#indiewebcamp 2012-11-18

2012-11-18 UTC
dascher and tantek joined the channel
#
@HeyTotah2012 "@t: #indieweb software should be easy to install/update/revert portably across web hosting services. cc @maxwellsalz #MVRM (ttk.me t4Lz2)"h
zztr joined the channel
#
tantek how odd
#
tantek and their tweet has already been deleted
jfranusic joined the channel
#
tantek welcome jfranusic!
#
jfranusic thanks!
#
jfranusic okay, so I'm moving all my data back to my own domain
#
jfranusic figuring out how to structure my (existing) data is a little daunting
#
jfranusic is there a suggested set of default sub-directories to have?
#
jfranusic for example /js, /css, /img, etc
dascher joined the channel
#
tantek jfranusic - those all seem like "support role" directories
#
tantek rather than the essentials of your URL structure for your permalinks
#
tantek somewhere on our wiki I think there may be some documentation of this
#
tantek I know tommorris has strong opinions
#
tantek jfranusic - the reason I point this out, is that I found that the structure for my js, css, images etc. *followed* from once I figured out the URL structure for my permalinks, posts etc.
#
tantek so consider starting with that design problem instead
#
jfranusic hmm, okay
#
tantek.com edited /Web_Actions (-1) "r" (view diff)
#
tantek.com edited /POSSE (+0) "/* See Also */ s/_/" (view diff)
#
tantek.com edited /Publish_Then_Syndicate_and_Replicate (+141) "see POSSE for current thinking" (view diff)
#
tantek.com edited /Publish_Then_Syndicate_and_Replicate (+18) "link session to 2011 schedule for context" (view diff)
#
tantek jfranusic - you can see some of my thinking on personal site URL structure here: http://tantek.com/w/Whistle
#
tantek !tell shaners read through the entirety of http://indiewebcamp.com/ActivityStreams for a summary of issues and how we (here in the #indiewebcamp community) are using pieces of ActivityStreams and ignoring other pieces.
#
Loqi Ok, I'll tell them that when I see them next
shaners joined the channel
#
Loqi shaners: tantek left you a message 3 minutes ago: read through the entirety of http://indiewebcamp.com/ActivityStreams for a summary of issues and how we (here in the #indiewebcamp community) are using pieces of ActivityStreams and ignoring other pieces.
morrocco_mole, dascher, tantek and lmorchard joined the channel
#
@xtof_fr "Haché ou cloné : á vous de choisir" http://christopheducamp.com/wiki/indielog-4 /by @barnabywalters luthier sur l'#indieweb
#
tommorris tantek: one of the things SemWebbers have got right is "follow your nose". standard URL structures bad
#
tommorris being able to discover where things are by following links good. it's good we now have the ability to point to a favicon in a <link> tag rather than having to put it in a specific place.
#
tommorris things like favicons are relatively harmless, but if people start dictating URL structures for content, it means people can't experiment.
#
aaronpk i know what you mean and I mostly agree with the discoverable URL structure
#
aaronpk but with favicons, it's nice to be able to always know where it is and what size it will be without needing to make an http call first and parse the page
#
tommorris well, content negotiation on /
#
aaronpk for example for example I know this will work on almost every site: <img src="http://slideshare.net/favicon.ico">
#
tommorris 200 GET /
#
tommorris Accept: image/vnd.microsoft.icon
#
tommorris ;)
#
aaronpk need a way to specify Accept headers in an img tag then ;)
#
tommorris Indeed.
#
tommorris tantek: btw, I signed up for UXCampBrighton. Am going to give a session on the evolution of posting UIs over the years. I'm no UX designer, but having a bunch of smart UX people in the room might be a way to sound out better ways of designing posting UIs.
catsup and barnabywalters joined the channel
#
tommorris indiewebsters, what I'm working on: http://cl.ly/image/3V2Y281F3w3D
friedcell and barnabywalters joined the channel
#
tommorris barnabywalters: you may be interested - http://cl.ly/image/3V2Y281F3w3D
#
tommorris first step on the path to indie checkin. ;-)
#
barnabywalters tommorris: looks good
#
tommorris using OpenLibrary
#
barnabywalters I wonder if extending the checkin verb to media dilutes it at all
#
tommorris it's in use on GetGlue, a check in service people are using
#
barnabywalters I’m all for minimising the number of verbs we deal with, but checking into a place and a piece of media seem fundamentally different
#
barnabywalters okay, so there’s some existing usage
#
tommorris http://o.getglue.com/tommorris/all?action=recent
#
barnabywalters so they seem to be using 'check-in' as an umbrella verb which covers think, read, watch
#
tommorris so, checking in to a place is one of the things I'm working on. in the application, I'm handling the geo checkins very similarly to the media checkins
#
tommorris.org edited /checkin (+20) "/* see also */ adding link to [[media checkin]]" (view diff)
tantek joined the channel
#
barnabywalters well, checkin to a place is really quite hazy and represents a whole load of use cases, so I don’t see any reason not to use it for media checkins
#
barnabywalters as a generic “consume” verb instead of read, watch, think, like…
#
barnabywalters so what is the UI for checking in to a piece of media? Search for it, click checkin, optionally add a comment?
#
tommorris yup
#
barnabywalters amazon affiliate links? ;)
#
tommorris potentially
#
tommorris using OpenLibrary because it's CC Zero and AnyoneCanEdit. ;-)
#
barnabywalters looks up openlibrary
#
tommorris.org created /media_checkin (+479) "Created page with "'''Media check-in''' is a type of check-in or post. == Existing services == * [http://getglue.com/ GetGlue] allows you to check in to TV, films, music and other media. [http://o..."" (view diff)
#
tommorris.org edited /media_checkin (+97) "adding implementation" (view diff)
#
barnabywalters I LOVE the fact they’re doing REST with extensions added to existing URLs determining the content type
#
barnabywalters tests to see if they’re handling Accept: headers as well
#
tommorris I'm currently just adding .json to the search URL
#
barnabywalters that’s great. I wish more people would offer that
#
barnabywalters aw, they only change the content type if an extension is provided, not if you send an accept header
#
tommorris the search ranking on OpenLibrary does kind of suck, sadly.
#
tommorris yeah, that kind of sucks too.
#
tommorris on Amazon, they add things like sales rank to the search ordering for book
#
tommorris so, a self-published book on OpenLibrary can rank higher than a major scholarly work from a proper academic publisher.
#
tommorris which wouldn't happen on Amazon
#
barnabywalters That’s a pity. Although I suppose for media checkins you’re more likely to be searching for a particular book anyway, not a general topic
#
tommorris I may bug them about it. perhaps have it so that books with Wikipedia pages rank higher
#
barnabywalters oh my god. this standard for embedding bibliographic data in HTML is hideous: http://ocoins.info/
#
barnabywalters cc tantek RE h-cite
#
tommorris heh, COinS is a horrible monstrosity.
#
barnabywalters The accessibility issues don’t bear thinking about
#
tommorris I've been using RDFa with BiblioOntology and Dublin Core.
#
barnabywalters btw tommorris, have you seen that RDFa is becoming a standard for
#
barnabywalters inline editable CMS data?
#
tommorris I hadn't
#
barnabywalters there are a couple of js libraries and CMSes which use it to denote which parts of a page are editable
#
barnabywalters http://createjs.org/
#
tommorris ah, it's the German crew who previously worked on dbpedia. they like their RDFa ;-)
jancborchardt, josephboyle and friedcell joined the channel
#
tantek tommorris, great idea on doing a session on posting UI/UX - eager to hear how it goes.
#
tantek barnabywalters, yeah, COINS is just *one* of *many* disparate attempts to do bibliographic/citation embedding - see http://microformats.org/wiki/citation-formats for tons of documentation and research on this.
#
tantek the short answer is - they're all overdesigned from a web publishing perspective. all of them. hideously so.
#
tommorris tantek: it's mostly going to be testing my contention that on-page-posting is necessary and seeing if there are people there who have better ideas.
#
barnabywalters oh god — that is a tonne of research
#
tantek I've been toying with an ultra simplified citation microformat just for citing stuff on the web, and then when a sessions was proposed (by someone else) at IndieWebCamp2012 (in Portland) on Web Citations, I jumped right in and worked on it there. Now it's on every one of my articles.
#
tantek http://microformats.org/wiki/citation-brainstorming#web_citations
#
tommorris the primary use case for me for citation is conversion. the amount of time I spend typing Wikipedia citation templates. and when I was a Ph.D student, I spent an equal amount of time fixing up BibTeX files.
#
tommorris make it all stop. there are a limited number of keys I can push in one day without turning into the RSI monster.
#
tantek here's the indieweb session that pushed me to publishing what I'd figured it and formalizing h-cite to make it ready for web use: http://indiewebcamp.com/2012/Academic_Citations_Web
#
tantek tommorris - word
#
tommorris I've been writing a BibTeX to Wikipedia converter. because most of the academic databases have copy-and-paste BibTeX
#
tantek especially re: "there are a limited number of keys I can push in one day without turning into the RSI monster." <-- I think that's best argument I've seen *for* microformats (especially microformats2) and against RDFa, microdata etc. which all involve a TON more typing even for (especially for) simple things.
#
tommorris I know there's a discussion going on about rolling out microformats2 h-cite on wikipedia. I probably ought to participate in that discussion.
#
tantek anyway - would be interesting in feedback on h-cite, especially by IndieWeb folks using it on their own sites (both for copy/paste markup like on my articles, and in the article markup itself when citing other things on the web).
#
tantek tommorris - seriously?!? whoa, didn't expect that so soon.
#
tantek please drop some links about that here or on #microformats so we can at least follow along
#
tommorris I haven't seen the discussion myself
#
tommorris I think Andy instituted it
#
tantek I can bump up my prioritization of writing up a more formal "h-cite" draft spec with more examples to make it easier.
#
tommorris someone asked me in my capacity as a general techie to say why Andy was wrong about it.
#
tommorris and I told them frankly that although there's no standard yet, and that I generally prefer things that are RDFish, compared to what we currently have (some combination of COinS and nothing), it'd be a big step up
#
tommorris And then I descended into obscenity at the awfulness of COinS
#
barnabywalters does wikipedia currently use COinS?!
#
tommorris barnabywalters: on Special:BookSources. not in citation templates, thankfully. we're not that crazy.
#
tantek tommorris, speaking of things RDFish, sometime (when we have perhaps more than few minutes), I'd like to chat with you about outputting canonical RDF triples from arbitrary microformats2 markup.
#
tommorris that should be relatively easy.
#
tantek I'm fairly sure it's possible, and I'm hoping that we can settle syntax arguments that way (since microformats2 syntax is easier to remember, less typing etc. it's more usable), since if you want JSON, you can get it, and then if you want RDF triples, we can make sure you can get those too.
#
tommorris I haven't looked at microformats2 in detail, but it should be easy enough to just declare a microformats2 namespace and define a parsing spec sorta like Hixie tried to with microdata
#
tommorris https://en.wikipedia.org/wiki/Template_talk:Citation/core#Removing_COINS_metadata
#
tantek right. I've already done the work of defining a parsing spec. barnabywalters has implemented it
#
tantek now all I need to do is add what triples are generated (in addition to the canonical JSON)
#
tantek parsing spec: http://microformats.org/wiki/microformats2-parsing
#
tantek ok, off for now, looking forward to figuring this out more later with you tommorris!
#
tommorris got my when-brain-functioning reading ready then. ;-)
#
tantek great
#
tommorris (may even be possible just to define a microformats2 -> json -> json-ld transformation)
#
tommorris json-ld is the newest sexiest way of doing RDF for the JavaScript crowd. ;-)
#
barnabywalters so it looks like json-ld is just json with some metadata
#
barnabywalters should be easy to add that metadata to canonical µf2 JSON
#
tommorris yup. ;-)
#
tommorris that said, it might be easier if I just define a standard RDF mapping. I've found problems with the RDF spec that W3Cers managed not to. ;-)
Poopman joined the channel
#
Poopman Hey
#
barnabywalters hey Poopman
#
Poopman Yay someone is here lol I'm lonely :(
#
Loqi :D
#
Poopman Hey loqi
#
Loqi who, me?
#
Poopman Both of u :)
#
Poopman You still there?
#
barnabywalters yes
#
Poopman Cool
#
Poopman What's your name?
#
barnabywalters Barnaby
#
Poopman Oh lol
#
Poopman I'm kian
#
Poopman Are you an operator?
#
barnabywalters no, I’m a regular user
#
Poopman Oh ok
#
Poopman Ill see you later bye!!
tilgovi joined the channel