and
TextThis is an automatically generated IRC chat log made by the logger bot from the Semantic Web Interest Group IRC chat at irc://irc.freenode.net/rdfig (also known as server irc.freenode.net channel #rdfig if that URI does not work for you).
NOTICE: #rdfig logs are being turned off 2004-12-03. Please
switch to the new and
shiny #swig channel for Semantic Web Interest Group chat.
Change your client to #swig and enjoy the new experience.
Or read the latest #swig logs
to see what you've been missing :)
Semantic Web Interest Group Logs > 2001 > 2001-08 > 2001-08-15 (Latest) (Search)
00:02:02 Topic now Semantic Web Chat: I want my data back! [ http://www.w3.org/RDF/Interest#irc ]
00:02:02 Users on #rdfig: logger Morbus karld DanC_ barstow deltab eikeon dc_rdfig nephrael em
00:02:02 <ChanServ> This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome
00:04:50 <deltab> F:|Apache's Handler Use
00:04:51 <dc_rdfig> titled item F
00:05:30 <Morbus> ?
00:05:57 <deltab> I gave the URL earlier, and it got chumped
00:06:19 <deltab> so I might as well fill out the entry a bit
00:07:01 <Morbus> i have no clue what you're talking about. but that's ok :)
00:07:21 <deltab> see http://rdfig.xmlhack.com/
00:08:05 <Morbus> ah. so the X: syntax posts to the site?
00:08:09 <deltab> yes
00:09:02 <Morbus> hmf. some of myother ssi's in this article.shtml page aren't being included now.
00:09:28 <deltab> is the regex matching those?
00:09:51 <Morbus> oh. poo. i'm still using redirectmatch. that's why.
00:10:00 <Morbus> sigh. i was one level lower than i thought cos of the revised syntax.
00:10:02 <deltab> heh
00:11:17 <Morbus> on a side note, doing ^/gm/ with Rewrite didn't work. 404 with no change in url
00:12:20 <deltab> F:Apache allows handlers to be called for specified directories and filename extensions
00:12:20 <dc_rdfig> commented item F
00:13:05 <Morbus> one of the bad things about this system is that without another include, i can't nicely cover the <title> element.
00:13:12 <deltab> F:Handlers can be provided by modules, or be arbitrary programs
00:13:12 <dc_rdfig> commented item F
00:13:30 <Morbus> s/be/by/
00:13:31 <deltab> hmm, that's annoying
00:13:55 <Morbus> yeah. i mean, theoretically, i coudl override it with my includes, outside the <head> but that makes me feel dirty
00:14:19 <deltab> that'd be bad html, and it wouldn't always work
00:14:24 <Morbus> exactly.
00:15:06 <Morbus> i wonder how i should handle that.
00:15:38 <Morbus> i don't knowmuch about js and dom. but if the title of the article has a certain class and that class only belongs to the title, could i grab it that way?
00:16:07 <Morbus> cos it's class=title in my css.
00:19:00 <deltab> dunno if that would work
00:19:11 <deltab> where are you getting the title from?
00:19:31 <deltab> is it in the article files?
00:19:48 <MorbusIff> welp, in my included data file, the actual title of the article is contained as class="title" and the contents are replicated in <title> (hard coded).
00:20:21 <MorbusIff> so if i can't modify the title before browser display, i'm wondering if after the SSI kicks in, I can js/dom class="title" into the <title> of the page.
00:20:45 <MorbusIff> take a look at any article on gamegrene.com for an example
00:20:54 <deltab> that'd certainly not work in search engines
00:21:11 <MorbusIff> hmf. yeah, that's true, and that's the important one.
00:21:50 <MorbusIff> does <head> have to come before body? could it be at the bottom of the page?
00:22:07 <MorbusIff> thats a silly question.
00:22:08 <deltab> yes it does
00:23:21 <MorbusIff> hmm... besides doing a title specific SSI, I cna't think of anything else.
00:23:47 <deltab> what about gamegrene?
00:24:14 <MorbusIff> oh, any article on there source'd shows the class title thing.
00:26:32 <deltab> as far as my browser's concerned, the html title there isn't connected to anything with class "title"
00:26:44 <MorbusIff> oh no, i know.
00:26:54 <deltab> maybe that's how it's done internally - I can't tell
00:27:20 <MorbusIff> no, right now it's not dynamic at all. the <title> and class are hard coded to what they are. the only thing SSI'd on that page is the sidebar and the ads.
00:27:28 <MorbusIff> everything else is pure static.
00:27:33 * deltab nods
00:27:55 <MorbusIff> for browsers at least, i may be able to do that js/dom thing, but i still can't think of a way to do it for engines, which means, i'm gonna be stuck with doing a title.include
00:28:00 <MorbusIff> which is a pain.
00:28:15 <deltab> if the title was in a separate file, you could include that separately
00:28:22 <MorbusIff> yeah, that's the pain :)
00:28:46 <MorbusIff> in my quest for templating, i'd be looking at #.title, #.article, and #.comments if I want it semantic.
00:28:58 <MorbusIff> #.title and #.data if I don't care about article/comment split
00:30:17 <deltab> I take it add conditional SSI to the article files is out of the question?
00:30:25 <deltab> ^adding
00:30:47 <MorbusIff> i'm not adverse to it if it's a good idea. what'd you have in mind?
00:31:22 <deltab> well, it's rather a hack
00:31:46 <deltab> the template could set variables indicating title or body context
00:31:58 <MorbusIff> you mean, more so than this "templating" system? :)
00:32:30 <deltab> then the included file has ssi conditions to display only the title in title context, and only the body in body context
00:32:39 <deltab> yes, much more so :-)
00:33:08 <deltab> I'm assuming you want the article files to be uncluttered by presentation details
00:33:16 <MorbusIff> exactly, yes.
00:33:28 <MorbusIff> MorbusIff is now known as Morbus
00:34:23 <Morbus> ok. so, the template would do something like include=#.article?title and #.article?article
00:34:37 <Morbus> and then then #.article file would check for those and display the right crap.
00:34:47 <deltab> ah, that's possible
00:35:16 <Morbus> heh. same thing, different way. aren't misunderstandings grand? <g>
00:35:42 <deltab> I was thinking <!--#set var="context" value="title" -->
00:36:28 <Morbus> right. the sad thing is i've never used set before. sigh. :)
00:36:34 <deltab> but using the query string would be marginally neater on the template side
00:36:40 <deltab> me neither :-)
00:36:44 <Morbus> hehehe
00:37:45 <Morbus> yeah, i could even save duplication by checking for article or title on the class=title and just article for the rest...
00:38:15 <Morbus> and since its commented, any parser i write would ignore it. so that would work nicely.
00:40:27 <deltab> inside each article though, it'd mean wrapping non-title text in an ssi if command that suppresses the html
00:43:20 <deltab> [if not title context]<h1>[end if]the title[if not title context]</h1>rest of article[end if]
00:48:34 <deltab> inside each article though, it'd mean wrapping non-title text in an ssi if command that suppresses the html
00:48:37 <deltab> [if not title context]<h1>[end if]the title[if not title context]</h1>rest of article[end if]
00:49:13 <Morbus> yeah, i just started realizing that as i'm playing with it here.
00:49:29 <Morbus> it may look cleaner just to do two ssi's and duplicate the title
00:50:03 <deltab> may I ask why you have such constraints?
00:50:13 <Morbus> which constraints?
00:50:40 <deltab> no cgi, simple article files
00:51:30 <Morbus> ah. well, i don't really. i work at the isp where all my crap is hosted. i could run slashcode if i really wanted to (heh), but i try to look forward to the inevitable evil of "host moving". if i can't get something set up in five minutes on a new host, i don't want to be down for 10.
00:52:07 <Morbus> the no cgi thing is a constraint in this case cos i don't want the extra overhead, either through mod_perl or repetitive hitting of a cgi. i'd rather just have apache itself and only apache do all the logic.
00:52:40 <deltab> is this ssi really required, or could you generate static pages?
00:52:54 <Morbus> i coded the perl posting system that makes gamegrene look that way - it just knows enough to create all the files in the right place and has an auth system, etc... but it only creates the static files so there's obviously no udnerstanding of what data is what.
00:53:06 <Morbus> welp, right now its generating static pages.
00:53:34 <Morbus> my problem is I don't feel like up'ing and down'ing 1000+ pages when i have to do a template change. i'd rather just change five or six files (which this template system would allow).
00:55:07 <deltab> I think there's something to be said for pages which, once published, stay the same :-)
00:56:07 <Morbus> heh. i know it :) ... my main company, disobey.com, is all static, with over a thousand pages.
00:56:44 <deltab> ah, I recognise that name
00:56:49 <Morbus> really?
00:56:56 <Morbus> yeah, i'm him. hi. :)
00:58:37 <deltab> aha, http://disobey.com/ghostsites/ :-)
00:58:56 <Morbus> yes indeed. thats where you know it from?
00:59:03 <Morbus> thats one of its most popular sections
00:59:14 <Morbus> i used to have netslaves.com too, but sold it off.
00:59:19 <deltab> no
00:59:53 <deltab> someone I know gives lots of URLs in irc
01:00:02 <Morbus> ah.
01:00:04 <deltab> I don't have time to follow all of them
01:00:35 <Morbus> hmm.
01:00:44 <Morbus> oh. duh. nevermind.
01:00:53 <deltab> no offence intended :-)
01:01:07 <Morbus> oh, none taken ;)
01:01:49 <deltab> I'd glanced at the front page before, but decided it was much too interesting to allow myself to start reading :-)
01:02:07 <Morbus> there ya go. now you're geting it <g>
01:02:24 <Morbus> yeah, see this is messed up.
01:02:34 <Morbus> the ssi's in my .article are not being interpreted.
01:02:41 <deltab> *so* many thing to do, so little time
01:02:52 <Morbus> i'm just seeing my conditionals spit out raw.
01:03:20 <deltab> what do the filenames end in?
01:03:34 <Morbus> dammit. nevermind.
01:03:36 <Morbus> hehehe
01:03:43 <deltab> and do you have server-parsed set as the handler for it?
01:03:47 <Morbus> suddup i'm tired <g>
01:04:02 <deltab> (they don't have to end in .shtml)
01:04:32 <Morbus> i know, i'll .htaccess in the right extensions later, for now, i just tweaked to shtml.
01:04:46 <Morbus> i'm testing this all off my os x box, config'd just like the linux box at the isp ;0
01:04:49 <deltab> hey, maybe shtml files can be used as handlers
01:06:55 <Morbus> oh, well this is queer.
01:06:57 <Morbus> i think.
01:07:41 <Morbus> you wanna know what's happening? man,
01:08:04 <Morbus> this is in my template file, neatly sandwiched in between by <title>'s...
01:08:06 <Morbus> <!--#include virtual="${QUERY_STRING}.shtml?title" -->
01:08:12 <Morbus> and it works rather nice.
01:08:25 <Morbus> this is for the article itself:
01:08:29 <Morbus> <!--#include virtual="${QUERY_STRING}.shtml?body" -->
01:08:55 <Morbus> however, by the time we get to that point, QUERY_STRING has been redefined to 'title' cos of the first include.
01:09:02 <Morbus> and thus it fails.
01:09:40 <Morbus> phoeey
01:10:02 <deltab> <!--#set var="article" value="${QUERY_STRING}" -->
01:10:03 <Morbus> the quick solution is to set a temp variable to remember the query string.
01:10:06 <Morbus> heheheh.e
01:10:09 <Morbus> yeah, same wavelenght.
01:18:15 <Morbus> goddamn. os x has this awful habit of locking up the modem connection if you have a lot of incoming traffic (caused, for example, by messing heavily with the server on your own system).
01:18:26 <Morbus> anyways, the set thing worked.
01:21:25 <Morbus> what's your email deltab? i'm gonna be writing this up for a evolt.org article and want to attribute you
01:21:50 <deltab> hrm, what address to give?
01:21:59 <Morbus> heh
01:22:08 <deltab> I have so many
01:22:40 <deltab> nearly infinitely many, in fact :-)
01:24:29 <Morbus> heh. i only have three. 1 is my "real" isp account, and then morbus@gamegrene.com and morbus@disobey.com which are so atypically boring.
01:26:17 <Morbus> heh. you own espnow.com perhaps? thus the infinitely many?
01:26:28 <deltab> not own, no
01:26:34 <deltab> work for
01:26:46 <Morbus> ah. you chose pokemon names? works for me :)
01:27:26 <deltab> yeah, the first machine was charizard, to tie in with the web hosting company blazingfast
01:27:34 <Morbus> hehehe...
01:28:09 <Morbus> i've got a rather boring name now. into our dialup ports, of course. even our work machines are named rather mundane.
01:29:16 <deltab> I can set up subaddresses for mail very easily, and occasionally I use some in an effort to track spam
01:29:31 <Morbus> yup. seed boxes.
01:30:32 <deltab> hasn't really worked though, and (ObRDFIG) it makes matching up my addresses harder for computers
01:31:02 <AaronSw> hah, morbus was making fun of me for having so many addresses
01:31:07 <AaronSw> the tables have turned, it seems ;-)
01:31:12 <Morbus> hehehehe...
01:31:20 <deltab> so, use deltab@osian.net
01:31:26 <Morbus> wha? not at all. i have 3. i only publicize one. phbbt.
01:32:05 <Morbus> k. i'll probably have the article up in a few days.
01:32:39 <Morbus> but if you want aaron, you can send me email to whatever @gamegrene.com or @disobey.com to make me feel special <g>
01:32:41 <deltab> out of interest, what parts of the pages need to be changeable? adverts, anything else?
01:33:12 <Morbus> welp, the adverts are cgi-based so the urls remain the same, everything changes elsewhere. the right ahdn column was already a ssi anyways.
01:33:38 <Morbus> this is more of a protection in the future. like i want to eventually add a return to top thing, and that means changing about 200 files right now and up/downing those over a 56k line.
01:33:52 <Morbus> and we're throwing about five new pages up a week. it'll get outta hand quickly.
01:34:37 <deltab> could you not just make the change to new files? or would the inconsistency be too much?
01:34:59 <Morbus> it would drive me slowly nuts. i can't stand inconsistency :)
01:35:06 <Morbus> its like some evil neurosis of mine...
01:35:09 <deltab> ah, okay :-)
01:36:09 <deltab> I'll just have to make sure snapshots of the pages are taken from time to time :-)
01:36:20 <Morbus> well, i'll brb. gonna take a shower and dwell on crap to make sure we're not missing anything.
01:36:48 <Morbus> heheh... i have archived versions of disobey.com back from 97 when i started. i should throw them up sometime so the progress could be shown.
01:36:49 <deltab> what about calling partial templates from the articles?
01:37:05 <Morbus> what do you mean?
01:37:33 <deltab> <!--#include virtual="header" -->
01:38:02 <deltab> and like wise for the footer
01:38:40 <deltab> or three, allowing for the part between the title and the body
01:38:43 <Morbus> well, the reason i never did that was cos i never could get embedded ssi's to work.
01:38:52 <Morbus> so the ads and rightcolumn never worked.
01:38:57 <deltab> ah
01:39:11 <deltab> well, they work for me
01:39:31 <Morbus> yeah, and they work for me now too. i think my problem before was the file etxension and it just never dawned on me.
01:40:10 <Morbus> now that i know they work, i can't think of a reason to go through this other templating system we just hacked together.
01:41:22 <Morbus> besides that the <title>thing would be a little uglier using header/footer.
01:42:09 <AaronSw>http://www.w3.org/RDF/Validator/
01:42:09 <dc_rdfig> G: http://www.w3.org/RDF/Validator/ from AaronSw
01:42:18 <AaronSw> G:|W3C RDF Validation Service
01:42:18 <dc_rdfig> titled item G
01:42:33 <AaronSw> G:A Web interface to [ARP|http://www-uk.hpl.hp.com/people/jjc/arp/]
01:42:34 <dc_rdfig> commented item G
01:42:42 <deltab> seems less ugly overall than rewriting urls to a template that includes the article twice, with conditional parts...
01:43:01 <AaronSw> G: barstow, it'd be nice if [SiRPAC|http://www.w3.org/RDF/Implementations/SiRPAC/] pointed here saying that ARP was better for validation...
01:43:01 <dc_rdfig> commented item G
01:43:18 <Morbus> yeah...
01:43:35 <Morbus> and of course, it wholly invalidates the need for an article on how to do something simple in a complicated way <g>
01:43:56 <deltab> hehe
01:44:13 <deltab> that happens to me a lot
01:44:15 <Morbus> but hey, at least we figured out the complicated way. that was fun <g>
01:44:34 <deltab> yeah, good learning experience
01:44:45 <Morbus> indeed.
01:45:04 * em returns and catches up with todays log
01:45:10 <Morbus> welp, i'll bb in 15.
01:45:14 <Morbus> Morbus is now known as Morbus-Shower
01:45:24 * AaronSw waves to em
01:45:36 <em> hey! welcome back :)
01:45:40 <deltab> em: much of it's me and Morbus talking about apache
01:46:10 * sbp waves to Aaron, em, Morbus, and deltab
01:46:12 <em> yes... looks like it.... hmmm, danc chumped the p3p visualzation
01:46:18 <AaronSw> back: well, i'm on vacation, sneaking in my Internet fix tonight
01:46:32 <em> and looks like AaronSw chumped the Validator
01:46:49 <em> anyone chump the webont group announcement yet?
01:46:56 <AaronSw> danc did
01:47:05 <AaronSw> but i can't seem to find it in the logs...
01:47:07 * em applauds AaronSw's addiction :)
01:47:34 <em> ahh.. i see it, good!
01:47:37 <deltab> we now return you to regular rdf programming
01:47:44 <AaronSw> hmm, must've been while logger was down
01:48:42 <em> apache Handler Use ... interesting.. deltab, are you working on this?
01:48:56 <deltab> I chumped it by accident, actually
01:49:18 <deltab> I did the same thing with sbp's contact details the other day
01:49:33 * em waves to sbp
01:49:41 * sbp waves back at em
01:49:43 <deltab> I really have to learn not to paste raw urls in this channel
01:50:09 <sbp> I kept doing that for example URLs...
01:50:20 <em> yeah... i think some sort of chump indicator would be nice
01:50:24 <em> :http:///...
01:50:26 * deltab nods
01:50:35 <AaronSw> ... or -> http://
01:50:41 <AaronSw> -> http://foo.org
01:50:43 <AaronSw> which scribey does
01:50:49 <deltab> or chump http:...
01:51:02 <sbp> yeah, "chump" seems good
01:51:12 <AaronSw> chump violates my made-up-on-the-spot IRC UI principle: don't screw with the conversation
01:51:35 <sbp> Oh... er... O.K.
01:51:59 <deltab> the word or the program?
01:52:08 <AaronSw> the program, as it stands now
01:52:15 <AaronSw> the "chump" proposal is fine
01:52:38 * sbp wants people to start writing these damn axioms down in one place
01:52:43 * em wonders if the dotGnu people have been around lately...
01:53:06 <sbp> no, but I hang about in #dotgnu
01:53:24 <AaronSw> BLURB: Principles of Usabe IRC Bot UIs
01:53:24 <dc_rdfig> H: Principles of Usabe IRC Bot UIs from AaronSw
01:53:49 <AaronSw> H: Principle 1: Don't interfere with the conversation (act only when called upon)
01:53:49 <dc_rdfig> commented item H
01:54:14 <deltab> being able to fix spelling would also be nice :-)
01:54:14 <sbp> thank you
01:54:17 <AaronSw> H:|Principles of Usable IRC Bot UIs
01:54:17 <dc_rdfig> titled item H
01:54:43 <AaronSw> H:Principle 2: Don't allow users to make irreversable mistakes
01:54:43 <dc_rdfig> commented item H
01:55:01 <em> sbp, are the primary dotGNU discussions taking place on lists or #dotgnu channel?
01:55:05 <AaronSw> shoot, i'm not doing well with speeling today...
01:55:14 <sbp> definitely on the lists...
01:55:25 * em aaargggsss...
01:55:26 <sbp> I'm subscribed to a couple of them too
01:55:36 <sbp> they're quite high volume
01:55:47 * em is definitly not looking for more email traffic...
01:55:51 <Morbus-Shower> Morbus-Shower is now known as Morbus
01:56:07 * Morbus waves to everyone who waved at him
01:56:07 <em> anyone have a good RSS 1.0 portlet environment yet?
01:56:12 * sbp finds the new DAML thing available in PDF over FTP, and decides not to bother
01:56:54 <deltab> sbp: I could convert it if you like
01:56:59 * em would really like a 'interested in, but not interested enough to subscribe to lists' RDF/RSS in a box portal environment
01:57:03 <Morbus> em, what do you mean by portley?
01:57:13 <Morbus> ah.
01:57:29 <em> sorry Morbus... too much Jetspeed on the brain
01:57:34 <sbp> Cool new WebONT group, BTW...
01:57:41 <em> :)
01:57:43 <ASwartz> ASwartz is now known as AaronSw
01:58:01 <em> thanks sbp
01:58:03 * sbp wonders who's going to join that, but thinks he can predict
01:58:18 <sbp> it's another one of those much needed long time coming SW things :-)
01:58:24 <em> i think it will be interesting to find out...
01:58:57 <AaronSw> em: have you tried Peerkat?
01:59:08 <AaronSw> ... and http://ilrt.org/discovery/rdf/
01:59:31 <AaronSw> and i hope to get an rss->email gateway working soon
01:59:39 <Morbus> didn't xml tree have that once?
01:59:41 <em> I've become a very big fan of just a few of the DAML properties... equalsTo, unique and unambiguous serve a tremendous amount of my current needs
01:59:59 * sbp wonders what "www-ret-logic" is in the WebONT charter...
02:00:00 <AaronSw> xmltree: yeah, but i think it's gone now and it wasn't very good
02:00:16 <Morbus> yeah, when i tried it, i was either getting nothing, or buried. quite fickle.
02:00:35 <Morbus> incidentally, that's been a feature request for amphetadesk. "load the page, send an email"
02:00:36 <sbp> :Sean :presumes { :em :means { :equalsTo daml:equivalentTo daml:equivalentTo } } .
02:00:42 <em> AaronSw, Peerkat, no...
02:00:46 <em> recommend it?
02:01:00 <em> yes to sbp
02:01:15 <AaronSw> Peerkat: well, I'm biased since Rael and I are the developers.... ;-)
02:01:19 <sbp> :em :replies { :Yes :to :sbp } . :-)
02:01:27 <Morbus> i'd play with it, yeah.
02:01:36 <em> ready for prime time?
02:01:44 <Morbus> requires python installed, and still early release, but good enough.
02:01:52 <em> prime == not somethign I have to debug in order to get to work
02:01:58 <AaronSw> yes to that
02:02:16 <AaronSw> it's definitely usable, even on some weird systems like Mac and Solaris
02:02:22 * AaronSw knows... he
02:02:26 <deltab> www.w3.org needs a bit of reconfiguring - it thinks the .dot files it serves are MS Word templates
02:02:26 * AaronSw knows... he's been doing the debugging
02:02:35 <AaronSw> heh!
02:03:06 <sbp> Good ol' w3.org
02:03:26 <sbp> It won't put the CVS Ids in my N3!
02:05:25 <sbp> Ugh, gotta run
02:07:20 * deltab reads w3.org's aural style sheet, and tries to imagine what it sounds like
02:11:31 <danbri> dc_rdfig
02:11:35 <AaronSw> Hi danbri
02:11:35 <danbri> dc_rdfig:view
02:11:36 <dc_rdfig> D: RDFWeb / Friend of a Friend "co-depiction" demo (http://swordfish.rdfweb.org/discovery/2001/08/codepict/)
02:11:36 <dc_rdfig> E: an RDF model of P3P (http://www.w3.org/2001/08/14-p3p/)
02:11:37 <dc_rdfig> F: Apache's Handler Use (http://httpd.apache.org/docs/handler.html)
02:11:38 <dc_rdfig> G: W3C RDF Validation Service (http://www.w3.org/RDF/Validator/)
02:11:39 <dc_rdfig> H: Principles of Usable IRC Bot UIs (blurb)
02:12:00 <danbri> G:done. kinda.
02:12:01 <dc_rdfig> commented item G
02:12:55 <danbri> G:I'd like to find a space to pay tribute to the role the (now ageing) SiRPAC parser played in RDF history. This'll do for starters! SiRPAC was critical in getting the first wave of RDF apps off the ground; we all owe a dept to Janne.
02:12:55 <dc_rdfig> commented item G
02:14:59 <AaronSw> G:I agree, SiRPAC is very important... however, there have been some problems with folks putting out bad RDF and saying "SiRPAC said it was fine"
02:14:59 <dc_rdfig> commented item G
02:18:03 <danbri> G:I agree, the SiRPAC codebase was showing its age, and ARP is a fine successor (for Java-heads).
02:18:03 <dc_rdfig> commented item G
02:20:22 <em> G: Notes the fact that Janne implemented SiRPAC from the spec only... it was an interesting test to see if people could develope to the RDF spec independant of working with members of the working group...
02:20:22 <dc_rdfig> commented item G
02:21:08 <danbri> hey eric!
02:21:24 * danbri remembers giving Janne grief about anon nodes, and literals-vs-resources.
02:22:01 * em waves to danbri :)
02:23:14 <danbri> D:Aaron and DanBri play at finding paths through the RDFWeb co-depiction graph...
02:23:14 <dc_rdfig> commented item D
02:26:32 <Morbus> on a side note, the default disobey and gamegrene channels are rss 1.0 from logicerror/blogifyYourPage
02:26:59 <danbri> D:From [libby|http://swordfish.rdfweb.org/discovery/2001/08/codepict/codepict.jsp?mbox=mailto:libby.miller@bristol.ac.uk] to [aaron|http://swordfish.rdfweb.org/discovery/2001/08/codepict/codepict.jsp?mbox=mailto:me@aaronsw.com], via [jan|http://swordfish.rdfweb.org/discovery/2001/08/codepict/codepict.jsp?mbox=mailto:jan.grant@bristol.ac.uk], [dave|http://swordfish.rdfweb.org/discovery/2001/08/codepict/codepict.jsp?mbox=mailto:dave.beckett@bristo
02:27:00 <dc_rdfig> commented item D
02:27:51 <danbri> D:...and [emiller|http://swordfish.rdfweb.org/discovery/2001/08/codepict/codepict.jsp?mbox=mailto:em@w3.org]. Maybe we should dump this all in prolog/n3 for logic geeks to play with?
02:27:51 <dc_rdfig> commented item D
02:28:37 <danbri> D:oh, I messed up the blog hyperlinks, but you get the idea.
02:28:37 <dc_rdfig> commented item D
02:32:24 <Morbus> yup, welp if you don't know the yahoogroups foramt, its described in the "finding more" doc linked in tiny letteers on the add a channel page
02:33:34 <AaronSw> ?
02:33:40 <AaronSw>http://www-106.ibm.com/developerworks/webservices/library/ws-phtt/
02:33:40 <dc_rdfig> I: http://www-106.ibm.com/developerworks/webservices/library/ws-phtt/ from AaronSw
02:33:54 <AaronSw> I:"Reliable HTTP" - HTTPR
02:33:54 <dc_rdfig> commented item I
02:34:05 <AaronSw> I:|A Primer for HTTPR
02:34:06 <dc_rdfig> titled item I
02:34:22 <AaronSw> I:Designed in part for business-critical web service stuff
02:34:22 <dc_rdfig> commented item I
02:46:02 <Morbus> actually, that deals little with rdf, but just hovers around a possible solution
02:48:44 <AaronSw> I:[Mark Baker responds|http://lists.w3.org/Archives/Public/www-ws/2001Jul/0027.html]: """"request" appears to be nothing more than a means of encoding an RPC style method invocation mechanism over HTTP POST. This doesn't fit well with the hypertext transfer application semantics that HTTP defines."""
02:48:44 <dc_rdfig> commented item I
02:50:48 <danbri> danbri has changed the topic to: Semantic Web Vapourware for the masses, http://www.w3.org/RDF/Interest/#irc
02:51:36 <danbri> the return of interesting /topics, which I intend to rule over crabbily and hypocritically as whimsy takes me :)
02:51:49 <Morbus> heh.
02:51:56 <AaronSw> AaronSw has changed the topic to: Semantic Web Chat: Vapourware for the masses ( http://www.w3.org/RDF/Interest/#irc )
02:52:12 <danbri> less Chat more Hack
02:52:13 * AaronSw keeps a whole list of 'em in case of emergency
02:53:04 <AaronSw> ooh, that's a good one
02:54:03 * AaronSw notes that he better go soon
02:54:13 <Morbus> anyone know much about proxies? i'm looking for header transcripts of auth and nonauth proxies.
02:54:46 <AaronSw> HTTP proxies?
02:54:50 <Morbus> yeah.
02:55:02 <Morbus> i'm sure how to handle nonauth'd, but i follow examples better than rfc's.
02:55:08 <AaronSw> I hacked one... http://logicerror.com/archiverProxy
02:55:11 <Morbus> i understand that i mime64 un:pw for auth'd proxies?
02:55:12 <AaronSw> it's in Python
02:55:37 <AaronSw> s/mime64/base64
02:55:42 <AaronSw> that's for basic auth, irrelevant of proxies
02:55:48 <Morbus> heh. yeah. thats what i meant.
02:56:45 <Morbus> bookmarked.
02:56:54 <AaronSw> there's some running Python code in urllib or somewhere nearby
02:57:01 <deltab> moprbus: you could set one up in apache, and watch what happens
02:57:16 <Morbus> my prob is i know absolutely nothing about proxies.
02:57:19 <AaronSw> use panelPatch -- it's very cool for those things
02:57:30 <AaronSw> altho a pain to get working
02:59:05 <AaronSw> OK, time to go... c'yas later
03:13:58 <Morbus> welp i gotta get outta here. i'll see everyone later
03:50:18 <danbri> danbri has changed the topic to: Semantic Web Vapourware for the masses; Interest Group chat, http://www.w3.org/RDF/Interest/#irc
12:18:23 <bijan> Morning libby.
12:18:41 <libby> hi bijan
12:18:48 <libby> afternnon!
12:19:03 <libby> how you doing?
12:19:08 <bijan> Ok! And you?
12:20:05 <libby> good thanks. sleep cycle's all messed up
12:20:15 <libby> how's semantic chimp going?
12:20:25 <bijan> Heh. I which I *had* a sleep cycle to mess up!
12:20:51 <bijan> Ok. Some bugs, some missing features. we're trying to get a release together for this week.
12:22:01 <libby> neat :)
12:23:01 <libby> anootations have been driving me nuts recently
12:23:45 <libby> it seems to me that since you can say stuff about the document items are contained in, then everything in the document should be reified
12:23:59 <libby> or at least the document source should be get-at-able through an api
12:24:05 <bijan> Hmm.
12:24:27 <libby> maybe this is similar yto what sbp was saying
12:24:34 <bijan> The question, really, is how explicit do you want the annotantionness to be.
12:25:09 <bijan> I mean, what reification buys you is explicitness.
12:25:46 <libby> well, what I decidede on for codipiction thing was something like http://swordfish.rdfweb.org/photos/rdf/000477.rdf
12:26:08 <libby> where you gave rdf:about="", creator=libby
12:26:20 <libby> and then talk about other objects within the document
12:26:34 <bijan> Yes.
12:27:00 <libby> this seems to me to be very explicit _if_ you can talk about the document's relationship to the triples inside it.
12:27:09 <libby> I'm sure people ahve takled about ythis before
12:27:29 <bijan> No :)
12:27:33 <bijan> except you!
12:27:48 <bijan> so the annotator is valid for the whole document?
12:27:57 <bijan> I can't tell how many annotations...
12:27:58 <bijan> ah.
12:28:00 <bijan> I see.
12:28:06 <libby> I'm wondering whether the problem is an implementation one....the stanford api (and jena, and mine/danbris implementation and redland) don't have provenence
12:28:09 <bijan> There's one annoation?
12:28:09 <libby> hehe
12:28:15 <libby> yeah one
12:28:20 <bijan> This is actually very chumplike.
12:28:29 <libby> but poeple using SQL databases usually do ahve provenence
12:28:41 <libby> but...the api has come from the non-provenenace end...
12:28:46 <libby> is it?
12:28:56 <libby> cool, then maybe I'm on the right track
12:29:02 <libby> :)
12:29:08 <bijan> Yes. this is the kind of thing you have to do to avoid reification.
12:29:25 <bijan> So, you have an item (in the chump)
12:29:31 <libby> uhuh
12:31:03 <bijan> which has a chump:target in it.
12:31:11 <libby> right
12:31:17 <bijan> Which is like your an:annotates.
12:31:22 <libby> yup
12:31:26 * danbri catches up
12:31:30 <bijan> Hey dan.
12:31:41 <bijan> (Though I like an:annotes better :)
12:31:41 <libby> hello
12:31:51 <danbri> hi
12:32:11 * libby doesn't care what it's calledd :)
12:32:13 <bijan> And then you have chump:comments
12:32:50 <bijan> Which contain description who's rdf:about is often the rdf:about of an:annotates.
12:32:55 <bijan> Yes, it's exactly the same.
12:32:56 <danbri> sum total of my knowledge right now: (1) all RDF is annotational (cos you're writing about the same set of things); (2) explicit reification is nearly always a pain in the butt
12:33:07 <bijan> Boo to 1
12:33:10 <bijan> Yes to 2
12:33:19 <danbri> howso?
12:33:30 <bijan> Well, I don't understand 1 :)
12:33:43 <bijan> "cos you're writing about hte same set of things"???
12:34:01 <libby> bijan do you have an e.g. of chump annotations?
12:34:34 <danbri> 1: It seems to me to be hard to draw a circle around those RDF apps that are 'annotations'. Forzample, restaurant reviews. "there is a restaurant; it has a homepage of ...; it has a yum-rating of ...; etc".
12:35:10 <bijan> Well, the rss in monkeyfist.com/SemantiChimp/
12:35:15 <libby> (1) might be true but you vcab't do interesting stuff - writing about objects and their relationships withoput either reification or being abole to group things somehow
12:35:27 <libby> thanks
12:35:46 <danbri> What makes an rdf app "annotational" seems to be a concern for making the provenance / who said what aspect clear at the data level, so apps w/out reification and provenance facilities can make use of different perspectives without geting confused.
12:36:44 <danbri> grouping: yup. A lot of what various of us are doing as "annotations" is a way of providing that sort of grouping.
12:36:56 <bijan> hmm. call me conservative but I tend to think of "annotations" narrowly: i.e., small comments on a text designed to be read with the text provide comment, explication, etc.
12:37:30 <libby> I kinof think of them as broader, but I'm pretty sure the issues are the same
12:37:37 * DanC_ waves from 1st telcon of the day
12:37:49 <bijan> Poor DanC!
12:37:53 <libby> wa
12:37:57 <libby> aw, I mean
12:38:14 <libby> well in fact they're mnot the same if the comment is just text, but they are if the comment is RDF
12:39:24 <danbri> that's sooo last year. I guess I have a few apps in mind: (1) medical content rating, eg x isCriticalReviewOf y; x isSupportingEvidenceFor y; here, the stuff you're putting in small textual comments is blown up into another 1st class document. (2) reviews of products and services, where its the thing _described by the target document_ not the doc itself that is of interest; eg. a review is of a restaurant, not of its homepage (typically).
12:39:29 <bijan> Well, by "A text" I surely include RDF documents and statments. And of course I think comments can be RDF.
12:39:59 <bijan> er...I don't see why 1 *Wouldn't* be annotative, by my standard.
12:40:01 <danbri> in which case you're on the slippery slope to generalism: everything's an annotation...
12:40:02 <libby> yeah, I know :)
12:41:11 <danbri> (1) is, but the confusion (in my mind anyway) is between the assertion that 'x isCriticalReviewOf y' being the annotaion, versus 'x' being the annotation. Of course x (and/or y) might contain the assertion...
12:41:42 * bijan has logically cleats...being on slippery slopes doesn't faze *me*.
12:41:50 <bijan> er..logical cleats.
12:41:53 <libby> heh
12:41:59 <danbri> :)
12:43:01 <libby> so, danbri, the difference between making the connection between 2 documents or a doc and an object and writing sdomething which is about but makes no reference to the object?
12:43:23 <bijan> I'd say that 'x' *and* 'x isCriticalReviewOf y' are annotative, perhaps part of a single annotation.
12:43:30 <libby> or, even if it makes a reference to the object, the connecting itself can have different qualities ....I think I get it
12:43:57 <libby> yeah, either is annotative.
12:44:26 <bijan> "annotation" is interest senstive and context relative, to an extreme.
12:44:35 <danbri> I found thinking of 'annotation' as a word for a type of document or information too slippery; thinking about an 'annotation' as a social/descriptive act slightly less so.
12:44:46 <bijan> But I think in most contexts, for the greater interest, it's better to be a *leetle* restrictive :)
12:44:59 <libby> heh
12:45:43 <libby> from a practical point of view, they key thing as you said it holding onto who said what. that's all that matters ovcer these different types of annotations
12:46:05 <danbri> restrictive, perhaps. I'm not being frivolously picky though. IMHO most of the more interesting SW apps that feel annotational aren't in the vein of the 'textual comment on document subsection'.
12:46:31 <bijan> well, i *didn't* say *textual* comment. I said "comment on a text" :)
12:47:16 * bijan points out that the *whole* point of SemantiChimp is to allow RDF annotations (as opposed to Annotea which seems to only allow textual ones.)
12:47:34 <libby> and it's great :)
12:47:40 <danbri> who said what: yes. I want to be able to mutter restaurant reviews into a mobile phone, or take pictures of fixed places on the planet, and find them again afterwards. I don't care really whether they're considered annotations, but I think they'll be built using pretty much the same techniques (modelling/database style) as 'annotations'.
12:47:52 <danbri> bijan: yup, that's why its interesting
12:48:17 <danbri> <bijan> Well, by "A text" I surely include RDF documents and statments. And of course I think comments can be RDF.
12:48:26 * bijan needs to look at chinese scholorship.
12:48:46 <bijan> They had annotation upon annotation.
12:49:29 <bijan> Actually, "who said it" can be relatively unimportant.
12:49:47 <bijan> I.e., assertional context may be important.
12:49:55 <bijan> But not actual authorship.
12:50:00 <libby> ok
12:50:13 <libby> the context is the important thing cos you can hang what you like off it
12:50:21 <bijan> I.e., you want to know that statment foo is an annotation rather than a part of the actual text.
12:51:12 <libby> right, the grouping of the annotation statements or stuff
12:51:51 <libby> I feel like I've been going round and round with this stuff for years :(
12:52:02 <bijan> Yep! :)
12:52:30 <bijan> In rdf, IMHO, the tricky bit is getting useful names without forcing an explicit reification.
12:53:10 <libby> I haven't found a good example of how you group them yet bijan
12:53:12 * bijan things of reification as *so* low level.
12:53:16 <libby> probaboy not looking hard enough
12:53:20 <bijan> Hmm?
12:53:24 <libby> ugh, horrible
12:55:16 <bijan> Sorry!
12:55:23 <libby> I mean, do you ahve an example of talking about RDF statement-type things to hand
12:55:35 <bijan> Ah! No.
12:55:37 <libby> no, reification is horrible..as we all know
12:55:45 <bijan> But I know how it would go in c humpland.
12:55:54 <libby> yeah? tell!
12:56:01 <bijan> If you notice, each comment has a uri.
12:56:12 <libby> yep
12:56:15 <bijan> documenturi/A1 or some such.
12:56:21 <libby> uhuh
12:56:30 <bijan> So I'd just pop that uri in the chump:content rdf:about.
12:57:01 <bijan> So the chump:comment stream is flat.
12:57:14 <bijan> You can't tell what they are comments *on* until to peek inside.
12:57:22 <libby> <chump:contentAbout rdf:about?
12:57:23 <bijan> They are commetns *of* their containing item.
12:57:28 <bijan> Yes.
12:57:47 <bijan> Essentially, I invented my own reification :)
12:57:54 <libby> heh, goog for you
12:57:59 <libby> s/goog/good
12:58:02 <bijan> Or rather I annotate the comment :)
12:58:07 <bijan> Er..the comment's content :)
12:58:22 <libby> cool. thanks bijan
12:58:30 <bijan> No probl.
12:58:40 <bijan> I antcipate *great* things of this for the chimp.
12:58:43 <bijan> Correctoins!
12:58:59 <bijan> :A1 chump:correction "New text"
12:59:10 <bijan> Delection!
12:59:32 <bijan> :A1 chump:visibility "false"
12:59:32 <libby> wow
12:59:45 <bijan> Er.. deletion that is.
12:59:47 <bijan> Delection!
12:59:54 <libby> ;)))
13:00:01 <bijan> :A1 chump:taste chump:delectable
13:00:18 <bijan> :)
13:00:31 <libby> heh, funny
13:00:58 <bijan> Designwise, what makes it all work is my being somewhat careful about the URIs.
13:01:22 <bijan> My app relies a bit on info encoded in the uris.
13:01:43 <bijan> Which is fine, IMHO.
13:01:48 <libby> right, I was wondering about that - F0, F03 etc
13:01:54 <bijan> Better to "cheat" a little than to have a horrid mess :)
13:02:03 <libby> have you writtne any of this stuff up anywhere bijan?
13:02:15 <bijan> Hmm. Maybe.
13:02:32 <libby> go on, write it up
13:02:37 <bijan> Oh, I plan too.
13:02:41 <libby> cool
13:02:42 <bijan> It'll be part of the release.
13:02:54 <libby> excellent. I look forward to it
13:04:16 <libby> see you later guys
13:04:47 <bijan> Ta!
13:48:57 * danbri crossposts to RDF-Interest and RDF-Logic about crossposting (and msg temparature)
13:49:08 <bijan> Spammer.
13:49:14 <danbri> now waiting to see how long it takes before someone
13:49:22 <danbri> ...accuses me of spamming :)
13:49:34 <danbri> thanks Bijan; if it had to be anyone, I'm glad it was you!
13:50:02 <bijan> My pleasure is service.
13:50:46 * danbri hopes Web Services leave a slot for such niceties in the protocol
14:04:19 <DanC_> msg temp? for example? what's up?
14:07:03 * bijan notes that the Timeline on http://www.w3.org/RDF/ seems a bit dated.
14:07:25 <DanC_> the /RDF/ page isn't very shiny, true.
14:08:28 <danbri> msg temp: a few messages recently seemed rather crabby / confrontational in tone. I don't want to pick at specific posters here.
14:08:57 <DanC_> I browsed the archive. you picked on a specific one by way of the In-Reply-To header field ;-)
14:09:18 <danbri> ;-)
14:10:10 <DanC_> speaking of crossposting: I was wondering whether to post the WebOnt announcement to rdf-interest as well...
14:10:46 <DanC_> ... but then I thought if it goes to general-purpose RDF lists, it should go to general-purpose XML lists too...
14:10:51 <danbri> Generally I'm happy seeing announcements go to wide readership.
14:10:59 <bijan> And that way lies madness!
14:11:22 <DanC_> ... then I thought I'd ask you (danbri) for advice, then I said, aw screw it, and just posted to www-rdf-logic, an audience to whom I am clearly obliged.
14:11:29 <danbri> Dare you to forward it to XML-DEV with a "thought you folks might be interested in this new XML-based Schema language from W3C..."
14:12:04 <DanC_> I haven't figured out how to get authorization to post to xml-dev without filling up my inbox.
14:12:37 <DanC_> (yes, that's lame)
14:27:59 <bijan>http://lambda.weblogs.com/discuss/msgReader$1673
14:27:59 <dc_rdfig> J: http://lambda.weblogs.com/discuss/msgReader$1673 from bijan
14:28:34 <bijan> J: | VML, ML extened with "hypothetical views"
14:28:34 <dc_rdfig> commented item J
14:29:15 <bijan> J: Via the [Lambda the Ultimate|http://lambda.weblogs.com/] weblog.
14:29:15 <dc_rdfig> commented item J
14:30:06 <bijan> J:"Operationally, a hypothetical view is a value with a representation that indicates how the value was created."
14:30:06 <dc_rdfig> commented item J
14:30:14 <bijan> J: Sounds familiar :)
14:30:14 <dc_rdfig> commented item J
14:34:16 <DanC_>http://www.oasis-open.org/committees/relax-ng/tutorial-20010810.html
14:34:17 <dc_rdfig> K: http://www.oasis-open.org/committees/relax-ng/tutorial-20010810.html from DanC_
14:34:26 <DanC_> K:|RELAX NG Tutorial
14:34:26 <dc_rdfig> titled item K
14:47:30 <lilo> [GlobalNotice] Hi all. Services will be coming back in a moment and you'll need to reidentify. It will be accompanied by a split.
14:50:27 <ChanServ> ChanServ has changed the topic to:
14:50:30 <ChanServ> This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome
14:51:33 <dajobe> dajobe has changed the topic to: Its Triples all the way down (or Quads, whatever) Semantic Web chat http://www.w3.org/RDF/Interest/#irc
16:05:04 <DanC_> D:I keep looking for the "view source" button/link. e.g. where's the actual query that's used? can I get the database in RDF/xml form?
16:05:04 <dc_rdfig> commented item D
16:08:20 <dajobe> hmm, the chump didn't rotate last night
16:08:38 <DanC_> umm.. where are the ftf minutes? I can't find them from http://www.w3.org/2001/sw/RDFCore/20010801-f2f/
16:10:51 <DanC_> ah.. http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Aug/att-0026/01-Overview.html
16:12:47 <deltab> F:They're used for processing files before output - useful for special markup, or adding headers and footers automatically
16:12:47 <dc_rdfig> commented item F
16:33:12 <tav> tav is now known as tav|talking_with_the_parrots
16:35:20 <libby> D: now you can look at the [list of urls used|http://swordfish.rdfweb.org/discovery/2001/08/codepict/scutterplan.txt]
16:35:20 <dc_rdfig> commented item D
16:35:31 <libby> D: the query used is at the bottom of the page
16:35:31 <dc_rdfig> commented item D
16:41:29 <deltab> what's happened to http://rdfig.xmlhack.com/ ?
16:41:45 <dajobe> what do you mean?
16:42:02 <deltab> The requested URL /index.html was not found on this server.
16:42:12 <dajobe> work for me
16:58:48 <tav|talking_with_the_parrots> tav|talking_with_the_parrots is now known as tav
16:59:33 <tav> works here too...
17:29:26 <dajobe> DanC_: what are Prince nodes?
17:29:48 <DanC_> anonymous nodes.
17:30:30 <DanC_> my view: existentially quantified variables. but the WG hasn't decided exactly what they are yet, other than there's one in <rdf:Description><prop>value</prop></rdf:Description>
17:30:40 <dajobe> is "Prince nodes" a well known term for it? Never come up anytime we have been talking about it so far
17:31:05 <DanC_> "Prince nodes" was not known to me before I read the meeting minutes.
17:32:15 <DanC_> ah... [editors note: I like the term suggested by Eric Miller -
17:32:15 <DanC_> prince nodes. I shall use that term for the rest of these minutes
17:32:15 <DanC_> where it seems appropriate]
17:32:20 <DanC_> -- http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Aug/att-0026/01-Overview.html
17:43:31 <dajobe> re chump not rotating ids at GMT midnight; that is a bug if first thing after midnight is an annotation not a URI
17:57:23 <nmg> I've not been following #rdfig for the last month or so - has there been any discussion on data smushing/resource folding that goes beyond danbri's notes from early this year?
17:59:16 * nmg wasn't able to find anything by searching the logs, but didn't know if a new term for it was now in favour...
18:00:19 <bijan> Not much, AFAIK.
18:00:56 <bijan> I've been thinking a bit about it from an application perspective...e.g., learning strategies.
18:01:41 <bijan> I remember chumping a paper about the difficulties of mereging knowledge bases with different inference engines.
18:02:04 <bijan> And there's some Loom papers on large scale distributed ontology merging.
18:03:04 <nmg> I've been applying heuristics to try and spot when two resources are likely to be the same.
18:03:30 <nmg> for example, are mailto:nmg@ecs.soton.ac.uk and mailto:nick@totl.net the same person?
18:05:02 <ambient> rather, do they belong to the same person?
18:05:08 <theran> nmg: do you have training data?
18:05:27 <nmg> currently, I'm having to rely on bits of contextual information (do they copublish with the same author, etc) which give an indication that they're likely to be the same, but which don't can't give a concrete answer.
18:06:35 <nmg> ambient: yes, effectively. if you go down the road of specifying resources in terms of their properties (ie. anonymous nodes) you have much the same problem.
18:07:21 <nmg> theran: I've been merging RDF dumps of my dept's personnel db and our publications db.
18:07:53 <nmg> there were already consistency problems between the two sources, but I think that these are exactly the problems that SW systems will have.
18:08:01 <theran> nmg: right. and you want to build a classifier.
18:09:51 <nmg> I suppose you could describe it as a classifier, yes, but you don't necessarily have an a priori list of the classes you're trying to group resources into (ie. a list of unique people)
18:15:39 <theran> nmg: are you trying to learn what the features are as well?
18:16:48 <nmg> no - still taking small steps here
18:16:53 <theran> nmg: I'm a bit out of my depth here, but if you had features (properties), then couldn't you use something like k-th nearest neighbor?
18:19:16 <theran> nmg: but in that case, you'd still be inducing a decision boundary of some sort.
18:20:10 <nmg> yes, but each feature could potentially be more complex than a simple triple (eg. has authored a paper which was published in a certain journal), so you'll have an extremely large feature set to work with
18:20:49 <theran> sure. though usually there are some features that count a great deal more. (that's why I asked if you wanted to learn them)
18:35:26 <nmg> thanks - that's something to think about
18:35:29 <nmg> \quit
18:35:35 <nmg> d'oh
18:39:06 <DanC_>http://www.webservices.org/
18:39:07 <dc_rdfig> L: http://www.webservices.org/ from DanC_
18:39:15 <DanC_> L:|Webservices.org
18:39:15 <dc_rdfig> titled item L
18:39:24 <DanC_> L:nifty web services portal
18:39:24 <dc_rdfig> commented item L
18:39:38 <DanC_> L:http://conferences.oreilly.com/p2p/
18:39:38 <dc_rdfig> commented item L
18:39:41 <DanC_> oops...
18:39:43 <DanC_>http://conferences.oreilly.com/p2p/
18:39:43 <dc_rdfig> M: http://conferences.oreilly.com/p2p/ from DanC_
18:40:02 <DanC_> M:The O'Reilly Peer-to-Peer and Web Services Conference
18:40:03 <dc_rdfig> commented item M
18:40:16 <DanC_> M:Washington, D.C. -- September 18-21, 2001
18:40:16 <dc_rdfig> commented item M
18:40:22 <DanC_> M:|The O'Reilly Peer-to-Peer and Web Services Conference
18:40:22 <dc_rdfig> titled item M
18:40:40 <DanC_> L:who runs this thing? I can't find an "about..." page
18:40:40 <dc_rdfig> commented item L
21:03:27 <dajobe> hi AaronSw
21:04:24 <AaronSw> hi there
21:05:08 <dajobe> I've been having a good coding day - squashed loads of Raptor bugs and threw out code for containers etc.
21:06:13 <AaronSw> Great!
21:09:04 * AaronSw waves to danbri
21:09:12 <danbri> hello!
21:15:46 <oierw|france> oierw|france is now known as oierw
21:18:29 <DanC_> danbri, didja see that P2P/web services item I chumped above?
21:19:08 <danbri> I knew about the ORA conference; the webservices.org thing I found much more interesting. Slashdot for SOAP enthusiasts...
21:20:22 <danbri> battle of the metaphors: SW vs P2P vs WS :)
21:20:48 <dajobe> rdfweb is a person-to-person semantic web service
21:25:34 <danbri> yeah, got my bases covered ;-)
21:28:11 <DanC_> rdfweb got a whole lot cooler with the codepiction demo.
21:28:36 <DanC_> it inspired me to (feel guilty about not) updating my smart-home.rdf file
21:28:42 <AaronSw>http://lists.w3.org/Archives/Public/www-archive/2001Aug/0024.html
21:28:42 <dc_rdfig> N: http://lists.w3.org/Archives/Public/www-archive/2001Aug/0024.html from AaronSw
21:28:49 <AaronSw> N:|Aaron's Thoughts on Web Services
21:28:49 <dc_rdfig> titled item N
21:29:03 <AaronSw> N:"""These are some excerpts to a response to questions posed by Clay Shirky as part of a research report he is doing for O'Reilly on Web Services."""
21:29:03 <dc_rdfig> commented item N
21:29:27 <AaronSw> N:I basically claim that Web Services is merely a facet of the Semantic Web
21:29:27 <dc_rdfig> commented item N
21:31:33 * DanC_ is afraid that GET is going to disappear with the Web Services wave.
21:31:51 <DanC_> I was heartened to see REST vs. RPC... somewhere... was it on webservices.org?
21:32:01 <AaronSw> REST vs. RPC was on FoRK
21:32:09 <AaronSw> yes, it was good to see a convert ;-)
21:32:32 <DanC_> who got converted? I think I only saw a pointer to the fork discussion
21:32:34 <danbri> It could help; people take GET for granted. By reminding the world that there are other things a URI-named resource can do, GET's selling points may come back into focus.
21:33:00 <AaronSw> converted: Jeff Bone, I think... just a sec
21:33:11 <AaronSw>http://xent.com/pipermail/fork/2001-August/002801.html
21:33:12 <dc_rdfig> O: http://xent.com/pipermail/fork/2001-August/002801.html from AaronSw
21:33:19 <AaronSw> O:|REST, RPC, mountains, molehills, and a retraction (sort of)
21:33:19 <dc_rdfig> titled item O
21:33:24 <danbri> Well, so long as popular SOAP tools support GET-encoded method calls, and WSDL etc allow one to say of a method that it has no side effects.
21:33:34 * DanC_ curses NS4 for lack of "search history titles" feature... thinks about resurrecting that browser-history daemon...
21:34:20 <danbri>http://www.urlblaze.com/
21:34:20 <dc_rdfig> P: http://www.urlblaze.com/ from danbri
21:34:30 <DanC_> inerconnectedness... Jeff Bone jbone@jump.net .. I used to work at HaL with one of the guys that started jump.net
21:34:35 <danbri> _P:URLBlaze, claim to be the first URL sharing network.
21:34:38 <dajobe> oierw: I found it on [http://www.oreillynet.com/weblogs/|O'reilly Network weblogs] from Tim O'Reilly - doesn't seem to be a URL for that particular blog
21:34:42 <dajobe> O:I found it on [http://www.oreillynet.com/weblogs/|O'reilly Network weblogs] from Tim O'Reilly - doesn't seem to be a URL for that particular blog
21:34:42 <dc_rdfig> commented item O
21:34:46 <danbri> P:|URLBlaze, claim to be the first URL sharing network.
21:34:47 <dc_rdfig> titled item P
21:35:18 <danbri> P:Kinda cheeky. They mean, they're using P2Pish techniques to share the contents of bookmark lists. I think.
21:35:18 <dc_rdfig> commented item P
21:36:08 <DanC_> shared bookmarks date back to WWW1 (1994) at least.
21:36:18 <danbri> quite!
21:36:27 <AaronSw> O: Aha, [TimO's thoughts|http://www.oreillynet.com/weblogs/date/2001/08/11]
21:36:27 <dc_rdfig> commented item O
21:41:37 <DanC_> frabnabbit. 404 at http://www.avogadro.com/ietf/http-futures/msg00006.html
21:42:08 <DanC_> wherein I said basically what J. Bone is saying: It turns out that the best argument for REST is this: *there are no
21:42:08 <DanC_> applications you can think of which cannot be made to fit into the GET
21:43:44 <DanC_> I also wrote on this topic in WebApps...
21:43:47 <DanC_> O
21:43:49 <DanC_> O:
21:43:50 <dc_rdfig>http://xent.com/pipermail/fork/2001-August/002801.html
21:43:50 <dc_rdfig> REST, RPC, mountains, molehills, and a retraction (sort of)
21:43:51 <dc_rdfig> (dajobe) I found it on [http://www.oreillynet.com/weblogs/|O'reilly Network weblogs] from Tim O'Reilly - doesn't seem to be a URL for that particular blog
21:43:52 <dc_rdfig> (AaronSw) Aha, [TimO's thoughts|http://www.oreillynet.com/weblogs/date/2001/08/11]
21:44:29 <DanC_> O:on coordination languages, see also [my WebApps essay of Mar/Apr 1997|http://www.w3.org/People/Connolly/9703-web-apps-essay.html]
21:44:29 <dc_rdfig> commented item O
21:54:31 <DanC_> hmm... here's one to watch... "I'm doing my part to make sure in all this RPC noise, that something REST-friendly emerges, viz a vis my bitching about
21:54:31 <DanC_> the SOAP 1.2 HTTP binding.
21:54:31 <DanC_> " -- Baker
21:54:42 <AaronSw> URI?
21:54:49 <AaronSw> Baker's a good guy
21:55:01 <DanC_> O:[Baker's reply|http://xent.com/pipermail/fork/2001-August/002806.html] is heartening.
21:55:01 <dc_rdfig> commented item O
21:58:39 * DanC_ asked (indirectly) for a GET (i.e. REST) related requirement for the XML Protocol requirements; goes to see if it got in there...
21:59:31 * AaronSw wishes it was a requirement of the WG to stay inline with Web Axioms
21:59:41 <AaronSw> such as REST/GET stuff
22:00:02 * AaronSw also sent a message to the list requesting a GET requirement for their work, but got no response from the WG
22:00:16 <DanC_> it is a requirement, in general; and it's up to the team (i.e. me, Hugo, ...) to see that WGs follow. that's why I'm looking.
22:00:27 <DanC_> no response: rasie holy hell.
22:00:47 <DanC_> i.e. wait a couple weeks or so, then phone hugo.
22:01:52 <DanC_> umm... I'm nervious... "GET" doesn't occur in http://www.w3.org/TR/xmlp-reqs/
22:02:08 <AaronSw> I sort of gave in after hearing all the talk about "rubber-stamping SOAP"
22:02:43 <DanC_> huh?
22:02:56 * DanC_ loses connection to MIT. crap
22:03:06 <DanC_> ah... it's back
22:03:16 <AaronSw> popular opinion seems to be that XMLP is just there to approve SOAP (perhaps with minor changes)
22:03:46 <DanC_> er... I don't recall popular opinion having much sway in your life before, AaronSw.
22:03:56 <AaronSw> :-)
22:04:03 <AaronSw> thankfully, it hasn't
22:04:07 <AaronSw>http://lists.w3.org/Archives/Public/xml-dist-app/2001Jan/0029.html
22:04:07 <dc_rdfig> Q: http://lists.w3.org/Archives/Public/xml-dist-app/2001Jan/0029.html from AaronSw
22:04:13 <DanC_> family time... (5pCT on Wednesdays)
22:04:17 <AaronSw> Q:|Integrating XP Into Web Architecture
22:04:17 <dc_rdfig> titled item Q
22:04:32 <AaronSw> Q:My request of the XMLP WG to follow the Web Axioms
22:04:32 <dc_rdfig> commented item Q
22:05:01 <AaronSw> Q:Written after reading the XP Reqs and not finding it
22:05:02 <dc_rdfig> commented item Q
22:06:30 <AaronSw> Q:In the end I got no response from the WG, and it ended up devolving into a discussion about how SOAP/XML-RPC always over POST were easier to use than preserving the GET/POST semantics
22:06:30 <dc_rdfig> commented item Q
22:07:30 <AaronSw> popular opinion: however, I didn't feel strong enough to fight against the whole WG (or at least most of it) on the issue
22:12:03 <bijan> Aaron, your strength is of ten because your heart is pure.
22:12:32 <AaronSw> Unfortunately, this Working Group is more like a hundred-strong.
22:12:44 * AaronSw emails Hugo about it
22:12:51 <bijan> Oh well then.Yer screwed.
22:14:46 <AaronSw> Well, DanC has given me a little more fighting power
22:15:12 <AaronSw> Q:I wrote [a message to Hugo today|http://lists.w3.org/Archives/Public/www-archive/2001Aug/0025.html] checking up on the issue
22:15:13 <dc_rdfig> commented item Q
22:17:32 <dajobe> AArong, could have added; "note you can GET this web service in order to read our discussion"
22:17:39 <dajobe> urg, AaronSw I meant
22:18:01 <AaronSw> :-)
22:18:16 <AaronSw> Heh, wonder how I missed that...
22:19:09 <AaronSw> GET: We put the :// in http://www.w3.org/
22:19:09 <dc_rdfig> Label GET not found.
22:28:48 * AaronSw wonders off
23:23:17 * AaronSw resurfaces
23:23:44 <danbri> hiya
23:24:17 <AaronSw> hi danbri
23:24:23 <bijan> Dan!
23:35:06 * danbri stirs stupidly on syndication list
23:35:58 <bijan> No!
23:36:03 <bijan> Don't do it, Dan!
23:36:11 <Morbus> heh.
23:41:33 <AaronSw> Oh dear, this is going to be fun, Dan
23:42:01 * AaronSw thinks about retopicing to "RDF = PICS-NG = Pornography Description Framework, it's time the truth be told..."
23:42:42 <danbri> Adobe might not be too happy...
23:42:49 <Morbus> hehehe
23:42:58 <AaronSw> Heh, I wonder how much PDF Porn is out there.
23:43:39 * bijan needs to develop an RDF schema for Mailing List Summaries.
23:43:43 <bijan> This is very tedious.
23:44:49 <AaronSw> You might be able to borrow from DanC's mr76 work
23:47:20 <AaronSw> Hmm, that might go over well in congress... "AOL Time Warner is leading distributor of porn" (via dmoz, of course)
23:53:14 <Morbus> i have this annoying habit of always typing blockquoyte instead of blockquote. phoeey
The IRC chat here was automatically logged without editing and contains content written by the chat participants identified by their IRC nick. No other identity is recorded.
Alternate versions:
and
Text
Provided by Dave Beckett. Hosted by Useful Information Company.