Timestamps are in UTC.
reload again...
marking up my resume that i mail to people with 30 wikipedia links just feels like cruft to me
Yoz and I are going to add microformats to Ning
pnhChris: I think it makes a lot of sense to have skills be links to something describing what it is you have skill at
hypertext or text, that is the question
I'm currently marking up the chronologically-earliest mention of $TECH as an hresume skill
linking to a singular definition or a tag space?
and in some cases i may already be using abbr to add more information
it just seems messy
and arbitrary
It feels very webby to me. I'd much rather have too many things linked up than too few.
there were times when I created a link I'd rather not have had
but had to for the sake of adding the skill
it doesn't feel like something i want to do on a resume
I'd rather not have people who want to read my CV following useless links
I guess I could make them appear as not links with CSS
but that seems a bit crap
i often hesitate to even link projects or past employers because of implications
"in 1999 i built a website for company XYZ: see!"
i just don't see value in adding arbitrary links to tell someone looking at my resume what HTML is
I agree
not to mention being compelled to have to closely monitor that tag space
ultimately, if i don't already feel that someone visiting the web page needs that additional information, why would i think someone on the other side of a consuming application need it
a link to a wikipedia article isn't going to help monster.com's parser digest and categorize my resume
if anything else it might read too much into those links
see some association there
or a non-mf aware parser reading some association there
kingryan is ryan king
[[hreview-faq]] http://microformats.org/wiki?title=hreview-faq&diff=0&oldid=5038 * Tantek * (+230) how to markup multiple reviews on one item on the same page.
http://technorati.com/faves
hey Kevin, are we live with that?
yes
as of 15 mins ago
hey Kevin, Ning also has ?tag=tagname style tag URLs
bleah
can they not change?
I coudl be Postelian about it in the spider (though ?tag=6747478 cruft I can't)
check out http://hreviewit.ning.com
Kevin, I've asked Yoz if they can do /index.php/tags/tagname
and he's looking into it
great!
mlinksva is Mike Linksvayer and from Creative Commons
hmm
is it possible to register at microformats.org?
sigh someone else will have to fix the link then.
http://microformats.org/wiki/xfolk-profile
has a link to xoxo profile
which has now moved to http://microformats.org/wiki/xoxo
it is still linking to technorati which may not be the current doc for xoxo.
[[recipe-examples]] http://microformats.org/wiki?title=recipe-examples&diff=0&oldid=5039 * HollyMarieKoltz * (+946)
tantek is Tantek <http://tantek.com> and works on Technorati and develops microformats <http://microformats.org>
[[recipe-examples]] http://microformats.org/wiki?title=recipe-examples&diff=0&oldid=5040 * HollyMarieKoltz * (+74)
[[recipe-examples]] http://microformats.org/wiki?title=recipe-examples&diff=0&oldid=5041 * HollyMarieKoltz * (-2) Moroccan Braised Beef -
[[recipe-examples]] http://microformats.org/wiki?title=recipe-examples&diff=0&oldid=5042 * HollyMarieKoltz * (-7) Next Steps -
[[hreview]] M http://microformats.org/wiki?title=hreview&diff=0&oldid=5043 * Tantek * (+82) Schema -
[[recipe-examples]] http://microformats.org/wiki?title=recipe-examples&diff=0&oldid=5044 * HollyMarieKoltz * (+0) Next Steps -
tantek is Tantek <http://tantek.com> and works on Technorati and develops microformats <http://microformats.org>
[[hreview-faq]] M http://microformats.org/wiki?title=hreview-faq&diff=0&oldid=5045 * Tantek * (+114)
i wasn't able to watch the whole feed today and missed the hatom stuff... anyone catch what was settled on (with content + author issues)
[[hresume-feedback]] http://microformats.org/wiki?title=hresume-feedback&diff=0&oldid=5046 * Dave Cardwell * (+463) General Comments - More feedback - skills.
tantek: you here?
bergie is lives in Finland and blogs at http://bergie.iki.fi/blog/ and Midgard CMS developer
blake is allegedly human. Blake, also known as Cortland M. Setlow, studies at swarthmore.edu and enjoys building things, exploring buildings, and physics. He currently sleeps during the day.
limbo_ is Eran and blogs at http://hellonline.com/blog/
[[recipe-examples]] http://microformats.org/wiki?title=recipe-examples&diff=0&oldid=5047 * HollyMarieKoltz * (+1196)
[[recipe-examples]] http://microformats.org/wiki?title=recipe-examples&diff=0&oldid=5048 * HollyMarieKoltz * (+532)
bergie is lives in Finland and blogs at http://bergie.iki.fi/blog/ and Midgard CMS developer
[[recipe-examples]] M http://microformats.org/wiki?title=recipe-examples&diff=0&oldid=5049 * HollyMarieKoltz * (+1) Cranberry Cherry Relish -
[[recipe-examples]] M http://microformats.org/wiki?title=recipe-examples&diff=0&oldid=5050 * HollyMarieKoltz * (+1) Next Steps -
bkdelong is B.K. DeLong, Head Research Analyst for HALO Worldwide - http://www.haloworldwide.com. Web: http://www.brain-stream.com. Email: bkdelong@pobox.com and lives in Salem, MA, USA (-5:00 GMT)
dglazkov is Dimitri Glazkov (http://glazkov.com) and lives in Birmingham, AL, USA (-6:00 GMT)
pnhChris is Chris Casciano, blogs at http://placenamehere.com/ , and a member of the Web Standards Project.
mlinksva is Mike Linksvayer and from Creative Commons
tantek is Tantek <http://tantek.com> and works on Technorati and develops microformats <http://microformats.org>
mlinksva is Mike Linksvayer and from Creative Commons
hober is Edward O'Connor and works for EVDB on http://eventful.com/ and lives in San Diego, CA (-08:00)
limbo_ is Eran and blogs at http://hellonline.com/blog/
amanuel is Amanuel, the social ambassador at http://otavo.com
microformats.org refused to let me register is registration disabled?
was your proposed username a WikiName?
amaneul, see FAQ
question #1
http://microformats.org/wiki/faq
I think so.
will read more carefully thanks
thanks tantek
I was trying Amanuel
AmanuelTewolde
is it XOXO or xoxo
depends on whether it's HTML or XHTML ;)
good morning everyone
the technology is XOXO
the class name is "xoxo"
we had some excellent discussions yesterday
yeah I left in the middle of it
not by choice.
I left by choice, but not by desire
[[xfolk-profile]] M http://microformats.org/wiki?title=xfolk-profile&diff=0&oldid=5051 * AmanuelTewolde * (-82) update link to the local "xoxo profile" instead of technorati's older copy.
thanks amanuel
np tantek glad to help
dglazkov, one of the things we talked about was the object "include" technique that has been proposed (and accepted) for hResume
http://microformats.org/wiki/resume-brainstorming#hCard_reference_via_object
I believe this is useful not only for hResume, but given experience with hReview (i.e. helping Ning add hReview support yesterday), for hReview as well
talk of xanadu brought back some warm memories
sad we still think of web as webpages
In addition to object includes for hResume and hReview, for a quite some time, a bunch of us have talked about adding the ability to do includes/imports to XMDP
but for that, the latest thinking was to use <a href="XMDPtoincludeURL" rel="import">
so given that this object include technique makes sense across multiple microforrmats
and we have discussed using <a href rel> for includes for XMDP
this feels like a building block microformat
I'm not sure it stands on its own as a microformat
so perhaps it is something like the abbr-datetime-pattern
include-pattern
and then document *both* the <object class="include" href="#"> method for local content includes for microformats, and the <a href="remoteprofile" rel="include"> method for including XMDP profiles (and portions thereof) into XMDP
OR
the include-pattern could *only* describe <object class="include" href="#"> technique for now, since that is the *one* technique which has demonstrated direct utility across multiple microformats (hResume, hReview, and potentially hListing as well)
and then simply add the <a rel="include"> feature to XMDP directly
what do folks think of that?
and while you're thinking of that...
one more question for the channel to ponder. We've had the fn->nickname shortcut proposed for hCard for some time, and there is a bunch of support
http://microformats.org/wiki/hcard-brainstorming#Implied_.22FN_and_N.22_Optimization_.28proposal.29
i want to include this in hCard today
any objections
?
tantek: re: XMDP include, I'm all for it, as you can probably gather from my mailing list post the other day
I'm wondering about the degenerate case: when an XMDP profile only has includes
<dl class="profile"> ... what goes here ... </dl>
maybe <dt>include</dt><dd><ul><li><a rel="include".......
re: fn->nickname shortcut: sounds good to me
[[events/xtech-2006]] N http://microformats.org/wiki/events/xtech-2006 * RyanKing * (+64) creating stub
kingryan is ryan king
hober, yes, I figured ;)
blake is allegedly human. Blake, also known as Cortland M. Setlow, studies at swarthmore.edu and enjoys building things, exploring buildings, and physics. He currently sleeps during the day.
[[events/2006-03-07-etech-microformats]] http://microformats.org/wiki?title=events/2006-03-07-etech-microformats&diff=0&oldid=5052 * Tantek * (+288) added details to top
Given a) the divergence between the W3C and WHAT-WG on head@profile, and b) the mf preference for visible metadata, I really really like the idea of head@profile="#profile", and then <dl id="profile" class="profile"/> which consists of the includes for the various xmdp profiles being used.
It's a nice way of being visible about it while also being conservative WRT future direction of HTML.
hober, don't worry too much about W3C/WHATWG divergence
i'm working on getting the proper feedback to the right folks at W3C
word
for now, let's move forward with the assumption that they're not going to break backward compat here
because doing so (breaking backward compat) would be idiotic
So far as I can tell, they're all for breaking backwards compatibility (think XHTML 2)
trovster is a web developer from the UK who writes on http://www.trovster.com and runs www.csslounge.co.uk
[[hcard-brainstorming]] http://microformats.org/wiki?title=hcard-brainstorming&diff=0&oldid=5053 * DimitriGlazkov * (+44) Implied "FN and N" Optimization (proposal) -
[[hresume-feedback]] http://microformats.org/wiki?title=hresume-feedback&diff=0&oldid=5054 * Dave Cardwell * (+168) General Comments - Forgot to add something.
tantek, imho include-pattern should definitely contain <object class="include" href="#">
what's the diff: rel="import" , rel="include"? Typo?
dglazkov, not a typo, and no difference really, hence I am working on converging
we've talked about rel="import" for a while for XMDP
why not rel="xmdp"?
while <object class="include"> is new (class name) as of yesterday
dglazkov, that's not precisely correct
you're not including the xmdp to *use* it in another xmdp
you're including the xmdp to have it count as part of another xmdp
thus it is an "import" or an "include"
oh, got it
rather than a rel="profile"
which means to use it as a profile for the current document
[[hcard]] http://microformats.org/wiki?title=hcard&diff=0&oldid=5055 * Tantek * (+754) added implied "nickname" optimization, cleaned up optimizations
I think i'd go with making rel="include" part of XMDP
dglazkov, yes, that is how I am leaning as well
I think of rel="include" as an atomic mf, which could/should be used by XMDP (and other things)
programmers already think of "includes" as nouns
a ".h" file is an "include"
thus "include" makes sense as a noun for a rel-value
except xmdp is so clean and nice right now
it describes the destination as a noun, in terms of its relationship to the referring document
rel="include" will allow XMDP composition
which will make it easier for authors to refer to a single XMDP profile
which can then compose all the XMDP profiles they are using
kind of like a shared style sheet
hober, give example of other things?
dglazkov: cross-document mf composition? (the cross-document equivalent of the <object/> transclusion hack)
bergie is lives in Finland and blogs at http://bergie.iki.fi/blog/ and Midgard CMS developer
hober, in practice, the only microformat content includees we have found are more amenable to using <object>, from a semantic point of view
alright, I've updated hCard to included the implied "nickname" optimization for one word "fn"s (as mfbot reported above)
please take a look and review and let me know if anything is confusing or if there are any ambiguities or errors
http://microformats.org/wiki/hcard#Organization_Contact_Info
read from that section down through the next four sections
as I've tidied them up a whole bunch
looks shiny. There is a period after whitespace in parens. Maybe no parens at all?
[[hcard-brainstorming]] M http://microformats.org/wiki?title=hcard-brainstorming&diff=0&oldid=5056 * Tantek * (+271) added contributors, moved implied "nickname" to hCard proper
[[hcard-brainstorming]] M http://microformats.org/wiki?title=hcard-brainstorming&diff=0&oldid=5057 * Tantek * (+3)
dglazkov, where do you see that?
is exactly two words (separated by whitespace.),
on "Implied "n" Optimization"
[[hreview]] M http://microformats.org/wiki?title=hreview&diff=0&oldid=5058 * KimBayne * (-7) Examples in the wild -
[[introduction]] http://microformats.org/wiki?title=introduction&diff=0&oldid=5059 * Tantek * (+251) added "SIDE" article
[[hcard]] M http://microformats.org/wiki?title=hcard&diff=0&oldid=5060 * Tantek * (+156) New Examples -
[[introduction]] M http://microformats.org/wiki?title=introduction&diff=0&oldid=5061 * Tantek * (+110) See Also -
[[hcard]] M http://microformats.org/wiki?title=hcard&diff=0&oldid=5062 * Tantek * (-1) Implied "n" Optimization -
thanks dglazkov
found it
Enric is a media Software Developer and Videoblogger located at http://www.cirne.com
bkdelong is B.K. DeLong, Head Research Analyst for HALO Worldwide - http://www.haloworldwide.com. Web: http://www.brain-stream.com. Email: bkdelong@pobox.com and lives in Salem, MA, USA (-5:00 GMT)
so, did we see a hatom content resolution yesterday?
chris, yes
i believe kingryan is writing up the resolutions
to hAtom-issues
and davidjanes will be updating the hAtom spec
in short:
entry-title, entry-summary, entry-content
all as a set
gotcha
since Atom essentially introduces very specific semantics for all three of those, which are not necessarily portable/reusable as "generic" terms in the microformats vocabulary
thus we've swallowed the bitter pill of prefixing them in order to make it easier for folks working with both Atom and hAtom
how bout the author... do you recall if <p class="vcard author"> outside of the hfeed element get picked up
chris, yes, that is part of the resolved issues as well
http://www.scripting.com/2006/02/22.html#whyFormatsLikeRss20Work
[[include-pattern]] N http://microformats.org/wiki/include-pattern * Tantek * (+2719) first draft
ok, documented object include pattern
[[hatom-issues]] http://microformats.org/wiki?title=hatom-issues&diff=0&oldid=5063 * RyanKing * (+37) Feed (atom:feed) - note about hfeed not being required
factoryjoe, that rss2 post seems like a long-winded way of saying, in effect no-one may change/evolve/iterate rss and still call it rss, which effectively means rss2 itself is frozen
[[hatom-issues]] http://microformats.org/wiki?title=hatom-issues&diff=0&oldid=5064 * RyanKing * (+41) Entry Title (atom:title) - note about entry-title
yeah
basicall
y
[[Main Page]] M http://microformats.org/wiki?title=Main_Page&diff=0&oldid=5065 * Tantek * (+36)
i don't know where the post came from (ie context?)
we believe in evolution, not stagnation
and on that note,
http://microformats.org/wiki/include-pattern
please take a look folks, especially those of you who have been participating in the hResume, hReview, hAtom discussions in recent days
whoa!! cool!!!
I would like to *normatively* reference this from the next (in progress) versions of hResume, hReview, and hAtom currently being written
tantek: what does normatively mean to humans?
[[include-pattern]] M http://microformats.org/wiki?title=include-pattern&diff=0&oldid=5066 * Tantek * (-30)
factoryjoe, normatively means what's "officially" in the standard as opposed to a tutorial for example
definition vs. explanation
ok, thought so
definitions are normative, and what you code to
got it
explanations and examples are *informative* and are guidelines
if the two contradict, normative beats informative
cool. you can now consider yourself a standards geek.
is it possible to normativelly reference rss 2.0?
yay!
limbo__: i think that would be abnormatively
ah
makes sense
;)
the rss2 post from Winer comes from the rss advisory board reactivation
that was a nice story
made for quite a bit of noise
let's not get too distracted by others' noise folks
let's focus on our own success
indeed
tantek: i might have some help w/ mFs in Flock
yay!
and dammit, we gotta get you to meet with the engineers
assaf and amanuel might be able to help
http://feedvalidator.org/check.cgi?url=http%3A//feedvalidator.org/testcases/rss20/element-channel-item/multi-enclosure-test.xml
"Undecipherable Specification Error" ?
http://feedvalidator.org/docs/warning/UndecipherableSpecificationError.html
awesome
wow
this is why we love markp
i think someone should maybe blog about that
sam did: http://www.intertwingly.net/blog/2006/02/22/Undecipherable-Specification-Error-Redux
i didn't do that!
it's sam
thanks kevin
ah, that's why like sam, then
he got fed up with the discussion on rss-public and wanted to force the issue
markp: would you be interesting in helping w/ a flock mf topbar?
it seemed lik somthing you'd do, mark
actually, i just tried flock for the first time yesterday
i doubt it'll work this time. winer is much harder to move than apple.
i'm unclear why they forked an entire browser when it seems like they could have accomplished the same thing with a few firefox extensions
but i'm sure that's a FAQ somewhere
markp: that's the #1 question
it'll make sense by 1.0
and the answer is?
ah
they == factoryjoe
for right now it does seem like we're just firefox + extensions and a pretty theme
but firefox will unlikely ship lucene
they.include?(factoryjoe) == true
and probably won't include "people in the browser"
nor a blogging client
tantek, include-pattern looks good
kingryan, good, now onto hReview 0.3 for me
lucene the java-based asl2-licensed search engine?
(now that I've written up the pre-requisites)
markp: besides, a little competition doesn't hurt anything
markp: i think we're using cLucene right now
markp: people in the browser mockup: http://flickr.com/photos/25419820@N00/102993817
ok, asl2/gpl/lgpl
is incompatible with mpl/gpl/lgpl
hrm
so no, firefox would never ship it
will have to pass that by bart
still don't understand why an extension couldn't ship it
but to each his own (browser)
exactly our point
;)
wow, all your friends are creepy white nerds with glasses
it's a network!
CNWGN
ryan: that can be arranged
CNWGN == technorati staff!
cnwgn.net is free
it's like Brat Pack 2.0, except with more glasses
and nerds
[[hatom-issues]] http://microformats.org/wiki?title=hatom-issues&diff=0&oldid=5067 * RyanKing * (+51) Entry Content (atom:content) -
what's the license of flock?
we have a whole page of them
can your code be reincorporated into firefox at a later date?
http://technorati.com/about/staff.html
[[hatom-issues]] http://microformats.org/wiki?title=hatom-issues&diff=0&oldid=5068 * RyanKing * (+55) Entry Summary (atom:summary) -
ah: http://wiki.flock.com/index.php?title=Licensing_FAQ
no
your code is straight gpl
[[hatom-issues]] http://microformats.org/wiki?title=hatom-issues&diff=0&oldid=5069 * RyanKing * (+37) Entry Updated (atom:updated) -
so... i would think that you could make an extension w/o the licensing issues?
(i'm not really a big license expert (see my rants against the RIAA, et al))
Enric is a media Software Developer and Videoblogger located at http://www.cirne.com
[[hatom-issues]] http://microformats.org/wiki?title=hatom-issues&diff=0&oldid=5070 * RyanKing * (+87) hi markp
grr
my point is that your code can never be reincorporated into firefox
too bad you can't ignore mfbot
hmm
without relicensing under the marvelous tri-license (cough)
yeah
i guess that goes both ways or something
at least the code isn't CC
;)
lol
anyway i mean, extensions can live in the middle i think
no, in fact, it does not go both ways
oh
ah
hrm
firefox code is tri-licensed, and it is up to the receiver to choose which license they want to use it under
i c
i have, personally, taken code from firefox and used it in a greasemonkey script under the gpl
i didn't worry about it much, since the GM script was not something that would reasonably be considered for re-inclusion into firefox
[[hreview]] M http://microformats.org/wiki?title=hreview&diff=0&oldid=5071 * RyanKing * (+9) proper semantics of 'class name'
markp, try /ignore mfbot
in your case, you are doing things that someone might very well want to roll back into the main firefox codebase
hmm
but they can't, because of licensing issues
i guess i was under the impression that it *could* go both ways
well it can't
you can take whatever you want from firefox
i admit that i'm plainly naive about these things
but if you decide to take it under only one of the licenses firefox offers, then it can't be reincorporated into the main code base
what about camino? did they take the other path?
it's not a gpl-specific thing, the same problem would occur if you licensed your code solely under the mpl, or lgpl
camino is also tri-licensed
see bottom of http://www.caminobrowser.org/development/programming/
anyway, you may or may not care that you're forking firefox with no hope of future reconciliation
but you should at least be aware that that's what you're doing
limbo_ is Eran and blogs at http://hellonline.com/blog/
markp: i just asked bart
and he said you're right
that's vaguely comforting
i suppose
but under certain circumstances, we can choose to license portions of our code under the trilicense
so for example
if you contributed something for parsing microformats for flock
and your condition was that it be trilicensed
that's something that we could discuss
[[hresume]] http://microformats.org/wiki?title=hresume&diff=0&oldid=5072 * RyanKing * (-178) removing 'reference' and making root class name explicit
actually i've already written code for parsing some microformats
and it's gpl-only :)
but that illustrates my point
if you take that code and integrate it into flock (and you are free to do so), that would just make it harder to contribute code "upstream" to firefox
since you would need each copyright holder's permission to relicense their bits
right
ok
i'm somewhat familiar with that bit
anyway, my hcard parser is here: http://diveintomark.org/projects/greasemonkey/hcard/
have you seen tails?
i ported that to python and put it in feedparser.py too
but you probably wanted the javascript/dom version
tails? no
RobertBachmann is Robert Bachmann <http://rbach.priv.at/> and lives in Austria (Timezone: 01:00)
markp, could you add your hcard parser to the Implementations section on the hCard spec?
kingryan, thanks for posting notes from yesterday's meeting
i could, but so could anyone else
heh
markp: http://blog.codeeg.com/tails-firefox-extension
[[hresume]] http://microformats.org/wiki?title=hresume&diff=0&oldid=5073 * RyanKing * (+78) Experience - added 'include' class name
factoryjoe: doesn't load, i blame the corporate firewall
aha
oh yeah, i've seen that
yeah
kinda useless, no?
it's a good start
yes
i suppose that would be the more diplomatic way of describing it
but better integrated.... like, in an "mf vacuum topbar" we could show all the mf items on a page
and suck them up as you browser
browse
then you could right click and do various things... "post to my yahoo calendar", "send email", "bookmark this person's urls"
what i'd really like -- and this is something i've been prototyping in greasemonkey -- is a way to suck up all that information and post it to my private atomstore
etc
yes
well
what does your atom store look like?
so i can do real searches on it
coz that's what i want to do w/ lucene
lucene should be indexing all the mfs
since we index your feeds and history
my atomstore is a cgi script on my own remote server that understands the atom api
then when you do a search after fact, we reveal the data in the pages
ah
well you could run the server locally and publish via flock
this way i'm not tied to browsing my history on a single server
er, client
ok
indeed
the theory is that i would install the same script on my work and home machines
and have a unified searchable/browseable history
exactly
but it's entirely possible that it's not worth involving a separate server
yeah, that's what i'd like to do w/ flock profile syncing
so you could login w/ any browser
yeah, that might be easier
but if you're using flock, you get big UI benefits
do everything locally, sync it up once in a while
markp, hey dude
hello
markp, we (flock) aren't triple licensing our code - just releasing gpl, basically
markp, so we don't have a licensing problem with clucene
sync globally, act locally?
yes, i am aware
yes, i agree
markp, it does make it hard to send it upstream though :(
ok, i might have misunderstood
yeah that's what mark was saying
yes, that's my only point
if you don't care about sending code upstream, then happy trails
ok cool
well, I'd like to send code upstream, but we probably can't send this component upstream
i got the impression from factoryjoe that y'all thought it would be easy to send code upstream
we were talking about markp and his work on microformats making it into flock...
and i was just pointing out that that was not, in fact, true
there is actually currently a problem that gerv has identified with us offering a single download, but that is not clucene specific
and trilicensing it so that it can go upstream
ok cool
cool
I'd like flock to be a place people can experiment with things that will make there way into the wider mozilla development community
(tangent: this is one of the very nuanced reasons to prefer straight-gpl-licensed code, because it guarantees that there are no legal impediments to merging forks)
yakk: i'm still not clear why that can't be done in extensions
markp, the scope of extensions is limited
markp or BSD, MIT, etc
we're also concerned about overall ease of use
markp, why wasn't firefox2's places implemented as an extension?
:)
offering 20 separate extensions would be hard to maintain consistency in
foolswisdom: no, that is exactly my point, i can fork a bsd-licensed project and go completely proprietary with it, and the forks can never be merged
whereas just doing the whole browser guarentees certain baseline functionality in our product
pure GPL is nice, but both Firefox and Flock have a trademark protection requirement that can't be satisfied :(
forks of straight-gpl programs must themselves be straight-gpl, thus ensuring license compatibility
thats why Firefox ships under essentially a proprietary license
i know it's not really relevant to the firefox discussion, that's why i marked it "tangent"
yes, debian-legal goes apeshit every now and then about the firefox license
markp: now, I see what point you were making ;-)
so speaking of
markp, and its one of the more legitimate debian-legal freak-outs
markp, we've been discussing OPML and FOAF @ flock lately
oh god
i've been pushing the microformats angle
i was wondering if you and yakk might have some thoughts to share
on why one might be better than the other?
you are primarily a consumer of content
if there is content in the world that is only available in opml, then you should support opml
ditto foaf
BUT
we want to do both
when you are acting as a producer of content (e.g. exporting information), PLEASE DON'T SUPPORT OPML
oh, I hate opml
for the love of god
the last thing the world needs is more OPML
bergie is lives in Finland and blogs at http://bergie.iki.fi/blog/ and Midgard CMS developer
one of our developers got excited about OPML recently and I've been trying to calm him down.
oh, there's always one
a new one every month, but there's always one
show him xoxo
cults are like that
hopefully he'll see the light.
we're trying to work out what kinds of formats to support for social network / profile / etc data, and there doesn't seem to be a microformat thats appropriate, while FOAF seems to be widely supported, mature and extensible
haha
sorry, that's very funny
FOAF is many things, but it is not stable
xoxo and opml both fail to solve the same problem
markp, in what ways is it unstable?
dan periodically spiders all the FOAF files in the world and looks for elements that people just made up
sorts them by prevelance and adds the most popular elements to the spec
well, its RDF, so people are allowed to do that, right?
oh
sure!
hmm
what could possibly go wrong?
thats kinda scary
"missing isn't broken", etc.
feh
i've not seen much support for consuming FOAF
most YASNs that support FOAF only do that for publishing
there are large social networks that publish FOAF though
very few will import it
whenever someone mentions the word "semantics", i reach for my gun
so it's the same as OPML - consume FOAF, produce XFN.
we're interested in consuming on the client
yakk: can you clarify that aside on xoxo?
and publishing presumably with blogs
limbo_, except XFN represents 10% of the information in FOAF
yakk, that's a *feature*
KevinMarks, its a blogroll format which isn't the problem we're trying to solve
lol
you know when you're *done* implementing XFN
whereas you are *never* done implementing FOAF/RDF
oh, just make up your own xml format and be done with it
step 1: admit you have a problem
step 2: state it
tantek, okay, but what if I want to represent a bunch of information about a person, not just their blogroll or just their vcard?
XOXO isn't just a blogroll format
tantek: you mean http://www.tbray.org/ongoing/When/200x/2006/02/07/Thumbnail ?
yakk, the bits in FOAF about a person are a needlessly reinvented mutation of vcard
yakk, there are 100000000x more vcards/producers/consumers out there than FOAF
yeah, why not export .vcf files?
I use XOXO as a generic serialization format
and thus hCard is based on vCard
i believe that XFN + hCard would give you everything you'd wanna use FOAF for (that makes sense)
oh wait, i forgot what channel i'm in
tantek, but vcard is old old old.
and thus with hCard, you actually get a lot more intereoperabtiliy *today* with apps/services etc. than FAOF
yakk, but it *works* and is *supported* by a HUGE eco-system
yakk: yes, that's the real problem, standards rot over time... </snark>
yakk, old = stable
or did you have a specific bug report?
TCP is pretty old too
tantek, well, the bigger problem is that it doesn't represent all the kind of personal information that I want to represent. Instant messenger ids springs to mind.
yakk, see hcard-example page
that's a solved problem
hcard-examples
yakk: i would start looking for examples that aren't supported
and then we can address them
i think you'd actually be hard pressed to find things that don't fit
and if they don't, maybe they're aren't generally useful
in which case shipping them in a browser is questionable
tantek, that suggests: <a class="email" href="mailto:joebob@hotmail.com">joebob</a> to represent an msn address
tantek, how on earth am I supposed to consume that?
"old old old" = 8 years?
you know the semantic that a "hotmail.com" domain address works as an MSN IM
done
tantek, but my msn id is ian@mckellar.org
ryan, in internet terms that's ancient.
tantek, its also my email address
my msn id is ryansking@mac.com
heh
doh!
"ancient" = "it now works, so lets disregard it"
kingryan: "ancient" = not-fashionable and glitzy enough
no, ancient means might need some work to comply with changes in the environment
I like Linda Stone's formulation: "obsolete enough to be useful"
"ancient" = we should just go reinvent it with new terminology
there's no mechanism for extending vcard?
URL
they never thought that someday someone might want to add fields to it?
vcard doesn't seem to have an extension mechanism and represents a model of profile information that is outdated
the mechanism is called the IETF
yakk, it's not outdated, it's interoperably exchanged by people all the time
I'd love for someone to point me to something that disagrees with that
via email clients
tantek: have you written up your point about the separation between xfn and hcard that's so essential against FOAF?
cell phones
etc.
factoryjoe, fundamentals of modularity
k
tantek, and its a real pain (at least it has been for me) to map the vcard data model to more modern applications
yakk, then post your pain points
specifics, please
I'm convinced we can resolve nearly all of them
yeah, that's really what we need to do
tantek, lets talk quickly about instant messenger
the real problem with MSN IM is that they failed to make up their own URL scheme, the way that AIM and YIM did
b/c inventing a new standard is an exercise in futility at this point
tantek, the real problem with AIM & YIM is that they made up their own URL schemes :)
you will experience thta pain when you try to import or export it anyway
yakk, are you praising Microsoft for not doing so?
;)
thats more legit than most - they are separate protocols
unlike, say, feed://
there is also skype: for example
right
tantek, url schemes have in recent years been invented as a tool to launch helper apps.
tantek, not to represent different kinds of ways of accessing resources
yes, because of the MIME-type/ extension fuckup
feed: makes me angry angry angry all the time
but IM networks *are* different kinds of ways of accessing resources
it makes me want to add html: and jpeg: support to flock
yakk, not as bad as the .tel TLD
for publishing "contact information"
you know about that right?
well, feed is arguably a differnt protocol, as it implies 'poll this at intervals'
tantek, true :)
actually, now that I think about it should the semantic of an IM address be captured with rel?
tantek, well, we could talk about how microsoft could have invented an url scheme, but that doesn't help us represent peoples' msn contact information in hcard
my concern is that because hcard aims to just be a reflection of vcard into the microformats world its limited in the same ways vcard is
tantek: i had sort of proposed that before
rel=im
really? sorry if i missed that factoryjoe
factoryjoe solves the problem
yeah actually i think i mocked that up
though because the networks are different
awhile back
we may need different values for each
same for photos, thought that's not quite as necessary
<a class="email" rel="msnim" href="mailto:joebob@hotmail.com">joebob</a>
yeah, that's decent
and if you use the same nic, just use multiple rels
rel="msnim aolim"
and that could be a microformat both independent of hCard, and usable inside hCard
see, if we just use FOAF we can spider to find these nececary extensions instead of just arguing :)
though no doubt it will be renamed 'liveim' shortly
;)
yakk: did you see the google study?
well, while we're arguing, what's the problem XOXO isn't solving?
they kind of did that
factoryjoe, no, but I blame wordpress
KevinMarks, I want to represent profile and social network information. xoxo only solves part of that.
yakk, xoxo is just for lists
XOXO has 2 uses
and outlines
Its not clear to me (I guess I should go read more) how to tie a variety of information together
lists of otehr things
lists != outlines?
and serialization
yakk, you don't one big massive monolithic format that is never finished and completely unimplementable
lists are degenreate outlines
you want a bunch of small modular, well defined microformats which can be implemented as needed
how do I tie a xoxo describing the sites I post to to an hCard that describes my contact info an XFN information about my friends?
XOXO supports arbitrary key:value pairs, but only makes sense where you want botht he keys and values human readable
yakk, take a look at the markup of my blogroll: http://tantek.com/log/
it uses XOXO+hCard+XFN
to describe my relationships to various people I know and their URLs
tantek, how do I know this is your blogroll?
tantek, that's the main source of my confusion
for XFN parsing, you can treat any links with the XFN rel values on the page accordingly
that's the point
you don't need to know that this is my blogroll
tantek, if I'm visiting your page and I want extract information about you and about your social network I can't?
yakk: address+hcard says it's tantek's page. so an xoxo+hcard+xfn on that page is his blogroll.
hober, ahh, cool, ok
well xfn is the social network
it might be a met list in a post
hober, where is that specified?
yakk, see http://gmpg.org/xfn/
and linked pages
it describes use of XFN on pages
and how XFN within blog posts indicates a time-stamp on a relationship etc.
tantek, (if you have atom I guess)
s/atom/hatom/
or atom
html inside atom
yakk: the whole ideal of using xfn rel values is that those links are relative to the current page
so if we have <Address><hcard /></address>
we know who's page it is
yakk: address is defined by html to be the author of the page
right
any xfn links from the page will be presumed to be related to the author's social network
correct
so if i'm on your blog
if they are scoped inside blog posts, then they are temporal in nature
and you link to someone w/ and xfn rel value
i can determine what teh relationship is between you and the linked page
[[include-pattern]] http://microformats.org/wiki?title=include-pattern&diff=0&oldid=5074 * Tantek * (+409) added how a parser sees the resultant included subtree
take a look at the source of this blog post: http://factoryjoe.com/blog/2006/02/21/rati-gets-favorites/
take a look at the link to derek
(in the source)
see here for more on hCard+XFN:
http://microformats.org/wiki/hcard-examples#hCard_and_XFN
it's that simple
we don't need blogrolls if you link to your friends w/ xfn
how widespread is XFN use? I took a look at myspace and livejournal and at a glance neither seem to use it
yeah, a heck of a lot simpler than parsing a bunch of XML/RDF
whereas FOAF, as i understand it, is all in one file
and you have to check that file for updates
yakk, XFN is used all over the place on numerous sites
and if it's fragmented, then what?
tantek, well, I have plenty of XML/RDF tools, but few to deal with microformats
it's distributed
with xfn, no matter where i'm publishing
i can link to my friends explicitly
s/parsing/publishing
xfn & hcard don't seem to me to be what we're after
but they reflect what people publish
you should be after what people visibly publish
see also: http://gmpg.org/xfn/faq
yakk: I wrote a patch for livejournal a while ago to allow for <lj user="foo" rel="xfn-values-here" />
but I don't think it was applied.
yakk: can you be more specific about exactly what we need?
hober, they generate foaf
and also, regarding how do you link to all the places you publish/profiles, see this: http://www.gmpg.org/xfn/and/#idconsolidation
i'm having a hard time seeing it...
yakk: I know
yakk, i'm working with livejournal folks to publish hCard
yakk: my patch was intended to let lj users refer to other lj users with xfn
right there on their *visible* profile pages
same thing for upcoming and flickr
yakk, re: tools, see aforementioned hCard parser by markp
parsing microformats is so easy that a bunch of folks have already written a bunch of open source tools
we want to be able to allow users to publish a profile that encapsulates profile information, links to all the sites the person publishes and social network information
for "inks to all the sites the person publishes" see http://www.gmpg.org/xfn/and/#idconsolidation
tantek, yeah - that looks good
the solution is so simple, just use rel="me"
tantek, I'd really like a way to say "my profile is over on this page"
tantek, so I can link back
much better than a format trying to devlop a taxonomy for all the different kinds of URLs
right, just put rel="me" on that link
to your profile
tantek, well, I'm not convinced that not developing a taxonomy is a good thing
tantek, I might want to say "here are my photos" and "here is a blog"
yakk: I follow the rel="me" link and find an hatom page. :)
or an hresume, etc.
yakk, it's a futile exercise
let the site-linked-to tell us what it is
because new types will always come out
here is my videoblog
here is my calendar
hober, but if both wordpress and flickr publish hatom....
etc.
useless to put it in an explicit taxonomy in a format
tantek, true, but thats not reason to not try
much better to just use tags
yakk: that sounds like a good thing to me! :)
and for that, you can use xFolk
yakk, actually it is
tantek, xfn tries to have a taxonomy of all kinds of relationships
no, just 80%
80/20
tantek, hcard tries to have a taxonomy of all kinds of phone number
hcard gets that from vcard
if we were to do it from scratch today, we would just use tags
ahh tags...
the utility is that tools/apps/services support that taxonomy from vcard
whereas no such thing exists for new formats invented today
tantek, what about xfn's taxonomy of relationships?
tags/folksonomy solves the taxonomy problem for such things so you (the format designer) don't have to
xfn's taxonomy of relationships reflects the 80/20 of what people already publish on the web
it reflects current behavior
xfn's invisibility is a downside
xfolk + rel=me is an interesting concept
i like that
*and* is extensible since folks can create their own XMDP
right factory joe
self-tagging
xflok is hreview wihtout an opinion ;)
yakk, in short, xfolk is a distributed way to publish the equivalent information as del.icio.us
xfolk, not xFlock
KevinMarks, actually, xfolk, has a "description" field
tantek, ahh
because there is no need to keep it all on a centralized service
tantek, what does xfolk provide that say, HTML's A does not?
you should be able to publish your tagged links *anywhere*
xfolk is semantic markup for delicious posts
xfolk is link + tags + description
yakk, that's exactly the right kind of question to ask
we always ask that for microformats
is it just tags?
if you can do it with just HTML, then don't invent a new microformat
tags+comment(description)
and yeah, that's it
doesn't A support description?
no
not in the way people publish delicious links now
right
people publish the link, the tags, and the description all inline, and all visible
xfolk was specifically designed with delicious in mind
tantek, ahh - title & name but no description
and name is deprecated
factoryjoe, but does delicious use it?
and no tags either
yakk, the point of xfolk is to enable bloggers etc. to publish their own aggregatable delicious on their own sites
it doesn't matter if delicious supports it or not
there just seem to be a lot of microformats with no common producers or consumers
yakk, not true at all
for each spec
see the "Examples in the Wild" section for producers
publishers
actually
i talked to them already
and "Implementations" section for producers (tools) and consumers etc.
tons of microformat producers and consumers
so many new ones it is hard to keep up with documenting them
ok
is it just rating + screencap that's new there now
i'm not kidding. i find that to be an actual personal challenge.
that isn't accounted for in xfolk
I don't see that many for for example xfolk (it looks like just one firefox extension consumes it)
ii need to follow up
but they're down w/ xfolk and hcard
yakk, that's a good question to ask on microformats-discuss
its a wiki! shouldn't people implementing be adding them? :)
ahem. yes.
yakk: as hReview is a superset of xFolk, a lot of implementors went with that instead
yakk, ping the mailing list about xFolk, and the author, Bud Gibson is sure to pipe up and provide more info
I'll take a look at hReview
I should join the mailing list
one of us, one of us
the mailing list is worht it for markp's rants alone
I miss markp's rants
http://microformats.org/discuss/mail/microformats-discuss/2005-October/001175.html
[[review-brainstorming]] M http://microformats.org/wiki?title=review-brainstorming&diff=0&oldid=5075 * Tantek * (+106) hReview 0.3 thoughts -
hmm
the 80/20 attitude concerns me
I want my users to have things work all the time
without them caring what technology they're using
yakk, I want perfect tools, too
80/20 doesn't mean 'fails one time in 5'
it means 'solves the core problem'
but I don't want them to be 'in beta' forever
80/20 is the analysis done on use cases
when my friend Maf was working on MS Word Mac, he told me that everyone he met told him:
a) Word is too bloated
b) Cna you just add these 3 features for me?
it's how you prune your feature set to something that a) you can ship, and b) your audience actually cares about (as opposed to esoterica)
kingryan, its hard dude
is this valid hcard:
it's a well established design methodology for practical design
<li class="vcard">
<a href="http://ian.mckellar.org/" class="photo"><img src="/images/avatar_ian_128.jpg" alt="Ian McKellar" /></a>
<span><a href="http://ian.mckellar.org/" class="url fn">Ian McKellar</a></span>
</li>
yakk, nicely done
more ideally, you want to put the class="photo" on the <img>
shouldn't photo be on the <img> tag, or shouldn't the <a> point to an image?
this is on flock.com/about/
yakk, you are correct
why have that extraneous span in there?
so factoryjoe screwed up, but without the right tools there's no way for him to know till some tools work and some don't
I said tools to many times
yakk, try using tails, it will show that stuff for you
that plus X2V works quite well for checking your content
take the url with the hCards, and put it after this: http://feeds.technorati.com/contacts/
tantek, that silently dropped the invalid image elements
[[hreview-v0.2]] N http://microformats.org/wiki/hreview-v0.2 * Tantek * (+34604) archived hReview 0.2 for reference purposes.
I'm going to go and prototype a profile page.
oh
is there a way in XFN to say there is a relationship but not specify what it is
try 'contact'
it just means 'someone I know how to get in contact with'
okay
is there a "reads" kind of concept?
i suppose we could invent rel="favorite" :p
hah
I sense a convergence
why not just tag it with 'reads'
?
kingryan, shouldn't tags be specified by the user rather than generated by software?
<ul xclass="xoxo readinglist"><li class="xfolk-entry">*
anyhoo, I'll hack around a little
[[to-do]] http://microformats.org/wiki?title=to-do&diff=0&oldid=5076 * RyanKing * (-373) WordPress patches for microformats -
I'd say the Netscape bookmark format was woth loking at there
[[to-do]] http://microformats.org/wiki?title=to-do&diff=0&oldid=5077 * RyanKing * (-352) hCalendar/hCard/hReview editor -
DanC, that's not the first time
nor will it be the last
coincides interestingly with bray's "on PHP" I18N follow-up
yeah
PHP needs a kick in the nads over that
there's a lot of utf improvements already in cvs for php6
...to be fair
yay.
i can parse rel-directory, cite-rel and some xfn
rel-tag?
using assaf's css selector ruby parser
i dont need rel-tag for this project :)
but it'd be just as easy.
maybe with your use of rel-directory
yep
code to parse rel-directory:
rule :dirlinks, "a[rel~=directory]", RelDirectory
class RelDirectory
include MicroformatParser
rule_1 :dirname, "a", "text()"
rule_1 :dir, "a", "a@href"
end
it's a pretty neat library.
yakk, you're correct about "tags [should] be specified by the user rather than generated by software"
tantek, for XFN there's not container is there?
[[to-do]] http://microformats.org/wiki?title=to-do&diff=0&oldid=5078 * RyanKing * (+729) Lazyweb - moving stuffs out of my section
yakk, no container
[[hresume]] http://microformats.org/wiki?title=hresume&diff=0&oldid=5079 * RyanKing * (+190) Field details -
so, could people take a glance at http://microformats.org/wiki/hresume#Field_details
I'm trying to communicate that some of the items in the list are class names, while others are just concepts which use html constructs
[[to-do]] M http://microformats.org/wiki?title=to-do&diff=0&oldid=5080 * RyanKing * (+109) Agenda (Wishlist) -
kingryan: I think it's pretty clear in your test
err, text
ok
I just don't want people to scan the list and think that 'contact' is a class name
I'm still having trouble with the 'why'
[[hresume]] http://microformats.org/wiki?title=hresume&diff=0&oldid=5081 * RyanKing * (+15) Job Titles -
is there a good article which describes why I want to embed information in a page rather than link to it?
also, is there any tool to convert an atom/rss feed into an hatom fragment?
limbo_ is Eran and blogs at http://hellonline.com/blog/
tantek, someone vandalized your wiki: http://developers.technorati.com/wiki/attentionxml
yakk, first of all, the DRY princple
the data's already there
so, repeating it elsewhere excessive and introduces opportunities for it to get out-of-sync
2. visible data has social pressure to keep it correct/up-to-date
'cause peeps are looking at it
These logs were automatically created by mflogbot on chat.freenode.net using a modified version of the Java IRC LogBot.
See http://microformats.org/wiki/mflogbot for more information.