:54 jp im especially interested in new web technologies
# 17:54 barnabywalters tantek: indeed, it’s not bedtime reading, but *much* nicer than the UA implementor-oriented specs for reference — easy to search too
# 17:54 jp and html5 for apps and the like
# 17:54 tantek jp - best thing to do is just get started with HTML5 using your own website
# 17:54 snarfed tantek: glad to see more work on that getting started page, thanks!
# 17:55 jp not yet i think i might buy one now
# 17:57 barnabywalters tantek: updates to getting started look good, I think the rest of the document needs some reprioritisation
# 17:57 jp if i host my website with a vps is it possible to still get mail accounts?
# 17:58 barnabywalters jp: setting up your own mail server is pretty tricky, but you can host your mail elsewhere and still point your domain to your VPS
# 17:58 barnabywalters that’s what I do on waterpigs.co.uk — HTTP requests go to my web server, but I get someone else to host my mail
# 17:59 jp i think google apps is discontinued :/
# 18:01 bear I use mailroute to be honest and then run a simple postfix server to receive mailroute's stream
# 18:01 bear layered on top of that is dovecot for imap,pop3 support
# 18:02 tantek similarly bear, how to use mailroute/postfix would be a good add to that page
# 18:02 tantek even just a sentence or two about what you're doing
KevinMarks joined the channel
# 18:03 bear i'll bring the page up and edit it while waiting for servers to update and such (i.e. my idle time during the day)
# 18:09 KevinMarks thoughts on better markup for these notes welcomed - should I add twitter accounts to the hCards?
# 18:10 KevinMarks as each line is also a tweet, I could link through, though that is a bit cumbersome
# 18:11 jp has no whois protection though :\ thats long
# 18:13 jp apparently .so domains do not support whois protection
# 18:15 barnabywalters KevinMarks: or maybe that livenote document is a h-entry which is also an h-feed
# 18:15 barnabywalters from a UI perspective, it is a series of short notes — that’s how it was authored and syndicated
# 18:16 barnabywalters and each item is replyable/mentionable/quotable and so probably deserves a permalink to itself
# 18:16 barnabywalters otherwise the tweet permalinks are more useful than the canonical permalinks ;)
# 18:16 jp im gonna set up some mails for my domain now ill have to check out google apps
# 18:16 Loqi barnabywalters meant to say: otherwise the tweet permalinks are more useful than the canonical permalink ;)
# 18:17 KevinMarks yes, that's why this is a bit odd. I should probably post each note as it happens, but that woudl mean using a doffernt process
# 18:17 mathpunk i'm trying to make my h-card and as usual i'm having trouble being the least bit professional
# 18:18 mathpunk barnabywalters: not trolling and graffiti'ing my own web presence :) mental problems this time, not technical
# 18:18 barnabywalters KevinMarks: it’s problematic when it’s unintentional, but .h-entry.h-feed > .h-entry might actually make sense
# 18:18 mathpunk BUG REPORT: user still not a grown-up
# 18:19 barnabywalters mathpunk: fortunately there are helpful people here. can you paste the code you’ve written in somewhere
# 18:20 barnabywalters KevinMarks: but I’d still try to give each item it’s own, non hash-fragment permalink
# 18:21 etymancer kind of like dishes
# 18:21 etymancer I'd much rather do someone else's dishes than mine.
# 18:22 mathpunk alright, i've selected a subset of h-card tags i'm willing to fill up, I'll stick some text on an etherpad. (etherpad's still a thing? cool.)
# 18:22 KevinMarks barnaby - the markup is generated inline on www.noterlive.com, then copy pasted at the moment;
# 18:23 KevinMarks I'd need to use something more databasy to make a permalink per note
# 18:25 barnabywalters mathpunk: is mathpunk.net still supposed to point to your wordpress site or is my cache not cleared properly?
# 18:25 jp oh theres no more free google apps
# 18:25 mathpunk no, we pointed it away
# 18:25 mathpunk i archived it and i'll figure out what to do with the content later
foo_ joined the channel
# 18:26 jp anyone using free alternatives for webmail?
# 18:26 jp ^on their own domains
# 18:26 mathpunk also i can't believe that people publish their email addresses on their sites-- it seems like that would encourage people to email you
# 18:27 jp would i install that on my server?
# 18:27 jp okay haha newbie with vps stuff but ill give it a go
# 18:27 aaronpk not saying I recommend it. I did say I *used* to use it :)
# 18:28 mathpunk there's this given-name/surname thing...
# 18:28 mathpunk seems a bit fine-grained~
# 18:28 barnabywalters yeah, I wouldn’t worry too much about the given name/family name unless you *really* want to publish it
benprew joined the channel
# 18:30 jp aaronpk: fair enough, think ill use outlook.com until i get a better grasp of installing stuff on a vps
# 18:30 KevinMarks well, I'd need a way to post the notes line by line like I do with the tweets, which would need a noting API
# 18:31 mathpunk Something I've just realized-- I prefer how "Thomas" looks in print even though everyone calls me "Tom"
# 18:31 mathpunk I'm going with the eye-rhyme
# 18:32 mathpunk I suspect I'm confusing the machine-reading with the person-reading
# 18:32 jp wish my name had a cool domain hack
# 18:33 mathpunk there are new top-levels with a weird smattering of nouns
# 18:33 mathpunk so maybe you can go for JULI.DIAMONDS
# 18:33 mathpunk but you have to become a mobster
# 18:33 barnabywalters mathpunk: don’t worry too much about the machine readable parts — just publish the bits of contact information you want in the what which makes sense, then scatter the mf2 classnames in afterwards
andreypopp joined the channel
# 18:34 bear mathpunk - I modified your name line to show an example
# 18:34 Loqi barnabywalters meant to say: mathpunk: don’t worry too much about the machine readable parts — just publish the bits of contact information you want in whatever way makes sense, then scatter the mf2 classnames in afterwards
# 18:34 mathpunk So style it like it's for a person, and then the attributes are how to let a parser know
# 18:35 barnabywalters because if you start with what you think will be interesting for a machine, it’ll be structured weirdly for real people, or just make no sense
# 18:36 mathpunk ohwow... the nested span made no sense, and then I realized it would let a machine greet me by first name while knowing my full name
# 18:36 mathpunk is enlightened
# 18:36 bear the <span> tag is a great way to split up the smaller bits but still keep the presentation as a whole item
# 18:39 etymancer mathpunk: ftfy
# 18:40 mathpunk you are "a great help"
# 18:40 etymancer goes to google h-card
# 18:40 etymancer yep. that's what google said too ;)
# 18:41 etymancer I like this microformat idea.
# 18:41 snarfed meanwhile: tantek forks kevinmarks's personal site, blows snarfed's mind
ozten and andreypopp joined the channel
# 18:44 mathpunk do people actually use mailto links? i use them just to remind myself that i have no intention of configuring an email client
# 18:45 barnabywalters I find them pretty handy. and they allow you to log to indieauth using persona, which is pretty handy
caseorganic joined the channel
# 18:54 KevinMarks well, he wanted to fix the typos in my notes from last night, and because I have a clunky publishing process, it's all in git
bnvk joined the channel
# 18:55 snarfed :P oh it's totally fine. i was just having fun with the meta/philosophical part. probably a sign of too much free time.
bnvk joined the channel
# 18:57 bear that's how I log into indiewebcamp - using persona via mailto sniffing
# 18:57 mathpunk what would a parser do if I have two p-regions? I was thinking of putting both the state (Oregon) and the bioregion (Cascadia)
# 18:58 jp is it safe to delete all the default zone files?
# 18:58 jp if my domain doesnt currently link anywhere
# 18:58 mathpunk I'm going to find out by doing. Maybe my site will make Cascadia real--
# 19:05 jp what do you guys recommend me installing on my vps if i wanna host my domain there?
# 19:06 barnabywalters mathpunk: I’m not aware of any consumers which actually *use* region data, but the default approach is to use the first value if only one is expected
# 19:06 bear jp, I would recommend using a service like FreeDNS if your providor doesn't give DNS support
# 19:07 bear running your own BIND daemon gets old really fast IMO (and i'm an ops guy in my dayjob)
# 19:08 jp i have a vps at digital ocean and i was thinking of using that
# 19:08 bear and I would start with nginx so you can host your static html5 site - then figure out as you find an itch what to install next
# 19:08 bear DNS should be provided by the folks you registered the domain name at
# 19:09 bear for me it's Hover - they have DNS support that is pretty decent
# 19:09 jp yeah i've got dns support at Gandi thats fine
# 19:09 jp just nowhere to point it to yet
# 19:09 bear oh, now I see why you said zone files - Gandi.net
# 19:09 bear yea, for the start define @ and maybe a www CNAME to point to @
# 19:10 jp do i point it to my vps' IP?
# 19:10 jp (sorry im oblivious to dns)
# 19:10 bear no worries, everyone is a newb at something every day
# 19:11 bear DNS is the internet's phone book - it maps bear.im to the address of the server that handles it's services
thatryana joined the channel
# 19:11 tantek snarfed, funny how github creates forks in your profile even when you just edit one file from another project.
b0bg0d joined the channel
# 19:12 snarfed yeah. makes sense though. it'd be nice if they cleaned up the fork afterward, but meh
# 19:12 tantek ideally I'd have most pages on my site editable by close/trusted friends.
# 19:13 tantek though I have to figure out where the boundary should be between generated pages (e.g. permalinks pages from storage) and standalone editable pages
# 19:14 tantek I *think* the boundary is pages that primarily have a date-time context vs. pages that primarily have a named/topic context.
# 19:15 bear oh, yes, I haven't indie-webified my code section
# 19:15 bear (this is a hand migration from wordpress)
# 19:15 tantek all that leaves is what to do about templates that the datetime-based pages use for generating home page, archive pages, permalink pages (maybe even tag pages)
# 19:16 tantek I don't have a URL design for template pages but perhaps I should - how bad would it be if people could see your page templates for generated pages?
# 19:18 tantek barnabywalters - yeah it is (was) not linked from anywhere
# 19:18 jp tantek: are you ex-twitter?
paulcp joined the channel
# 19:19 tantek jp - never worked at Twitter. Joined in the first year, gave a lot of in-person feedback back when they were <20 people, and blogged about it early on.
# 19:19 tantek hope he escaped reference in the Hatching Twitter book.
# 19:19 jp havent read it yet haha
# 19:20 tantek barnabywalters - so far I only have two things in that "template" that get filled in:
# 19:20 tantek <div class="section stream"> </div> <!-- where the home page composite stream goes -->
# 19:20 tantek <div class="section recent-articles"></div> <!-- what I coded at IndieWebCampHollywood -->
# 19:21 tantek my templates are valid markup in and of themselves (never understood why CMS templates had to use random other languages / escape codes)
# 19:22 tantek is wondering if we do publish our templates publicly, would it make sense to come up with microformat conventions for the pieces of a template that are meant to be "filled in" by a publishing engine that reads the template as well as a datastore of some sort.
# 19:22 tantek right now I'm using class names "stream" and "recent-articles" on my home page as "fill in this element" template hooks.
# 19:23 tantek barnabywalters - hence why I stuck with simple class names rather than anything microformatty.
# 19:23 tantek would rather avoid designing a big template engine.
# 19:23 tantek would sharing templates across systems be useful?
# 19:24 tantek one possible advantage - front end devs/designers could focus on making nice templates, and then swap out backend systems written by others.
# 19:24 tantek barnabywalters - I'm trying to use PHP files purely as code execution files (not the original intent of PHP)
# 19:25 barnabywalters you could just use HTML files with microformats and dummy/empty content in as the templating language — if the data to go in is microformats2 structured then you don’t need another vocabulary
# 19:25 barnabywalters tantek: I mostly do the same, and keep template files well separated from code files
# 19:26 barnabywalters PSR-0 requires files to either produce side effects or define symbols (e.g. functions, classes, constants) but never both, which makes a lot of sense
# 19:27 tantek barnabywalters, re: don't need another vocabulary - except in *every* instance I have of "filling something in" - it's had a special meaning beyond existing vocabulary
caseorganic joined the channel
# 19:27 tantek e.g. "stream" - yes it is a "feed" but it's a very specific type of feed - that for the *home page* (and yes I should have probably named it more specific)
# 19:28 tantek same thing with "recent-articles" - that's a box of just the article names linked to their permalinks
# 19:28 tantek and just a few of them - not sure how to capture that semantic of "how many recent articles"
# 19:29 tantek (right now it's a hardcoded constant in the code - though perhaps that number, 3, should come from the template? suggestions welcome)
# 19:30 tantek e.g. instead of "stream" what's a good way to represent "recent updates for the home page including all post types" or "… all post types except… fitness raw data" (or something) ?
# 19:30 barnabywalters tantek: hm, I’d say that kind of logic (figuring out special meaning) is the job of the code calling the template
# 19:30 barnabywalters rather than having to give the template the smarts to figure out exactly what sort of thing it’s displaying
# 19:30 tantek barnabywalters - except the template is typically *designed* with specific presentational intent, thus it should convey that information to the code filling in the template.
# 19:30 barnabywalters code calling the template knows what should go in the template based on the URL, the template should know where to put it
# 19:31 tantek like how should the code filling in the template know or care about 3 recent articles or 4? shouldn't the template determine that?
# 19:31 tantek KevinMarks - I think to work around a bug in an older version of my code. I could probably take that out.
# 19:32 tantek maybe I need to just study existing template (markup) systems like Blogger, Tumblr, WordPress (any other suggestions) to figure out what information the templates are choosing to convey to the engines underneath.
# 19:32 tantek I hoped someone here with template design experience would have opinions to offer.
# 19:33 barnabywalters it would be kinda cool to have declarative templates able to do queries of things
# 19:33 tantek right. but simple queries. I don't want to end up down the XPath/XQuery/SPARQL rabbithole.
# 19:33 barnabywalters typically they don’t at the moment because reusability, but in this case it might be better to put the reusability on the layer above
# 19:34 tantek simple queries based on real world template needs?
# 19:34 barnabywalters because then you can just put data-stream="articles" data-max="3" in *any* template and get useful stuff
# 19:34 tantek wouldn't it be nice if you could switch from Tumblr to WordPress or vice-versa but keep your templates that you'd put work into designing?
# 19:34 tantek sure - but the data-* attributes are supposed to be system-specific - not for anything cross-system or portable.
# 19:35 KevinMarks a lot of the templating languages assume a JSON blob being passed to them
# 19:35 tantek KevinMarks - because designers love authoring JSON blobs in templates?
# 19:36 tantek barnabywalters - yeah - that's pretty clear about them in the spec
# 19:37 tantek barnabywalters - however, since you can install Taproot on any site, it is independent of any specific site.
# 19:38 barnabywalters tantek: but if falcon and taproot both supported these templates, and they *used* them, then they would be… *using* them
# 19:39 mathpunk etymancer, barnabywalters: check it-- mathpunk.net bio card. not good, but better than bad~
# 19:39 tantek so they have hooks with data-twitter-* attribute just for their libraries
# 19:40 mathpunk and it appears to h-card validate
# 19:40 KevinMarks eg Jade is a pretty terse way to make HTML from a JSON structure
# 19:40 tantek not intended for any other software or libraries
# 19:40 etymancer mathpunk: woot!
# 19:40 etymancer defintely > bad
# 19:40 mathpunk etymancer: I have no clue where you-know-who is storing vim swap files... do you?
# 19:41 barnabywalters it’s a pity — they’re an effective way of applying user-customisable key-value pairs to elements
# 19:41 etymancer mathpunk: sudo lsof -n|grep '\.sw'
# 19:42 etymancer (typically they're in the same directory as the open file)
loleg joined the channel
# 19:44 tantek barnabywalters - yeah - I think that's why people resort to JSON blobs for that
# 19:45 barnabywalters tantek: not sure I understand — JSON blobs are data given by the calling code to the template, not data given by the template about what should be put in it
# 19:45 tantek barnabywalters <script id="blobnnn">JSON blob</script>
# 19:45 tantek in fact a lot of pages now send down data like that
# 19:46 tantek view source on your public Facebook profile page
# 19:46 barnabywalters oh right — that’s actual data to fill the template with rather than data given by the template about what it does
# 19:47 tantek it's a template that's filled in by clientside JS rather than serverside
# 19:47 barnabywalters but the equivalent in the original context (server side templates) would be storing all your notes in the template
# 19:49 barnabywalters okay, so the problem I was brainstorming solutions about was the template file itself giving the calling code information about what to fill it with, e.g. the number of posts to fetch
# 19:50 KevinMarks yeah, most of them have iterator syntax, not sure you can easily do a 'just the first 3'
# 19:51 tantek on my home page. the information about having *3* recent articles in particular - where should that go? I'm thinking that information should somehow be in the template, as I expect the designer of the template to determine that. Not the calling code.
# 19:51 KevinMarks I suppose you could emit all the html and hide beyond 3 with css, but thats a bit odd
# 19:53 tantek also potentially more expensive on the back end
# 19:53 mathpunk waaaaaaaat-- I'm moving on to the "h-entry" project but... articles are also h-cards? (??)
# 19:53 tantek "all the html" - you mean go get ALL my past articles?
smus joined the channel
# 19:53 tantek (of bandwidth - to send down that much in the home page)
# 19:54 tantek mathpunk, no articles are not h-cards. articles are h-entry which has a p-author which is an h-card
# 19:54 tantek KevinMarks - but it's not paginated. it's just a simple sidebar box of 3 recent article names linked to their permalinks
# 19:55 mathpunk tantek: Does all that h-card data have to go into the p-author? :/
# 19:55 tantek mathpunk - what do you mean by "all that h-card data"?
# 19:55 tantek KevinMarks - sure, and no reason to query for any more than the presentation needs
# 19:55 mathpunk mmm... well, I thought the h-card was a sort of bio-slash-contact card... so I put in my name and pic and location and homepage and whatnot
# 19:56 tantek mathpunk - typically, a simple URL+photo+name is great for a p-author h-card
# 19:56 mathpunk but I put that in a div for the end of the page.
# 19:56 mathpunk I was thinking of it as a sort of "about the author" footer
# 19:56 tantek KevinMarks - I don't understand why it is two pass.
# 19:57 tantek typical practice is to have mini-h-cards for the p-author
# 19:57 mathpunk but it gets duplicated in the posts? that seems weird
# 19:57 mathpunk maybe it just seems weird because I'm handcoding this
# 19:57 tantek or you can use rel=author to link from a mini name only h-card to your home page with a more expansive h-card
# 19:58 tantek KevinMarks: 1) read template, extract information about what is needed, 2) query datastore, 3) fill in template with information to generate a page.
# 19:58 tantek in step 3 you already know all the places to fill in so you're not doing a "pass"
andreypopp and jp joined the channel
# 19:59 tantek that's only one "read" (ergo pass) of the template file
# 20:00 tantek it's just holding onto pointers from the DOM elements found in pass 1
# 20:00 tantek that's not a "pass" in that nothing is "passed" by/over
squeakytoy joined the channel
# 20:03 KevinMarks you're always walking 2 things at once, the template and the data structure
# 20:04 tantek I guess I don't understand. I just pass/parse the template *once* into a DOMDocument. Then query it to find the nodes to fill in. Get the data. Fill them in. Then output the page.
jp joined the channel
# 20:06 jp lol laptop runs out of charge so plug it in then turn back on
# 20:07 jp and then it turns off again after realising it wasnt turned on in wall
# 20:09 etymancer waaaaaaaat++
# 20:09 etymancer (striving for balance)
# 20:10 etymancer no doubt. silly ascii, overloading punctuation
# 20:11 tantek Get the data is a batch query with parameters based on a) which template, and ideally also b) information from the template (like # of recent articles - which is currently hard coded in a constant in the code)
josephboyle joined the channel
# 20:19 tantek KevinMarks - it does seem like typical/popular templates do a lot more than we really need for indieweb sites.
paulcp joined the channel
# 20:25 tantek are there any template formats that let you edit "by example" and then fill in your "example data" with real data from the datastore?
# 20:26 KevinMarks I think google had over 20 internal template libs when I was there
# 20:26 tantek one challenge I have found with all templating languages/formats is that there is no way to quick-preview what a template might look like when you're editing it
LauraJ joined the channel
# 20:33 KevinMarks 'cos the {{...}}
in mustache/handlebars lets you insert queries on the JSON blob
# 20:35 mathpunk thanks for yr help y'all
b0bg0d1 joined the channel
# 20:39 tantek KevinMarks (or anyone) - any suggestions for semantic HTML for marking up a filename?
# 20:44 KevinMarks the question is are you writing in a suitable DSL or having to do it some annoying way
# 20:45 KevinMarks at one end you have regexes, at the other you have big nested datastructures you traverse for the bits you want
# 20:46 tantek KevinMarks why do I need a DSL rather than "just" using HTML+classes for the template hooks themselves?
paulcp joined the channel
b0bg0d joined the channel
# 20:49 KevinMarks the question is which ones turn into YAML and which ones stop at JSON, to use a metaphor
# 20:50 tantek KevinMarks - yeah I guess I'm naively confused about why they end up descending into so much complexity (and DSL-ness).
# 20:51 tantek So I'm going to gather use-cases I've encountered and just try to use simple HTML+classes for myself, and document as such.
# 20:51 tantek Kevinmarks - "did"? that sounds like past tense. Plenty of CSS3 modules still in progress. ;)
# 20:52 tantek (a lot of that, on CSS3, is spending non-trivial time sweating lots of little details)
# 20:52 tantek barnabywalters - really appreciate the prior art - was not aware
ozten joined the channel
# 20:53 tantek would love to see *any* of them document, hey this feature is here because of the following use cases.
# 20:54 tantek barnabywalters - what kinds of template use-cases does Taproot have and how does it solve them?
# 20:55 barnabywalters tantek: at the moment, dumb ones. I have some data in an array and I want to make it into HTML quickly
# 20:55 tantek it's ok, I'm pretty sure Falcon's use of templates is even dumber ;)
# 20:55 tantek but I figured I'd start the documentation somewhere
# 20:56 tantek huh, so far Aral's Tally is the least complicated.
# 20:56 barnabywalters yeah, TAL is from Zope, which AFAICT was big and complicated, with a strong focus on XML
# 20:58 tantek barnabywalters - yeah, I'm ignoring anything XML based figuring XSLT has that covered.
# 20:58 tantek I don't even have that. I have pieces of my code that look for specific template constructs.
# 21:00 bret jekyll sites are nearly 100% liquid templates
# 21:01 tantek bret - not sure what you mean by "liquid templates"
# 21:02 barnabywalters how come all these templating languages look almost exactly the same, yet they’re all completely incompatible
# 21:02 barnabywalters so someone could make a meta template library which could render data to all of them
# 21:03 barnabywalters what’s the name for when it’s socially acceptable to reinvent the wheel in some minuscule with a shiny site and a fancy name?
# 21:07 barnabywalters bret: ooh, so someone *has* made a classname (or at least CSS-selector) based HTML templating language
# 21:07 Loqi barnabywalters meant to say: bret: ooh, so someone *has* made a classname (or at least CSS-selector) based HTML templating library
# 21:08 tantek.com edited /Falcon (+685) "/* Working On */ archives, after other things in the list, link to templates as a reminder to iterate on that too." (
view diff )
# 21:09 KevinMarks the JS based ones are DSLs that compile to functions like my fugly ones
# 21:11 tantek This: "Special syntaxes invent another language to intermix with HTML and thus add programatic concepts to a declartive syntax—which is not clean separation no matter what you name it. "
# 21:11 tantek and this! "If you are trying to replace HTML or CSS with JavaScript, you are doing it wrong and have just signed a maintenance contract from hell, with yourself, for yours and your data’s life. "
# 21:12 bret XD also note in the view source… no class names! He is like an html5 monk
# 21:16 KevinMarks I suspect the use cases for a lot of these are "I have a big JSON blog from an arbitrary API, and I want to make something pretty"
# 21:16 tantek Kevinmarks - so they're working backwards then, from storage to UI?
# 21:17 KevinMarks or they're client-side people who don't get to change the API
# 21:17 tantek as opposed to, I've designed this awesome UI, now I need to specify how the variable content pieces of it can be retrieved from some datastore
# 21:19 tantek where the simple markup example of an element that is repeated?
# 21:20 barnabywalters so in this example the count is still determined by the data and not the template
# 21:20 tantek designers like to determine counts from the front end
# 21:20 barnabywalters but as they’re both loaded at the same time it would be trivial to slice $data by some attribute in the template
# 21:21 tantek sure - just wondering if that system has such a mechanism
# 21:21 barnabywalters agreed, I think having the template broadcast it’s capabilities/constraints is a great idea
# 21:22 KevinMarks I suppose I'm thinking single pass because of that, as the process is interactive
# 21:22 KevinMarks I never end up closing the last blockquote as I don't have a button for that.
# 21:23 tantek well "Templating With DOM" is the closest prior art to what I'm looking for so I'll probably use ideas/concepts from there
# 21:23 KevinMarks you could use the number of children in the source to decide how many times to iterate
# 21:23 tantek but I agree strongly with this goal: "… the HTML file itself can be designed independently of the software, and that whoever does the HTML doesn’t have to know PHP"
# 21:24 barnabywalters that has the advantage of the template looking as close to the intended output as possible
# 21:24 tantek and that way you could even fill in the children with sample output just to see what the design looked like while editing the template
paulcp joined the channel
paulcp joined the channel
# 21:26 tantek Kevinmarks - I like the idea of using elements to represent explicit counts as it's more designer-literal / preview friendly
# 21:27 barnabywalters this also completely avoids the need to create the “helpers” every templating language accumulates
# 21:28 barnabywalters and where they are more reusable, instead of being wrapped in framework padding
# 21:29 KevinMarks does that presume you always use <ol> or <li> for repeating elements?
# 21:29 KevinMarks or is the knowledge of what will repeat be encoded in the code?
# 21:30 tantek I think yes, there is an implication of (re)using HTML semantics for template semantics.
# 21:31 tommorris Remarkable looks... interesting. That it turns "//google.com" into "http://google.com" seems silly though. Protocol-relative URIs are a thing.
# 21:31 tantek Kevinmarks, so even <ul> vs. <ol> as to whether its ordered or not
# 21:32 tantek perhaps even using the new "reversed" attribute to indicate reverse ordering, e.g. most recent posts first
# 21:36 jp finally got my domain pointing to my vps and nginx installed but i'm not sure where the html files are located - theres no www folder
# 21:37 jp anyone have any ideas
# 21:38 bret there should be a default site defined there
# 21:39 bret someone dragged me through that configuration file the other day… still learning about it myself
# 21:39 jp there's something there... its not the default index.html its all a mish mash in notepad let me open in notepad++
tilgovi joined the channel
# 21:40 bret jp that is the config file… there should be a section that covers the default site
# 21:41 bret thats all i can help you with though… thats as far as I got
# 21:43 bret jp in that section there should be a path to the default site. This is all from memory though… you are going to have to poke around
andreypopp joined the channel
# 21:43 KevinMarks that was odd. my phone rang, and I accidentally hit the hang up button, and it transferred to hangouts on the chromebook
andreypopp joined the channel
# 21:44 KevinMarks reading this camendesgn site I want to send him a pull request to fix typos like tantek did for me
# 21:45 bret you could keep a charged, unpaid for cell phone around for emergency and just use google hangouts as a phone number for free…. aaaand you are no longer paying AT&T
mathpunk joined the channel
# 21:52 barnabywalters so, not sure whether I want to build my own DOM templating from scratch or try to work on kroc’s code
# 21:53 jp still need to fix the zone file thingies
# 21:53 jp and the MX records arent in the right place
# 21:54 jp need to redo it and use <article> <nav> etc
obensource joined the channel
# 22:06 barnabywalters if I reuse the querySelector and querySelectorAll method names from the JS DOM, will people be annoyed if they have to pass a context node in a la jQuery?
# 22:06 barnabywalters instead of having to wrap every single DOMElement in yet another wrapper class
andreypopp and abrereton joined the channel
# 22:19 barnabywalters tantek: I looked at krocs code and I am sure now. But I will document my thoughts anyway
# 22:19 tantek I'll likely consider your use cases while improving my own templating support
andreypopp and caseorganic joined the channel
b0bg0d joined the channel
caseorganic joined the channel
# 22:48 KevinMarks thinking about this a bit more, is not a page marked up in microformats effectively a template?
# 22:48 barnabywalters I’m not building a system which strongly couples classes to behaviour, basically just a DOM manipulation toolkit optimised for template use
# 22:48 barnabywalters KevinMarks: no reason it can’t be, just make the templating library replace mf2 classnames with values
# 22:49 KevinMarks you can also parse it to get the shell left by that JSON, and replace it with something else
# 22:49 barnabywalters HTML -> mf2 parser -> JSON structure -> transformation -> HTML -> (loop back)
# 22:52 KevinMarks 'cos tally is telling you what to replace in the element, but mf2 does that too.
# 22:53 barnabywalters I’m hitting up against an issue: css selectors can’t select individual attributes
# 22:54 KevinMarks !tell tantek mf2 syntax is complete enough to specify what part of the element is replaced with what, just like tally
# 22:54 Loqi Ok, I'll tell him that when I see him next
# 22:55 barnabywalters so is another method required? setAttribute('element selector', 'attribute', 'value')
b0bg0d and bnvk joined the channel
# 23:12 barnabywalters bah, I can’t get DOMDocument to retain whitespace in the <head>, only in the body
CheckDavid, tpinto, poppy, kylewm and tantek joined the channel
# 23:58 Loqi tantek: KevinMarks left you a message 1 hour, 4 minutes ago: mf2 syntax is complete enough to specify what part of the element is replaced with what, just like tally