#indiewebcamp 2013-11-12

2013-11-12 UTC
#
barnabywalters
goodnight all
#
Loqi
sweet dreams
thatryana joined the channel
#
KevinMarks_
in other words, complicated
bnvk joined the channel
#
pdurbin
aaronpk: at least for higher ed in the US, there's InCommon, where institutions trust other member institutions: https://incommon.org ... I'm just learning about it though... fuzzy on the details... but I've been asked "Is your Service Provider already registered with InCommon?"
#
aaronpk
wow fascinating
#
XgF
aaronpk: Bridging OAuth and SAML is one of the reasons why OAuth 2.0 is a bajillion specs and incomprehensible
#
KevinMarks_
and then there's SCIM
#
XgF
Theres OAuth 2.0 + Bearer tokens (like OAuth 1.1 but simpler)
#
XgF
OAuth 2.0 + HMAC (Pretty much OAuth 1.1)
#
XgF
OAuth 2.0 + SAML (Enterprisey!)
#
XgF
OAuth 2.0 + JWT ("We reinvented SAML in JSON because we want to sell you our software again")
#
aaronpk
oh is JWT similar to SAML?
#
XgF
That was the whole impeutus
#
aaronpk
oh good
#
XgF
JWT is how you sign an assertion. SAML is how you sign an assertion
#
XgF
Of course the whole JOSE committee are a bunch of retards
#
XgF
(Hey, we're designing a spec in the 21st century! Of course mandate support for RSA with PKCS#1v1.5 padding, that hasn't been broken for OVER A DECADE!)
#
KevinMarks_
"However, as a rich and highly extensible framework with many optional components, on its own, this specification is likely to produce a wide range of non-interoperable implementations."
#
aaronpk
that was one of Eran Hammer's last edits to that document
#
XgF
The sensible members of the IETF basically said we're not publishing it unless you rename it a framework and point out that everybody's version of OAuth 2 is different
#
aaronpk
I laughed a lot when I saw him add that
#
XgF
In reality OAuth 2 is pretty simple, it's just unfortunately defined in about 5 times as many specifications as is actually necessary
#
bret
Is there some sort of Oauth2.thegoodparts? A subset of Oauth 2 that is deemed tasteful?
#
bret
or is that what oz was trying to do?
#
XgF
OAuth 2 Bearer Profile is pretty simple. OAuth 2 with HMAC is as secure as OAuth 1.1. Both are good
#
XgF
Unfortunately, you've gotta piece things together over 3(?) specifications for each of them
paulcp joined the channel
#
XgF
I wish somebody would build an "OAuth 2.0 Basic Web /Protocol/" which takes those three and packages them together as a replacement from 1.1
#
KevinMarks_
isn't that what's happening de facto?
#
XgF
Well Bearer is what the webby people are going to use anyway
#
XgF
(Bearer over TLS, HMAC over HTTP)
#
pdurbin
Eran Hammer... the author of this post I was just looking at: http://hueniverse.com/2012/07/oauth-2-0-and-the-road-to-hell/
paulcp, caseorganic, snarfed and tantek joined the channel
#
tantek
brianloveswords there are no good books on the subject of federated networks because there are no good federated networks. yet. email may have had a glimmer but it being crushed under overwhelming spam and will die as a result.
#
tantek
there are many aspirational posts about federated social networks (such as the one hadleybeeman pointed out) but few and far between are actual good posts about them. here is a list of a few: http://indiewebcamp.com/Posts_about_the_IndieWeb
paulcp joined the channel
#
tantek
perhaps this would be a good FAQ
acegiak, astrolin, snarfed, skinny, melvster, bnvk, neirpyc, tantek, b0bg0d, earplugs, bret, andreypopp, cweiske, indiewebcamp-vis, pfenwick, jgraham909, skinny_, LauraJ and Jihaisse joined the channel
bnvk, LauraJ, KevinMarks, josephboyle, friedcell, julien51 and barnabywalters joined the channel
#
tantek.com
edited /IRC (+30) "w3c irc page"
(view diff)
#
tantek.com
created /Email (+19) "r"
(view diff)
#
tantek.com
edited /FAQ (+641) "Any books or posts on federated networks"
(view diff)
bnvk and pfefferle joined the channel
#
waterpigs.co.uk
edited /FAQ (+0) "/* Any books or posts on federated networks */ typos"
(view diff)
#
waterpigs.co.uk
edited /FAQ (+4) "/* Any books or posts on federated networks */ linky"
(view diff)
#
@pfefferle
@haschek und das macht das #indieweb eben sehr gut http://indiewebcamp.com/posts#Kinds_of_Posts es definiert beispiele und produziert funktionierenden code
(twitter.com/_/status/400214718104616961)
caseorganic, skinny and andreypopp joined the channel
bnvk, pfefferle, andreypopp and acegiak joined the channel
#
@haschek
@pfefferle Coverspace scheint tot, Repo leer, Link tot. Webmention und die RSVP party gingen auch durch die FSW/RWW-Comm., Danke für Idno.
(twitter.com/_/status/400228444211904512)
#
@haschek
@pfefferle Ich mag das #IndieWeb auch, 2 Apps von mir gehen auch in die Richtung #PESOS: #Foafpress und #PubwichFork (keine Archivierung)
(twitter.com/_/status/400229204551143424)
adactio, caseorganic, scor, andreypopp, Acidnerd, LauraJ, caseorga_, tpinto, XgF, tantek and bnvk joined the channel
#
Jihaisse
pfefferle: hello, Have you get the time to have a look at the plugin ?
#
pfefferle
Jihaisse I had, but it's very confusing with all the nested "if"s
#
pfefferle
the wordpress stuff is looking fine and the rest seems working ;)
#
Jihaisse
pfefferle: yeah, the code in the "Social" part of the plugin, is directly copy/pasted from the Social plugin
#
Jihaisse
I started by thinking : "this works, don't change it"
#
pfefferle
should be ok though ;)
#
Jihaisse
ok, do you think it would be easy to add another plugin ?
#
Jihaisse
pfefferle: btw, do you make POSSE on your wordpress ?
#
pfefferle
I had a look at the Jetpack plugin, but I couldn't find the ids of twitter or facebook, seems that they don't save it.
skinny joined the channel
#
pfefferle
yes, I use JetPack
#
Jihaisse
there is nothing in the DB ?
caseorganic joined the channel
#
pfefferle
Jihaisse no
#
Jihaisse
too bad
ozten, heath, brianloveswords, LauraJ, bnvk and josephboyle joined the channel
#
aaronpk
ozten: nice
#
ozten
aaronpk: Sitting in draft folder for too long
#
aaronpk
ah yeah
#
ozten
but I haven't been hacking on anything concretely
#
ozten
so I guess putting the perspective out there is better than nothing
#
ozten
and I hope to start prototying some of the flows
#
aaronpk
that would be cool
#
barnabywalters
ozten: nice post!
#
barnabywalters
another consumer-oriented OS concept
#
ozten
barnabywalters: No, I just got that reference an hour ago in #identity (mozilla)
#
ozten
I look forward to checking it out
#
ozten
thanks for the reference
#
@momowilly
Help bring back @ReaganGomez's webseries 'Almost Home Season 2' http://www.indiegogo.com/projects/almost-home-season-2/cstw/368172 #indiegogo #indieweb #indiefilm #womendirectors
(twitter.com/_/status/400296536539541504)
#
@ReaganGomez
RT @momowilly: Help bring back @ReaganGomez's webseries 'Almost Home Season 2' http://www.indiegogo.com/projects/almost-home-season-2/cstw/368172 #indiegogo #indieweb #indiefilm #wome…
(twitter.com/_/status/400296944599199744)
#
tantek
ozten - cool. love the brainstorming out loud. looking forward to seeing how you take steps toward your vision with your own online presence :)
#
ozten
tantek: email would be a big checkmark - I already host quite a bit of me
#
ozten
but, ya… get crack
#
tantek
barnabywalters - what is that?
#
Loqi
that is deemed tasteful
#
ozten
crack'n :)
#
barnabywalters
tantek: GDS? another consumer-oriented server OS concept
#
tantek
barnabywalters - there doesn't seem to be anything consumer-oriented. all the docs seem developer-oriented.
#
barnabywalters
yeah, very much so.
#
tantek
so it's a lot of blue sky and no actual building then?
#
tantek
well at least people are showing a desire to solve the problem(s)
#
tantek
maybe they'll eventually be frustrated with lack of progress and come around to shipping something simple/minimal on their own site.
#
aaronpk
that does seem to be a theme
#
@PoliticalJones
RT @momowilly: Help bring back @ReaganGomez's webseries 'Almost Home Season 2' http://www.indiegogo.com/projects/almost-home-season-2/cstw/368172 #indiegogo #indieweb #indiefilm #wome…
(twitter.com/_/status/400298195571728384)
#
tantek
aaronpk, choosing what to minimally build and build it is hard. I'm hoping that /Getting_Started and /IndieMark help with that.
#
tantek
and deploying something minimal and incomplete on your own site is scary. I'm hoping all the "IndieWeb examples" sections on all the building blocks help with that by showing that it's ok to be awkward and incremental.
#
tantek
it's so much easier and safer to simply brainstorm APIs.
#
tantek
and wishlists
#
tommorris
awkward and incremental is the only way to be. perfection requires time. time otherwise taken by having a job and a long commute. ;)
_6a68 joined the channel
#
tommorris
back in the old days of Wikipedia there was a guideline called "leave something undone" - still sort of documented here - https://en.wikipedia.org/wiki/Wikipedia:Make_omissions_explicit
#
tantek
it's a good collaborative principle
#
tommorris
if someone comes to my site and says "this sucks, I can build something better", we win. :)
#
barnabywalters
kinda like how discussions based on facts tend towards better facts, discussions based on opinion tend towards shouting
LauraJ and b0bg0d joined the channel
#
barnabywalters
how does one tell whether or not the name of a h-entry should be shown in addition to it’s content?
#
barnabywalters
e.g. for an article, the name will be short and different from the content, and expected to be shown
ryana joined the channel
#
barnabywalters
whereas for a note, the name is either the content, or a substring of the content
#
barnabywalters
and is not expected to be shown
#
barnabywalters
at the moment I’m just comparing the name to the plaintext value of the content, but that doesn’t always work for various reasons
#
aaronpk
good question! i wonder that too
#
barnabywalters
e.g. different whitespace collapsing, JS in note name but not content (in the case of idno checkins)
#
barnabywalters
just seeing whether or not the name is *shorter* than the content will work, given some constraints…
#
barnabywalters
1) that the name is not a truncated version of the content (which I’m pretty sure noone is doing at the mo)
#
barnabywalters
oops, I take that back, the content *does* have the js in in that case
friedcell joined the channel
#
tantek
KevinMarks: "…moving discussion of my videos to reddit instead"
#
tantek
so she's talking about moving to a comments silo, rather than indieweb
#
tantek
or did I miss something?
bnvk, Acidnerd, bnvk_, benprew and snarfed joined the channel
#
KevinMarks
She's moving her posts and videos to her own site. We should help her with comments
#
aaronpk
oh I saw her talk at xoxo!
friedcell joined the channel
#
XgF
barnabywalters: I find the existing delivery of notes unsatisfying. IMO they should be marked up /without/ title
#
barnabywalters
unsatisfying? how so?
#
XgF
Because it often includes a title which is just a duplicate of the content
#
tantek
barnabywalters - the simple algorithm is if the name of the note is the same as the content, don't show the title
#
barnabywalters
mf2 structures always have name properties due to implied parsing. better that it’s explicitly defined by the author than left to whatever the innertext of the root mf element is
#
barnabywalters
tantek: in theory that works, in practise it isn’t
#
aaronpk
tantek: thats the problem, determining whether they are the same
#
tantek
you can add to that: if the name of the note ends with an ellipsis (… or … (unicode)) then drop the ellipsis and if the result is a prefix for the content, then don't show the name
#
barnabywalters
probably because of whitespace collapsing inconsistencies in the parser
#
barnabywalters
but that can be normalised by the consuming code anyway
#
tantek
barnabywalters - alright, that sounds like a challenge to write up the algorithm and start collecting issues based on real world examples
#
XgF
barnabywalters: Then I'd say MF2 is buggy
#
barnabywalters
so I’ll try improving my implementation until I’m happy with it, then draft an algorithm based on that
#
barnabywalters
XgF: implied name parsing has many significant benefits
#
XgF
Such as?
#
barnabywalters
<span class="h-card">Barnaby Walters</span>
#
barnabywalters
instead of the longer version which I can’t be bothered to type
#
barnabywalters
also, ensuring that all mf structures have *some* textual representation to help identify them e.g. in lists
#
XgF
Perhaps the appropriate thing to say is that "For objects of type note the name may (should?) be ignored by recieving applications"
#
barnabywalters
there is no note type, unless you could h-as-note
#
barnabywalters
everything’s just an h-entry
#
XgF
As someone who deals a lot with ActivityStreams its' easy to forget what MF2 is missing...
#
barnabywalters
s/could/count
fmarier joined the channel
#
tantek
XgF - how do you deal "a lot with ActivityStreams"? they seem to have nearly no adoption on the public web.
#
tantek
(and have numerous issues of being overdesigned, see: http://indiewebcamp.com/activitystreams#Issues )
#
@g_m_dunn
#goldencheetah 3.0 up and controlling the #computrainer on Ubuntu 13.10, @TheSufferfest #OwnYourData
(twitter.com/_/status/400314588899454976)
josephboyle joined the channel
#
tantek.com
edited /note (+983) "add note display and note type algorithm for starters"
(view diff)
#
tantek
I'd be interested in seeing what exceptions you've found to that
#
tantek
as when I last looked at this problem and even wrote a little code - those two cases handled it no problem
#
tantek
maybe the trim() was taking care of the whitespace differences you are seeing
#
tantek
but if so maybe we need to better specify whitespace handling (I thought we had)
#
tantek
(in uf2)
#
tantek.com
edited /note (-1) "/* Note Type Algorithm */ of"
(view diff)
#
aaronpk
I think whitespace is still a problem. not sure if the problem is in the source markup or the parsing though.
#
aaronpk
and if in the source, then there's an issue with authors not wanting to change the markup to support a better parsed result
#
barnabywalters
the question is should the parser collapse whitespace?
#
aaronpk
certainly not if it's inside a <pre> tag... and then there's the CSS issue
#
barnabywalters
that’s the difficulty
#
barnabywalters
it’s also throwing away information which the consumer of the parsed version can trivially remove, but impossible to put back in
#
tantek
it's usually worked out better for whitespace to *not* be collapsed at parsing time
#
tantek
right
#
tantek
I guess let's see some real world examples that don't work in that algorithm and we'll figure it out from those
#
barnabywalters
so maybe add a whitespace collapsing stage to the “is it a note” algorithm
#
barnabywalters
tantek: I’ll see if I can find some
#
tantek
I haven't seen a need for it except in before/after space which trim() takes care of
#
barnabywalters
where whitespace is collapsed in the name but not the content, interestingly
#
barnabywalters
that might be a parsing bug
#
barnabywalters
I might be using the nodeValue when I should be using textContent or some such nonsense
#
XgF
tantek: So these days there's selfdogfooding (a bunch of people on the activitystreams list are pump.io developers, for example)
#
XgF
clients (see every pump.io client, pump.io itself...), the predefined object types/verbs aren't a problem (in fact they're useful),
KevinMarks joined the channel
#
@kevinmarks
"@waxpancake: @gary_hustwit on the long-term benefits of self-publishing your own work directly to fans. http://t.co/LezUMtpkfN" #indieweb
(twitter.com/_/status/400318854888292352)
#
KevinMarks
If you use an HTML5 parser, isn't whitespace collapsed as part of the spec? Maybe need test cases
#
barnabywalters
uses PHP DOMDocument and doesn’t trust it to follow any spec
#
barnabywalters
KevinMarks: if whitespace is collapsed by default, how does white-space: pre work?
#
KevinMarks
Good point. Clearly I need to make test cases for http://indiewebcamp.com/note#White_space
ryana joined the channel
#
tantek
KevinMarks - it's not collapsed by default
#
tantek
or rather - only collapsed in very minimal ways (e.g. before/after tags, sometimes)
#
tantek
barnabywalters - ah, line breaks
#
KevinMarks
I'm still haunted by evil white space that you can type in windows but is invalid utf8
#
tantek
ok that should be easy to deal with, let's treat a line break in the content as termination and see if it is a prefix of the name
#
tantek
oh wait
#
tantek
this looks like a parsing bug
#
tantek
because the same node text is being used in both cases
#
tantek
<div class="e-content entry-content"><div class=""><p class="p-name">
#
tantek
barnabywalters the p-name and the e-content's .value key should be identical - why are they different?
#
tantek
they should probably both preserve the internal whitespace
#
tantek
using the nodeValue
KevinMarks, benprew and paulcp joined the channel
bnvk joined the channel
#
tantek
this: " It didn’t seem worthwhile to spend a lot of time learning how to configure it correctly and create a complex theme, when what I really wanted to do was quickly manage my design and focus on writing. Enter static site generators."
cweiske joined the channel
#
tantek
XgF - have any of those supposed activitystreams selfdogfooders bothered to update the AS wiki pages accordingly? and if it's all pump.io - that sounds almost like an implementation specific technology.
scor joined the channel
#
XgF
tantek: There are other people using AS, but Pump is the only big "public" deployment
#
tantek
XgF - then why isn't it documented on the AS wiki?
#
XgF
The wiki pages are a bit tumbleweedy, mostly because what needs saying gets said in the specs
#
tantek
I guess I'm not taking your word for it in IRC
#
tantek
well - lack of documentation of implementations is a problem
#
tantek
it indicates insufficient critical mass / community IMO
#
tantek
the specs themselves are bit of a tldr mess
#
tantek
that would benefit from broken into a modular set of minimal pages on the wiki
#
XgF
The JSON serialization spec (the "core spec") is pretty simple (and can easily be grokked in a single reading). The base schema spec could do with better organization, yes
#
XgF
I'll investigate that (though I need to find out whats happening ATM
#
XgF
there was talk of the W3C wanting to adopt ActivityStreams as an official workgroup)
#
snarfed
to be fair, the google+ API's data format could arguably count as AS too
#
snarfed
but it is tweaked, and it's only the data model, not any of the API itself, and the API is extremely limited…so it's not much
#
tantek
XgF - the talk at the osfw3c meetup in August was a bit of a disaster
#
tantek
the Google schema folks were called out for bullying changes in ActivityStreams
#
tantek
and meanwhile the JSON-LD folks were pressuring the AS folks into adopting JSON-LD
#
tantek
it had all the makings of a bureaucratic trainwreck
#
tantek
snarfed- yeah - Google+ API's data model is a conceptual fork of AS
#
tantek
some reuse of ideas
#
snarfed
yup. more than just ideas
#
snarfed
most of the core properties, enum values, etc are the same
#
snarfed
they add extra stuff, but they don't mess too much with the AS base, so it seems like most code is interoperable with only minimal changes
#
snarfed
regardless, it's academic since we don't work much with AS here
#
snarfed
…i do have AS to mf2 json/html conversion code, though, so i'll be able to use that for google+. https://github.com/snarfed/activitystreams-unofficial/blob/master/microformats2.py
#
tantek
G+ has been very hard to interoperate with in any capacity
#
tantek
interesting
#
tantek
so perhaps we can subscribe to people who post on G+ via h-entry?
#
tantek.com
edited /WordPress (+1354) "add Switching away from section with examples of another project, including statis site egnerators"
(view diff)
#
XgF
tantek: Yeah, being pulled from both directions
#
tantek
XgF - it was depressing to watch :(
#
XgF
To be fair w.r.t. JSON-LD, the initial starting point for "AS 2.0" was to build it as a profile of JSON-LD
abrereton joined the channel
#
tantek
JSON-LD is a massive distraction for anyone who wants to ship/iterate/interoperate ASAP
#
tantek
because the crowd behind JSON-LD do not actually ship/iterate/interoperate on their own projects/sites with it.
#
tantek
paying attention to JSON-LD is not technical nor practical, only political
amz3 joined the channel
#
XgF
Theres certainly /utility/ there. You could within about ~25 lines build a JSON-LD context which would map ActivityStreams JSON to "ActivityStreams RDF"... but yeah, I want to avoid the Schema.org quagmire
#
XgF
schema.org is an unmitigated mess
#
tantek
indeed, that we can agree on
#
XgF
[FWIW, I quite like RDF as a data model. It's certainly easier to work with as a "database storage format" than JSON in some ways)
fmarier joined the channel
#
tantek
from original poor design to grab bag additions - schema is bloating up as a maldesigned mess
#
tantek
XgF - have you met tommorris?
#
XgF
I don't think so
#
tantek
he too has some appreciation for RDF
#
tantek
FWIW, mapping to "ActivityStreams RDF" is not really "utility". that's just mapping between things that have no innate utility themselves
#
tantek
interesting about your perspective/experience that " RDF as a data model. It's certainly easier to work with as a "database storage format" than JSON in some ways"
#
tantek
that's the exception in my experience
#
XgF
RDF (or other "object oriented" style) databases become easier the minute you start having cyclic graphs :-)
#
tantek
so either you've figured out something clever about RDF, or are making JSON harder than it should be
#
aaronpk
this is basically an argument against pure JSON as a data store http://www.sarahmei.com/blog/2013/11/11/why-you-should-never-use-mongodb/ Possibly related?
bnvk joined the channel
#
XgF
aaronpk: Pretty much
#
aaronpk
XgF: cool, do you know of a simple example like that of using RDF as a data store?
#
XgF
Not sure if there are any deployed projects out there. I was having a bit of a play with Python + Pyramid + RDFLib + a custom "ORM".
#
XgF
From my experience with ZODB, it's pretty much as nice to use (but matches the data structure better if all your data is identified by URI)
jacook joined the channel
#
amz3
XgF what do you mean by “become easier the minute you start having cyclic graphs”
#
XgF
Using ActivityStreams as an example, a comment is inReplyTo its' parent post, and the comment should be in parentPost.replies
#
tantek
aaronpk - that seems more like an argument against denormalization rather than JSON
#
aaronpk
kind of
#
tantek
using URLs for references to people, posts etc. seems to get around that problem
#
XgF
Document stores tend not to be ACID, so don't do normalization too well
#
tantek
sticking all the data as strings inline in every instance seems to be the problem
#
amz3
XgF I know what cyclic graphs, I don't understand the first part of the statement «RDF database become easier», why do you mean by easier? Since they handle ACID acrod entities, it makes things easier?
#
amz3
s/why/what
#
Loqi
amz3 meant to say: XgF I know what cyclic graphs, I don't understand the first part of the statement «RDF database become easier», what do you mean by easier? Since they handle ACID acrod entities, it makes things easier?
#
amz3
s/acrod/across
#
tantek
btw XgF - go ahead and add yourself to http://indiewebcamp.com/irc-people
#
XgF
RDF becomes easier to work with than JSON because your ORM can do the lookups for you
#
XgF
tantek: Will do as soon as I unbreak my rel=me auth :-)
#
tantek
XgF - unbreaking your rel=me auth is easier than any RDF :)
#
amz3
I never understood the value that brings a RDF datastore, except the RDF query languages that I never studied deeply but seemed very cryptic and difficult to learn, not very scientific, yet I prefer graph databases
paulcp joined the channel
#
XgF
tantek: I finally got sick of Wordpress and MySQL, so my website is now (provisionally) migrated to idno. I haven't yet figured out what benwerd did to get his profile on /...
#
XgF
amz3: The single biggest benefit (IMO) is the distinction between strings and links
#
XgF
I half suspect it's just something trivial like "rewrite / /profile/name/;"
paulcp, vrypan, andreypopp, paulcp_, josephboyle, tpinto and benprew_ joined the channel
#
owenshepherd.net
created /User:Owenshepherd.net (+179) "Created page with "[http://owenshepherd.net OwenShepherd.net]"
(view diff)
#
Loqi
powered by [[idno]]. [[POSSE]] to [[Twitter]]
#
Loqi
hopefully soon to [[pump.io]], looking towards implementing full integration given time."
barnabywalters, bnvk, paulcp and vrypan joined the channel
abrereton, smus and caseorganic joined the channel
bnvk and tilgovi joined the channel
#
bret
aaronpk: interesting article about mongo
tantek joined the channel
#
bret
also RE osfw3c, that was quite a scene to watch unfold. I'll attest to the accuracy of tantek's description
#
bret
++tantek
#
bret
aaronpk: are you going to be at http://calagator.org/events/1250465193 ?
#
Loqi
MaptimePDX Hack Night on Wednesday, Nov 13, 6:00pm at Esri R&D Center (aka Geoloqi HQ)
#
tantek
bret, bullies really don't like to be called out as such.
#
bret
also ozten i like your noble web ideas! that was a good read too
#
bret
ozten++
#
Loqi
ozten has 2 karma
#
bret
and tantek++ fixing the above attempt
#
bret
ill get it later
#
aaronpk
tantek: oh cool, she was at the QS meetup when I wrote that article, we talked for a bit!
#
tantek
aaronpk - great!
#
bret
I'm starting to eye http://docpad.org/ as a next step for my site, but part of me wants to do it from scratch to avoid not fully understanding someone elses project
#
aaronpk
lots of people interested in this comment thread too https://news.ycombinator.com/item?id=6718422
#
ozten
bret: right on
caseorganic joined the channel