Timestamps are in UTC.
[[xfn-clarifications-fr]] M http://microformats.org/wiki?title=xfn-clarifications-fr&diff=0&oldid=19552 * ChristopheDucamp * (+678)
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:Rd9N4t" with an expiry time of infinite: spam
[[profile-examples]] M http://microformats.org/wiki?title=profile-examples&diff=0&oldid=19553 * Tantek * (-15255) Reverted edit of Rd9N4t, changed back to last version by Brian
[[downloads-brainstorming]] http://microformats.org/wiki?title=downloads-brainstorming&diff=0&oldid=19554 * DenisDefreyne * (+111) Adding Spangle/Perrier link
[[geo]] M http://microformats.org/wiki?title=geo&diff=0&oldid=19555 * ChristopheDucamp * (+0) Examples in the wild - moved txioka example to Top of list
[[geo-fr]] M http://microformats.org/wiki?title=geo-fr&diff=0&oldid=19556 * ChristopheDucamp * (+471) exemples dans la jungle - sync'd
[[twitternanoformats-fr]] N http://microformats.org/wiki/twitternanoformats-fr * ChristopheDucamp * (+6052) [fr: translation -> voir aussi picoformats à remettre à jour]
[[twitternanoformats-fr]] M http://microformats.org/wiki?title=twitternanoformats-fr&diff=0&oldid=19557 * ChristopheDucamp * (+121) Proposition de nanoformats pour Twitter (ou Jaiku) -
[[twitternanoformats-fr]] M http://microformats.org/wiki?title=twitternanoformats-fr&diff=0&oldid=19558 * ChristopheDucamp * (+1) 2nanoformats (syntax:param) - typo
[[twitternanoformats-fr]] M http://microformats.org/wiki?title=twitternanoformats-fr&diff=0&oldid=19559 * ChristopheDucamp * (-1) 2nanoformats (syntax:param) -
ddfreyne is nothing
Ronnos is Ron Kok, a friendly student Communication and Multimedia Design in The Netherlands
[[Talk:blog-post-formats]] M http://microformats.org/wiki?title=Talk:blog-post-formats&diff=0&oldid=19560 * ZmfArx * (+15274)
[[Talk:blog-post-formats]] http://microformats.org/wiki?title=Talk:blog-post-formats&diff=0&oldid=19561 * ChristopheDucamp * (-15274) [revert spam]
Ronnos is Ron Kok, a friendly student Communication and Multimedia Design in The Netherlands
Mr_Elusive is not a programmer from id but makes his home at http://eswat.ca
'lo
[[Talk:testimonials]] MN http://microformats.org/wiki/Talk:testimonials * RtdQx4 * (+15485)
[[Talk:rel-tag-spaces]] MN http://microformats.org/wiki/Talk:rel-tag-spaces * SooAat * (+15565)
[[governance-issues]] http://microformats.org/wiki?title=governance-issues&diff=0&oldid=19562 * BenBuchanan * (+17) Petition -
tantek is Tantek <http://tantek.com> and works on Technorati and develops microformats <http://microformats.org>
[[Talk:vcard-suggestions]] MN http://microformats.org/wiki/Talk:vcard-suggestions * JcgFm9 * (+15280)
[[hcard-supporting-user-profiles-fr]] M http://microformats.org/wiki?title=hcard-supporting-user-profiles-fr&diff=0&oldid=19563 * ChristopheDucamp * (+8) typo
[[Special:Log/delete]] http://microformats.org/wiki?title=Special:Log/delete&diff=0&oldid=0 * Tantek * (+0) deleted "Talk:vcard-suggestions": spam
[[Special:Log/delete]] http://microformats.org/wiki?title=Special:Log/delete&diff=0&oldid=0 * Tantek * (+0) deleted "Talk:blog-post-formats": spam
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:SooAat" with an expiry time of infinite: spam
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:JcgFm9" with an expiry time of infinite: spam
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:RtdQx4" with an expiry time of inifinite: spam
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:ZmfArx" with an expiry time of infinite: spam
[[Special:Log/delete]] http://microformats.org/wiki?title=Special:Log/delete&diff=0&oldid=0 * Tantek * (+0) deleted "Talk:rel-tag-spaces": spam
[[Special:Log/delete]] http://microformats.org/wiki?title=Special:Log/delete&diff=0&oldid=0 * Tantek * (+0) deleted "Talk:testimonials": spam
[[Talk:requirements-testing-examples]] MN http://microformats.org/wiki/Talk:requirements-testing-examples * BndO37 * (+15577)
[[social-network-portability-fr]] M http://microformats.org/wiki?title=social-network-portability-fr&diff=0&oldid=19564 * ChristopheDucamp * (+20) Design Patterns et Recettes - typo
tantek is Tantek <http://tantek.com> and works on Technorati and develops microformats <http://microformats.org>
[[social-network-portability-fr]] M http://microformats.org/wiki?title=social-network-portability-fr&diff=0&oldid=19565 * ChristopheDucamp * (-1) Le Problème -
Ronnos is Ron Kok, a friendly student Communication and Multimedia Design in The Netherlands
[[Main Page]] M http://microformats.org/wiki?title=Main_Page&diff=0&oldid=19566 * IMZ * (+689) Drafts - added some short descriptions (taken from the formats' pages)
[[Talk:hresume-feedback]] MN http://microformats.org/wiki/Talk:hresume-feedback * Wf7Vud * (+15319)
[[to-do]] http://microformats.org/wiki?title=to-do&diff=0&oldid=19567 * Tantek * (+300) Analyze existing microformats for opportunities to enhance CSS and propose to W3C
[[Special:Log/delete]] http://microformats.org/wiki?title=Special:Log/delete&diff=0&oldid=0 * Tantek * (+0) deleted "Talk:hresume-feedback": spam
ddfreyne is nothing
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:Wf7Vud" with an expiry time of infinite: spam
[[Main Page]] M http://microformats.org/wiki?title=Main_Page&diff=0&oldid=19568 * Tantek * (-200) edited/simplified summary text down a bit
[[to-do]] M http://microformats.org/wiki?title=to-do&diff=0&oldid=19569 * Tantek * (+33) in general -
[[Talk:hcalendar]] M http://microformats.org/wiki?title=Talk:hcalendar&diff=0&oldid=19570 * Fw7B9j * (+15465)
[[Special:Log/delete]] http://microformats.org/wiki?title=Special:Log/delete&diff=0&oldid=0 * Tantek * (+0) deleted "Talk:hcalendar": spam, not using talk pages
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:Fw7B9j" with an expiry time of infinite: spam
the amount of spam on the μf wiki is awful... shouldn't there be some way to prevent most spam?
ddfreyne, people keep building smarter spam bots
on the subject of spam, new users need to be encouraged to use "Preview" so discussions in the IRC room don't get interuppted by a deluge of edit notifications
tantek: sure, but forgive me if I have the impression that very little is being done to prevent spam
Wouldn't be very difficult to add some kind of capatcha? That'd stop most?
captchas suck ass
just out of curiosity, i tried microsofts' audio captchas once
WilliRaab, SignpostMarv, please add to: http://microformats.org/wiki/to-do#Wiki_improvements
And by doing so, they prevent a lot of spam?
couldn't figure out what the fuck was going on
I dislike graphical captchas... a few days ago I tried signing up at some site and couldn't get past the captcha check myself!
captchas can be programatically decoded into plain text
captchas annoy legitimate users more than they prevent spammers
[[widget]] N http://microformats.org/wiki/widget * Tantek * (+263) create root page to link to specific pages that worked on a widget microformat
I dissagree, it stop most spam, and most users don't seem to mind (as far as I've heard...?).
[[to-do]] http://microformats.org/wiki?title=to-do&diff=0&oldid=19571 * SignpostMarv * (+186) Admins - added encouragement of use of the Show Preview button
I personally quite like the idea of negative turing tests... which presents the user a text (i.e. "microfarmats rock.") and a command ("fix the typo") and then the server can check whether the result is correct ("microformats rock.")
Not getting as annoyed as they'd be by the spamming...
suggestion:
has very little accessibility issues, and unless spam bots learn english, they're impossible to get around
ddfreyne: hehe, good idea, that.
SpamKarama2 for WordPress can be configured to present the user with a captcha if the system thinks it's a spammer
the default comments form has no such field- it is only added for possible spam comments so "humans" can have a second chance
ddfreyne: but that'd take more 'brainactivity' from the user, which I think would bring some over to being annoyed by it.
SignpostMarv: nice.
[[widget-examples]] M http://microformats.org/wiki?title=widget-examples&diff=0&oldid=19572 * Tantek * (+265) needs split into -examples and -formats
if captchas were to be installed on the wiki, I'd push for "only present captcha if an edit is identified as spam, not before"
WilleRaab: that's one drawback... spambots have no brains though :)
anywoo
on the topic of captchas
[[widget-examples]] M http://microformats.org/wiki?title=widget-examples&diff=0&oldid=19573 * Tantek * (+99)
I know a few blogs that have been using it for months, and not a single spam comment came through
does it make sense for class="captcha" to be applied to captcha-related elements ?
that'd be POSH, so, sure... why not?
class="captcha" on any non-form data field would imply "this is a captcha", wheras class="captcha" on input, textarea, select, etc would imply "this is where you put shit"
for multiple captchas, see the following pastebin
[[widget-brainstorming]] M http://microformats.org/wiki?title=widget-brainstorming&diff=0&oldid=19574 * Tantek * (+99) see also
[[Main Page]] M http://microformats.org/wiki?title=Main_Page&diff=0&oldid=19575 * Tantek * (+43)
SignpostMarv?
http://pastebin.ca/654618
sorry, was typing :-P
ah sry. :)
sorry.
haven't been on it in a while,
but Kings of Chaos uses a multi-input captcha method
it displayed an image with a series of numbers
kings of chaos?
you had to click a button for which number appeared twice
ah
one of those lame browser based games
There's a very simple example I've seen on a blogg: "What's the result if you put 7 and 4 together" and then you simply write the number. NEver seen any spam on that blog... then again - it's in Swedish so maybe it wouldn't work in Englsh... =)
regex
[[to-do]] http://microformats.org/wiki?title=to-do&diff=0&oldid=19576 * WilleRaab * (+238) Wiki improvements - < + spam-prevention
Another thing'd be to add more admins to the wiki so there's more ppl removing spam....
http://www.kingsofchaos.com/recruit.php?uniqid=33hw2686
good tihng i didn't delete my account :-P
=)
looking at this example,
Ah, that's a new one.
i think there should be two class values for explictness rather than one implicit value
although I'm a bit mixed. if anyone can find an example of a captcha response that uses links or a flash-based form, then we need an explicit value
thanks WilliRaab
[[xoxo-brainstorming]] M http://microformats.org/wiki?title=xoxo-brainstorming&diff=0&oldid=19577 * TomMorris * (+5) Changed the OPML Manager link description from 'xoxo' to 'unordered list', since it's not actually xoxo.
[[User:SignpostMarv/sandbox/hCaptcha]] N http://microformats.org/wiki/User:SignpostMarv/sandbox/hCaptcha * SignpostMarv * (+1822) randomly off the top of my head
^does that make sense to peeps ?
SignpostMarv: I'm not sure why it's useful
ah
=)
[[social-network-portability-fr]] http://microformats.org/wiki?title=social-network-portability-fr&diff=0&oldid=19578 * JeanChristopheCapelli * (+1) Les Objectifs -
[[social-network-portability-fr]] http://microformats.org/wiki?title=social-network-portability-fr&diff=0&oldid=19579 * JeanChristopheCapelli * (+4) FAQ Portabilité de Réseau Social -
[[Talk:blog-description-examples]] MN http://microformats.org/wiki/Talk:blog-description-examples * NapZvw * (+15387)
[[User:SignpostMarv/sandbox/hCaptcha]] http://microformats.org/wiki?title=User:SignpostMarv/sandbox/hCaptcha&diff=0&oldid=19580 * SignpostMarv * (+1504) intended use, assesment of nefarious use
does that answer the question ? :-P
tommorris is a UK blogger, lives in East Sussex and studies philosophy in London, but hasn't yet said whether he's a quark or a lepton and blogs at http://blogs.opml.org/tommorris
The more I think of it, the better I like the idea ...
I'd like some kind of "capatchablock" to mark up the whole section that's containing the capatcha (and nothing but the capatcha).
the idea of having a microformat that would allow the development of firefox extensions for automatically filling in captchas, rendering captchas pointless ? :-P
SignpostMarv: =)
sshh... ,-)
lol
personally the intent would be to make it more obvious to the end user, and to help disabled or just plain picky users
Wide implementation of OpenID + BotBouncer would solve a lot of the CAPTCHA buttache.
a blind user would have no use for image-based captchas
so there's no point in wasting parsing the image element's alt attribute if there's an audio captcha available
Some standard CAPTCHA links would be useful. <div class="captcha"><a href="[mp3]" rel="alternate" type="[mp3]">Audio version</a></div> etc.
ah
rel="alternate captcha" :-P
alternate = alternate content for document i believe
alternate-captcha would probably be more accurate
or just plain rel="captcha"
Surely alternate is only alterate in /html/head/link?
Ah, the problems of pre-defined class names... ;)
alternate is defined for rel which works on both link and a href
[[to-do]] M http://microformats.org/wiki?title=to-do&diff=0&oldid=19581 * WilleRaab * (-2) Wiki improvements - << spelling is cool, gotta learn it...
would the use of rel="alternate" be borking the spec when referring to an alternate captcha instead of alternate content for the current document ?
"Designates substitute versions for the document in which the link occurs"
So, no, not appropriate.
rel="captcha" or rel="alternate-captcha" to indicate an alternate captcha link, or would be class="captcha" on a link be sufficient ?
Well, your choice of rel or class is dependent on your interpretation of the HTML4/XHTML1 specifications.
question for "why not to use rel at all": rel can't be used on img, object or span can it ?
As far as I'm aware (and I'm not a walking compendium of the HTML specifications), rel is an attribute only for a and link elements.
giving the option to use rel would mean having both rel and class would be annoying
^on a link
e.g. <a href="http://example.com/foo.png" class="captcha" rel="captcha">captcha</a>
the rel is redundant
When parsing for a 'elemental microformat' style format, I generally parse for class and rel.
(you could argue that the class is redundant)
(In this case, class=nsfw, an unsanctioned microformat, I thought that people could be liable to use class or rel - so it's better to specify class but parse for either)
<link rel="bookmark nsfw" href="http://example.com/nsfw.html" /> in the document http://example.com/nsfw.html would indicate that nsfw.html is NSFW
rel be better :-P
Well, the intended use was on 'a' rather than 'link', since the use case is "I'm linking to a document I think may be not safe for work". You generally do that in the body of the page, not the head.
And you can put a class attribute on a link element.
using link for self-linking would be useful for google safe search
You could use <body class="nsfw"> for that.
hrm
well the usage of both would be useful for marking up inline content i suppose
[[table-examples]] M http://microformats.org/wiki?title=table-examples&diff=0&oldid=19582 * UjfPwk * (+15235)
The semantics of it are extensible, what with profile URIs.
[[show-formats]] M http://microformats.org/wiki?title=show-formats&diff=0&oldid=19583 * LxoK80 * (+15601)
heh
[[rel-tag-faq]] M http://microformats.org/wiki?title=rel-tag-faq&diff=0&oldid=19584 * FpaNmu * (+5890)
[[rel-tag-other-evidence]] http://microformats.org/wiki?title=rel-tag-other-evidence&diff=0&oldid=19585 * AndyMabbett * (+118) Further reading
[[widget-examples]] M http://microformats.org/wiki?title=widget-examples&diff=0&oldid=19586 * Chris Messina * (+5)
[[Main Page]] http://microformats.org/wiki?title=Main_Page&diff=0&oldid=19587 * Chris Messina * (+93) Added figure links
tantek is Tantek <http://tantek.com> and works on Technorati and develops microformats <http://microformats.org>
[[User talk:WebOrganics]] http://microformats.org/wiki?title=User_talk:WebOrganics&diff=0&oldid=19588 * WebOrganics * (+6079)
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:FpaNmu" with an expiry time of infinite: spam
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:LxoK80" with an expiry time of infinite: spam
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:UjfPwk" with an expiry time of infinite: spam
[[table-examples]] M http://microformats.org/wiki?title=table-examples&diff=0&oldid=19589 * Tantek * (-15235) Reverted edit of UjfPwk, changed back to last version by Phae
[[rel-tag-faq]] M http://microformats.org/wiki?title=rel-tag-faq&diff=0&oldid=19590 * Tantek * (-5890) Reverted edit of FpaNmu, changed back to last version by TobyInk
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:NapZvw" with an expiry time of infinite: spam
[[Special:Log/delete]] http://microformats.org/wiki?title=Special:Log/delete&diff=0&oldid=0 * Tantek * (+0) deleted "Talk:blog-description-examples": spam
[[show-formats]] M http://microformats.org/wiki?title=show-formats&diff=0&oldid=19591 * Tantek * (-15601) Reverted edit of LxoK80, changed back to last version by Tantek
[[Main Page]] M http://microformats.org/wiki?title=Main_Page&diff=0&oldid=19592 * Tantek * (+20)
tommorris, SignpostMarv, that use of rel (rel="nsfw") is an abuse of rel and not in accordance with the semantics of that attribute
if you want to note that the destination of a link is nsfw, use xfolk to tag the link with nsfw
see http://microformats.org/wiki/rel-faq
tantek: I'm not advocating for or using rel=nsfw. I'm using class=nsfw. See http://tommorris.org/profiles/nsfw
xFolk is not appropriate for human authoring. It's fine if it's being pumped out by a script. But there's no way that I'm writing out an xFolk entry to tag a link.
tommorris using class for content (which ostensibly "nsfw" is) is an abuse of class
"nsfw" is semantically a tag on a link
the best we have right now of representing that semantic is xFolk
if you want to add specific issues with authoring xFolk to /wiki/xfolk-issues , that would be appreciated
(I'm not debating your assertion that you don't find it appropriate for human authoring, I just want to capture it and move forward.)
(and avoid abusing the class attribute in the mean time)
I'm still not sure how <a class="nsfw"> is an abuse of @class
because nsfw is *content*
it's something you would translate/localize
class (and rel for that matter) are not things you would translate/localize, like element names
it's for type/property names etc.
and hiding human content (which the string "nsfw" is) in the class attribute is also an abuse.
i need to write up an FAQ point on this
because obviously it is a subtle point
Well, it can be displayed by the author in any way they like. I use CSS generated content to add a string after links marked NSFW.
I have also thought about writing a DOM script that would pop up a box saying "Hey, you clicked on a link that the author has marked NSFW. Are you sure you want to continue?"
ISTM that writing <a href="...">... (NSFW)</a> captures the semantic better than class="nsfw"
requiring CSS to show content is a non-starter
Some people may choose to add it as a string to their page.
we already had this discussion
In fact, most of the uses on my site are del.icio.us links that I tag 'nsfw' (and are then, by extenson, xFolk). But for human authoring, i found it a useful pattern to share.
search the lists for rel="xxx"
it's an anti-pattern
as explained above, abuse of class attribute
I've seen the discussions on uf-discuss, and that is the reason I chose not to propose it.
But it's still something that I find useful. If other people don't want to use it, that's fine.
ok, we'll just call it shady POSH for now ;)
That's fine.
I will extend the XSLT so that anyone choosing to use the nsfw profile will have xFolk-with-nsfw-tagged-links marked as NSFW in the RDF version.
It seems like using class="nsfw" as a hook for DOM scripting that solicits user confirmation before following the href is conflating two separate semantic bits here, anyway
So separate the confused bits. <a class="protected" href="...">... (NSFW)</a> (or the xFolk equivalent)
use .protected as your hook for the DOM scripting
and describing the link as NSFW is content on the page
[[xfolk-issues]] http://microformats.org/wiki?title=xfolk-issues&diff=0&oldid=19593 * TomMorris * (+238)
[[xfolk]] http://microformats.org/wiki?title=xfolk&diff=0&oldid=19594 * TomMorris * (+124) Many more examples - added self to examples in the wild
[[figure-examples]] N http://microformats.org/wiki/figure-examples * Chris Messina * (+4796)
[[downloads-brainstorming]] http://microformats.org/wiki?title=downloads-brainstorming&diff=0&oldid=19595 * DenisDefreyne * (+203)
julianstahnke is Julian Stahnke and works for last.fm and implements microformats wherever he can
[[downloads-brainstorming]] http://microformats.org/wiki?title=downloads-brainstorming&diff=0&oldid=19596 * DenisDefreyne * (+837) Adding To Do section; adding a few more issues
factoryjoe: hi
[[downloads-brainstorming]] http://microformats.org/wiki?title=downloads-brainstorming&diff=0&oldid=19597 * DenisDefreyne * (+510) Adding version numbering issue
factoryjoe: are you staying now? ;)
hi ddfreyne
Alright, the downloads-brainstorming page has 3 more issues now
one about version numbering (how to mark up version numbers)
one about naming (appcast/downloads/blah)
and one about a smaller downloads microformat... although i'm not entirely sure how useful this would be
will need to investigate using hListing instead of hAtom...
and marking up changesets (trac, websvn?) using this as well
right
and should we bother with beta/major releases?
prereleases/betas/point releases?
also
do we know the xml format of OSX Software Updates?
actually
we do
that's actually something that is being discussed by some Sparkle people... something called "branches" that uses posets etc... let me find a link...
http://sparkle.andymatuschak.org/wiki/Documentation/PosetVersioning — but that's just brainstorming, random ideas flying around
hm... OS X software updates... I've never seen that format
one sec
i can get it
what the...
I ran a packet sniffer, and I'm detecting lots and lots of porn being downloaded by this computer
?!
are you on a PC?
no, on a Mac, which is very weird
http://ax.phobos.apple.com.edgesuite.net/WebObjects/MZStore.woa/wa/com.apple.jingle.appserver.client.MZITunesClientCheck/version
check that out
that's all of the apple software updates
oops
try this: http://phobos.apple.com/version
Ahh, yes, I've seen that before
(I mean formats)
i need to restart
can you addd that?
this porn problem is extremey, extremely awful
will do
if you figure out the porn issue, let me know
that's very very odd
brb
[[downloads-formats]] http://microformats.org/wiki?title=downloads-formats&diff=0&oldid=19598 * DenisDefreyne * (+1031)
wb
ok back
porn issue is solved; it's not my computer...
but let's just ignore that
ah, one thing the SU XML file has is a distinction between build numbers and version numbers
that's something appcasts should have as well... sparkle uses "version" and "shortVersionNumber"
"shortVersionNumber" contains the actual version number if "version" contains a build number, IIRC
yeah
that's good
we should look at overlap w/ changesets
I'm not sure I like the names "version" and "shortVersionNumber"... revision/build/something? hm.
i don't, but we just need to document what's out there
[[downloads-brainstorming]] http://microformats.org/wiki?title=downloads-brainstorming&diff=0&oldid=19599 * DenisDefreyne * (+449) Adding "Overlap with Changesets" issue; adding hListing todo item
Hm, there's also http://swscan.apple.com/content/catalogs/index-1.sucatalog
the one you linked to is probably iTunes-specific
[[downloads-formats]] http://microformats.org/wiki?title=downloads-formats&diff=0&oldid=19600 * DenisDefreyne * (+127)
yeah you;re right
that's the phobos apple server
all itunes/ipod/iphone specific
something like this is a bit too much for what we need
http://swcdn.apple.com/content/downloads/33/03/018-2946/JBXBxSWJvhjdFzWXKM3JTSpn5hyNRpMjzv/AppleKeyboardSoftware.dist
whoa
but that's useful
what does microsoft do?
for msft updates?
I'd have to reboot into windows
[[downloads-formats]] http://microformats.org/wiki?title=downloads-formats&diff=0&oldid=19601 * DenisDefreyne * (-736) Removing Apple SU XML sample and adding link to a typical .dist XML file
hm, I can't find anything on MS' update mechanism
"function canHasDriver()" — do I detect lolcat speak in the software update XML file?
consider it detected
awsum thx
invisible highfive!
also... what do you think about appcasts with downloads for multiple OSes, architectures, ...? I've been trying to find a way to mark up such appcasts, but it's hard
have you considered "going backwards"- e.g. rather than trying to spec a microformat, spec an XML data format- then a microformat from that
Sparkle would need a way to filter out the updates it can't use... that would require marking up the OS, architecture, ... of each download, which (I think) leads us a bit too far
there already is an XML data format—RSS with Sparkle extensions
does that support multi OS/arch ?
but brainlessly porting a XML format to a microformat doesn't always seem to work well
things can be fixed in translation
it doesn't; there are HTML pages out there that have downloads/release notes for multiple OSes and architectures
microformats are meant to be based off of real-world examples
hCalendar, hCard etc
obviously
1:1 translation
if there is a data format that does the job, translate it and see what happens :-3
I'd like to forget about multi-* appcasts, unless someone can come up with a great way to mark them up
didn't I do that already ? :-P
you did?
can't have been that great if I already forgot about it. :P
lol :P
but please add it to the wiki... [[downloads-brainstorming]]
well hold on
um
i mean i guess i see the os native thing
defining each OS and each architecture as an hCard
and referring to it in the appcast microformat via <object>
SignpostMarv: you'll have to mark up the OS type, OS version, architecture, possibly application kind (pro/lite/etc)
SignpostMarv: and I think that's bloat
can the details for an OS and architecture be marked up as an hCard ?
yes/no
bloat bloat bloat!
no hcards!
I think having a hCard for Mr. Windows Vista is rather odd
hcards are for people!
hcards are soylent green!
could they, not should they
could they not!
the point i'm making is this:
anyway architecture could just be <span class="architecture">
if you use <object> to link to data elsewhere in the document or on the web, then you avoid the problem of the infinite number of class names you could generate when referencing an OS
a "platform" is a combination of OS + architecture, right ?
oh man
SignpostMarv: have you been around here much?
you're talking crazy!
:P
modularisation of microformats
plain text is probably insufficient to describe a platform
It's not a question of how it should be done; it's more like a question of whether it really should be done
(multi-* appcasts)
it's about how it's *already* being done
modularising the declaration of a platform out of the appcast microformat would remove a lot of duplication of data, and allow for more data to be described about a platform
actually
DUH
architecture is just tags!!
it's compatible w/ hAtom that way
since many OSX updates will be PPC AND Intel
those are just tags
what data is associated with architecture
what data?
well, processor-native code
tantek is Tantek <http://tantek.com> and works on Technorati and develops microformats <http://microformats.org>
yo tantek
marking up a page like http://dev.mysql.com/downloads/mysql/5.0.html would still be a big task, heh
wow look who has returned from a long absence ;)
That page's a bit extreme, but still :)
;)
DTDs are hosted by the peeps who developed the spec. why not have a microformat that defines a platform/os/arch hosted by the people who develop it ?
SignpostMarv: repeat your idea for tantek
<SignpostMarv> DTDs are hosted by the peeps who developed the spec. why not have a microformat that defines a platform/os/arch hosted by the people who develop it ?
because that's the wrong point in the analogy
do you mean XMDP profiles?
probably
and people can, it's just that in practice fewer people pay attention to XMDP profiles than do to DTDs, and those are quite few
also, a built in assumption of DTDs is one per document
the idea:
whereas XMDP profiles can be multiple per document
microformat for describing a platform embeded into http://www.microsoft.com/windows/products/winfamily/ie/default.mspx
so the "why not" question doesn't make much sense
ask it as a positive instead
if the software runs in IE, <object class="platform" data="http://www.microsoft.com/windows/products/winfamily/ie/default.mspx" />
you want to encourage bindings to proprietary binary formats why again?
in stead: <a rel="tag" href="http://www.microsoft.com/windows/products/winfamily/ie" />
that's an HTML document
that's tagging something with "ie"
right
not a binary format
tantek: we're talking about architectures in download formats
like PPC, Intel, Windows, etc
i suggested that we use tags
I'm suggesting we don't :-P
so you mean runs in Vista
there are multiple questions
whether you want a taxonomy or folksonomy etc.
a registry ala MIME Content-Types like text/html
What if an appcast contains two actively developed branches, say 1.0 and 2.0... would you still use tags to distinguish between 1.0 and 2.0?
ddfreyne, is "two actively developed branches" the 80/20 ?
solve the 80/20 first
of real world examples
wordpress 2.0, 2.2; PHP 4,5
before asking such theoretical "what ifs"
actively maintained branches
Apache 1.2, 2.0, and whatever it is we're on now
1.3/2.2
yeah
let's go back to the original application...
"appcast" still sounds like a really poor name
delivering HTML-based application updates w/ payloads
"apps" are a *fraction* of downloads
plugins
tantek: we're not finalized on "appcasts"
appcasts is used in the wild
hDownload :-P
shouldn't even use them in the discussion
like podcast
tantek: I agree that the name "appcast" isn't great, but "downloads" isn't great either... these are versioned downloads
http://microformats.org/wiki/process#Naming_considerations
the problem is that "downloads" "software" and similar "client-specific" names will be obsoleted
you don't always "download"
yeah you do
the UIs of all these update thingie say "downloading..."
hDistribute ?
SignpostMarv, please see prev URL
no
they say "updating"
no, they say "downloading updates..."
hehe, i'm just kidding.
tantek: when you return please provide evidence
i have seen "downloading..."
but i have seen "updating..." as well
http://flickr.com/photos/factoryjoe/tags/updater
updating is installing
http://flickr.com/photos/factoryjoe/511599881/
no
updating is updating
downloading is getting it
installing is typically done once
once you have installed something, you update it
you install an update
http://flickr.com/photos/factoryjoe/470068648/
"install..."
not "download"
you download data, then you install it
I'd say updating is a combination of downloading and installing
an update is installing a newer version of software that you already have
e.g. upgrading
downgrading is installing an older version of software you already have
(or less/more feature complete)
^pro/lite etc
installing is just installing software you don't have
so you can't install an update?
are you defining an update as a whole piece of software or a patch ?
a whole piece of software
i don't think it matters
in practice, people send whole apps
not patches
that'd be an upgrade, not an update methinx
i don't think it matters
people call it update all the time
meh. semantics of the english language suck
I'd still like having a "download" microformat... something that adds extra data to a rel-enclosure
the point is you can download data in 1 of 4 cases: you don't have any other version of it, you have a previous version, you have a newer version, or you have the same version
and that download microformat could also be used by the, eh, "update" microformat
we need to document macupdate.com, versiontracker.com and iusethis.com
oldversion.com
<div class='header'>DESCRIPTION</div> ew.
;)
we document
we don't necessarily use
my admiration for macupdate just vanished
wouldn't it be a good idea to look at the specs for CVS and SVN ?
well, that's for changesets... which are similar... so changesets could perhaps be merged into all this
i'm referring more to how it describes the transaction
you wouldn't redownload a file if the version hasn't changed
similarly, you wouldn't redownload an application if the version hasn't changed
sparkle handles that
is sparkle an application or a data format
sparkle is an application (app updater framework), and it uses RSS appcasts
okay, so drop sparkle out of the discussion and focus more on data formats and data exchange
sparkle downloads the appcast RSS feed, extracts application versions, and checks whether any version is newer than the application that is being updated
sparkle is an appcast client, not an appcast
sparkle is used in many mac os x apps, so ignoring it is probably not a great idea
sparkle is a framework, technically
brb
what i mean is, i'm cringing every time you mention sparkle
sparkle is one thing
it does shit
start considering other stuff :-P
there's no point designing a microformat based around one thing
Many apps that had their own update scheme have switched to sparkle
</end vent>
also, appcast are not sparkle-specific
if everything uses sparkle, then there'd be as much need for an appcast microformat as there is for hAtom
e.g. little or none
but sparkle did help shape appcasts, and it's pretty much the most popular app update framework out there
so sparkle influenced appcasts
what influenced sparkle ?
first there were appcasts (RSS feeds with application update enclosures), and then there was sparkle
which is pretty much the appcast "killer app"
if sparkle does the job perfectly, translate it and leave it at that
that's pretty much what I've done so far
but there are a few issues; check the downloads-brainstorming page
[[downloads-brainstorming]] http://microformats.org/wiki?title=downloads-brainstorming&diff=0&oldid=19602 * DenisDefreyne * (+77)
is it a case of "we want to do stuff that would basically fork sparkle" ?
I don't want to fork sparkle at all
then don't. any situation where you want to add a feature to the microformat that doesn't exist in sparkle, get sparkle to add it to their spec first
if sparkle is the shiznit, then that missing feature won't be missed very much
that's pretty much the reason why I don't like multi-OS/arch/etc appcasts
is that feature ever going to be in sparkle ?
no idea :)
then just focus on doing hSparkle till it's added
leave the headache up to the peeps who maintain sparkle
everything you keep saying keeps leading back to sparkle
[[Main Page]] M http://microformats.org/wiki?title=Main_Page&diff=0&oldid=19603 * MvtPzw * (+1318)
[[Main Page]] M http://microformats.org/wiki?title=Main_Page&diff=0&oldid=19604 * Tantek * (-1318) Reverted edit of MvtPzw, changed back to last version by Tantek
it does... but that's because pretty much the only implementation that uses RSS appcasts is Sparkle, and it's also the reason the appcast microformat was started
so please, for the sake of my sanity, just do hSparkle :-P
[[Special:Log/block]] http://microformats.org/wiki?title=Special:Log/block&diff=0&oldid=0 * Tantek * (+0) blocked "User:MvtPzw" with an expiry time of infinite: spam
just put the platform info in the feed description and have multiple hAtom feeds in the document for the multi platform junk till sparkle changes
Ahh, sparkle 2.0 is going to have something like that... probably worth following the discussion on the sparkle mailinglist and seeing how they are doing it
and in the mean time, I think multi-* appcasts aren't worth it
multiple hAtom feeds within a document implies multi-* appcasts :-P
no, still a single hatom feed
http://sparkle.andymatuschak.org/wiki/Documentation/PosetVersioning
why not one for each * ?
that could work too
ther you go, problem solved :-P
anyway, I'll just ignore multi-* appcasts for now... it's been discussed for weeks and we're still nowhere
just whack 'em all in <body class="hSparkle"> and you'redone
(or whatever class name you'd be using)
you really make everything look trivial
ideas are trivial
implementation might not be :-P
conception & implementation are two different things, an thankfully, conception is much easier to do :-P
well, I'm heading off
g'night!
toodles
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.