Semantic Web Interest Group IRC Chat Logs for 2002-04-26

00:37:50 <sbp>http://www.ninebynine.org/RDFNotes/RDFForLittleLanguages.htm

00:37:50 <dc_rdfig> A: http://www.ninebynine.org/RDFNotes/RDFForLittleLanguages.htm from sbp

00:38:05 <sbp> A:|RDF for "Little Languages" - Query, Transformation and Report Generation

00:38:05 <dc_rdfig> titled item A

00:38:38 <sbp> A:"This note describes an experimental software development in which RDF/N3 is used to code query and report generation functions performed on RDF data." also: [http://www.ninebynine.org/Software/N3ReportGenerator.zip|Python code]

00:38:39 <dc_rdfig> added comment A1

00:39:36 <sbp> A:The N3 parser is interesting - uses some outdated forms and introduces some strange new directives, but a new approach is always welcome

00:39:36 <dc_rdfig> added comment A2

00:41:01 <sbp> let's speed test it...

00:42:12 <sbp> ugh, bug in the test code

00:47:05 <sbp> [[[

00:47:05 <sbp> N3 directive error: Name expected

00:47:05 <sbp> @prefix : <<<--

00:47:06 <sbp> ]]]

01:00:22 <bijan> Hmm. THere seems to be a category error in the following statement:

01:00:27 <bijan> """HTTP, or HyperText Transport Protocol, is used by virtually every Web page on the Internet."""

01:00:37 <bijan> (From: http://news.zdnet.co.uk/story/0,,t269-s2105076,00.html)

01:01:58 <bijan> Hmm, and perhaps a transcription error:

01:02:02 <bijan> """"Microsoft has some ideas (on how to break the independence on HTTP), IBM has some ideas, and others have ideas. We'll see," he said."""

01:02:09 <bijan> s/independence/dependence/ :)

01:04:28 * danbri has some ideas too

01:04:41 <bijan> Yay! Let's hear 'em.

01:05:14 <danbri> Let's reimplement HTTP as a set of SOAP/WSDL interfaces! Then we wouldn't need Google at all -- we could find documents using our handy local UDDI repositories

01:05:20 <bijan> I have an interesting *prima facie* argument for the Semantic Web that was spawned by all this Google nonsence.

01:05:26 <danbri> do tell

01:05:51 <bijan> Hmm. let me run it past you, if you can't point me to anyone else who's made it, I'll write it up for mf.com

01:05:57 <danbri> ok

01:06:03 * danbri claims to have thought of it first

01:06:06 <bijan> 1) Web links are one way untyped.

01:06:15 <danbri> what's 'one way'?

01:06:23 <bijan> I.e., not two-way.

01:06:31 <bijan> sorry,

01:06:31 <danbri> thanks for the clarification

01:06:33 <danbri> what's 'one way'?

01:06:41 <bijan> 1) web links are one-way, untyped links.

01:06:48 <bijan> They only point in one direction?

01:07:04 <danbri> OK f'now. Been thinking about this a bit, that's all. excuse the interruption.

01:07:04 <bijan> It doesn't matter.

01:07:12 <bijan> 1) Web links are untyped.

01:07:19 * danbri nods agreeably

01:07:30 <bijan> 2) If you get enough of them, you can do really cool things. (Existence proof: Google)

01:07:46 <danbri> true

01:07:56 <bijan> (In fact, incrediably cool things one wouldn't have expects. I refer specifcially to pagerank and other datamining techniques)

01:08:05 <danbri> uhuh

01:08:13 <bijan> (This is above the "local level" deliberate coolness that they have.)

01:08:31 <bijan> 3) The basic point of RDF/SemWeb is that links should have "semantic" types.

01:08:36 <danbri> very useful data extracted where you'd expect none...

01:08:49 <danbri> muhuh

01:08:53 <bijan> (Unlike, say, XLink types which are more "syntactic" (inclusion, etc.))

01:09:11 <danbri> xlink also has roles or somesuch, and a (strained?) mapping to rdf.

01:09:13 <bijan> 4) One should expect correspondingly googlecool effects.

01:09:23 <danbri> html links had types too (but nobody used them...)

01:09:29 <bijan> Yes.

01:09:34 <danbri> 1-4: works for me

01:09:39 <bijan> But those were mostly structural, yes?

01:09:45 <bijan> link, prev, etc.?

01:10:01 <danbri> I think this captures a common, not often explcitly articulated assumption many of us have

01:10:18 <danbri> sitemappy... but also 'help', 'bookmark' ,'comment' etc.

01:10:20 <bijan> This may be old hate to everyone, but coming from a logicy background this isn't obvious.

01:10:24 <bijan> Yeah.

01:10:33 <bijan> So, fundemental challenges:

01:10:49 <bijan> 1) Untyped links are harder to "screw up" since there's less to screw up.

01:10:59 * danbri wasn't complaining: articulating stuff nobody thouht to write down is a _good_ thing

01:11:05 <bijan> 2) Typed links generally require more overhead (selecting the type), etc.

01:11:06 <danbri> sometimes it makes less sense when written down!

01:11:16 <danbri> yep

01:11:21 <bijan> 3) Hard to see what the unanticpated benefits will be (by definition)

01:11:40 <bijan> note that this is a prima facie argument only. Might turn out that a big web of typed links sucks.

01:11:42 <danbri> html4 linsk - http://www.w3.org/TR/html401/types.html#type-links

01:12:15 <bijan> 4) RDF type links are more granular==more overhead.

01:12:22 <deltab> I wouldn't say that nobody uses html's typed links - I've seen and used them on a number of sites

01:12:25 <bijan> Actually, I was talking with jim about this.

01:12:41 <bijan> But htey aren't well supported in browsers.

01:13:01 <bijan> Goodness! There are a lot of types!

01:13:16 <bijan> I wonder if google uses link types.

01:13:18 <deltab> it's getting better - mozilla has a toolbar for using them

01:13:29 <bijan> icab used them too, I think.

01:13:30 <danbri>http://www.csd.uwo.ca/~jamie/.Refs/link-types.html

01:13:30 <dc_rdfig> B: http://www.csd.uwo.ca/~jamie/.Refs/link-types.html from danbri

01:13:48 <deltab> lynx and links display them too

01:13:50 <danbri> B:|HTML Link typing draft, TimBL and DanC (1993)

01:13:50 <dc_rdfig> titled item B

01:14:08 <bijan> Hmm. SemWeb types should evangalize them.

01:14:18 <bijan> And then offer rdf as 'XML for link types' L)

01:14:29 <danbri> B:"transcribed" by J. Blustein <jamie@csd.uwo.ca>

01:14:30 <dc_rdfig> added comment B1

01:14:40 <danbri> that has been suggested... :)

01:14:47 <bijan> Heh.

01:14:56 <bijan> Well, this gives me a research project.

01:15:00 <bijan> Another one, at least.

01:15:21 <bijan> Oh, so i was talking about htis with jim.

01:15:25 <danbri> We should propose X-Link v2.0, allowing links to things where pointing is done using 'reference-by-description'

01:15:37 <bijan> When he was describing his "virtual portal" idea.

01:15:47 <danbri> eg. an xlink from my homepage to "the movie that has an encoding that has a sha1sum of <squiggles>"

01:15:49 <bijan> And I pointed out that the way we tend to write web pages is by using a lot of googling

01:15:51 <bijan> At least me :)

01:15:57 <bijan> At least I do.

01:16:17 <bijan> If "linking" to ontology terms was at least as easy as linking to webpages, it be more popular.

01:16:24 <danbri> google - me too

01:16:29 <bijan> Janus (from jim's lab) is supposed to do some of that.

01:16:33 <danbri> yeah

02:08:02 <deltab> deltab is now known as in

02:08:17 <in> in is now known as deltab

02:18:05 <SethR> links by description is a great idea :)

02:19:15 <SethR> <a [movie called "Gone With The Wind"]>Gone With The Wind</a>

02:19:51 <SethR> .google

02:19:51 <xena> usage: google <query>

02:19:52 <xena> - returns the url of top google search result for the query.

02:20:01 <SethR> .google Gone With The Wind

02:20:01 <xena> Gone With The Wind: http://www.franklymydear.com

02:21:07 <SethR> beats link rot

02:46:05 <AaronSw> interesting! http://info.alexa.com/search?q=scripting

02:46:59 <sbp> indeed

03:12:37 <DanCon> hmm... I'm afraid I found something interesting courtesy of DaveW: "My favorite new app development environment, PythonCard, is about to meet up with my favorite new operating system, Mac OS X."

03:12:48 <DanCon> -> http://radio.weblogs.com/0001285/2002/04/24.html

03:13:11 * sbp is not sure that he can ever forgive DanC :-)

03:13:46 <sbp> heh, it's not even from DW

03:16:15 <sbp> DanC: did you ever hook the SPARK N3 stuff up to an RDF store?

03:18:08 <DanCon> umm... I don't think so. yapps blows away spark for N3

03:18:46 <DanCon> Dan Shafer ... is that the same guy who wrote this HyperTalk book I've had on my shelf since like 1990? seems to be!

03:18:46 <sbp> aha, that's just what I wanted to know :-)

03:19:12 <sbp> ooh, an interconnectedness event

04:04:13 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome

04:21:01 <mnot> Hmm, new Amaya. No OSX :(

04:39:30 <DanCon> the fink guys support Amaya on osx. no worries.

04:53:38 <AaronSw> mnot, have you seen amk's work on a REST framework for python?

04:55:48 <mnot> yeah, I tried it with fink, but it didn't seem to want to work.

04:56:04 <mnot> AaronSw: saw that, not sure exactly where he's going. I should drop him a line.

04:56:37 <mnot> I think he's more interested in something that can read/write WRDL or similar; I'm just putting together REST helper libraries (OK, it's a little more than that...)

04:57:21 <mnot> DanCon: but thanks for the heads-up, mate!

04:57:35 * mnot feels his adopted australian roots

04:58:01 <AaronSw> sounds a bit like that to me: "I ended up writing a class to represent each object and sub-object; so if there was a Service URI, and Services contain Tickets, I needed to write Service and Ticket classes."

04:58:18 <mnot> Oh, I didn't see that. Interesting.

04:58:28 <mnot> cool, thanks, I definately will have to talk to him.

04:58:42 <AaronSw> ooh, there are binary amaya packages now

04:59:00 <mnot> what platform do you speak of?

04:59:04 <AaronSw> OS X

04:59:06 <AaronSw> (of course)

04:59:09 <mnot> in fink?

04:59:17 <AaronSw> yeah. `sudo apt-get install amaya`

04:59:40 <mnot> is it amaya 6?

04:59:45 <AaronSw> 5.3-2

04:59:57 <mnot> hmm, couldn't find package amaya. time to run dselect, perhaps

05:00:08 <AaronSw> try `sudo apt-get update`

05:00:27 <mnot> ooh, it's working!

05:00:55 <AaronSw> oh, cool. Damian Steer did the Amaya packages!

05:01:00 <mnot> I need to get more RAM to run X decently, tho, and that means spending $$ :(

05:01:13 <AaronSw> RAM's pretty cheap...

05:01:30 <mnot> I'm on a tight budget :/

05:01:40 <AaronSw> I will have to bug damian for WorldWideWeb packages. ;)

05:01:53 <mnot> now *that* would be cool

05:01:56 * AaronSw dccs mnot some RAM

05:02:09 <mnot> hey, thanks!

05:05:01 <mnot> it works! slow as hell, but it works! now I can edit my site with PUT again...

05:05:21 <mnot> hmm, the CSS on http://www.mnot.net/ doesn't seem too amaya-friendly... whoops.

05:08:22 <AaronSw> heh, cool!

05:44:35 <AaronSw> woo! X works again

05:44:40 <AaronSw> had to delete ~/.Xauthority

05:44:45 <AaronSw> and amaya works too. great!

05:50:19 * AaronSw grabs OroborOSX

05:50:20 <AaronSw> wow, neat!

05:57:17 <mnot> why does oroborosx ring a bell?

05:57:51 <AaronSw> dunno, but I highly recommend it: http://wrench.et.ic.ac.uk/adrian/software/oroborosx/

05:58:00 <AaronSw> makes X programs look like native os x apps

05:58:07 <mnot> wishes one could point Advogato, slashdot, etc. at a single RDF profile/diary/etc that you maintain on YOUR site (or hell, even one of theirs), instead of keeping the data so close to their chests.

05:58:12 <mnot> oh, *kewl*

05:58:22 <AaronSw> they even minimize into the dock

06:03:08 <mnot> that's amazing. It's a pig, of course, on my little machine, but very, very nice.

06:03:29 <mnot> thanks

06:03:46 <AaronSw> sure

09:38:23 <libby_> libby_ is now known as libby

10:27:02 <jang> q: what's the difference between "ontology" and "schema"

10:27:09 <jang> AaronSw: 1 point in scrabble.

10:27:30 * jang curses an IRC client that ruined a perfectly good punchline

10:33:12 <JibberJim> Ah it worked okay, it just looks like you were the straight man to Aaron...

10:34:44 <danbri> answer: only one of them gets a w3.org site if you search on google

10:34:52 <danbri> ...and that is for xml schema

10:35:20 <danbri>http://www.w3.org/2001/12/rubyrdf/examples/redrubytests.rb

10:35:20 <dc_rdfig> C: http://www.w3.org/2001/12/rubyrdf/examples/redrubytests.rb from danbri

10:35:29 <danbri> C:|Calling Redland from Ruby

10:35:29 <dc_rdfig> titled item C

10:36:15 <danbri> C:Thanks to Dave for help with this (configure/make/autoconf and other mysterious voodoo...)

10:36:16 <dc_rdfig> added comment C1

10:36:32 <danbri> C:You need the latest SWIG (from their CVS) to get it running. And Redland from CVS.

10:36:32 <dc_rdfig> added comment C2

11:01:55 * danbri -> ilrt/lunch

11:02:00 <danbri> will be online later, fwiw

11:24:46 <dajobe> C:I got 'make test-ruby' now works, not by learning ruby but by reading your code

11:24:46 <dc_rdfig> added comment C3

11:24:56 <dajobe> C3:I got 'make test-ruby' on example.rb working, not by learning ruby but by reading your code

11:24:56 <dc_rdfig> replaced comment C3

13:05:45 <lilo> [GlobalNotice] Hi all. Just a reminder: From time to time, non-critical announcements, comments and detailed OPN administrative information can be found on wallops. Set your usermode to +w if you want to see them. Thanks.

13:10:32 * AaronSw saves up vocabularydescriptionlanguage for his next scrabble game

13:15:45 * larsbot wonders how big Aaron's scrabble board is

13:39:42 <AaronSw> Uh oh, Edd is a source of Stop Energy.

13:45:16 <AaronSw> .acronym ptdbd

13:45:17 <xena> ptdbd:

13:45:45 <AaronSw> oh. pecked to death by ducks

13:51:18 <bijan> Ugh. This David Watson's technique of posting nothing but a URL to a blog article on his site with *no* other text is very annoying.

13:52:33 <bijan> (example in: http://www.xml.com/cs/user/view/cs_msg/558)

13:52:37 <AaronSw> Well, don't you read his weblog anyway? ;)

13:53:23 <bijan> i don't mind him "keeping his content" on his website and posting pointers *per se*, but I wish he'd add a *little* abstract.

13:53:45 <bijan> Of course, proper inclusions (hi Ted!) would fix this too :)

13:55:31 <AaronSw> That David Watson doesn't sound like the friendlies guy, from reading his blog postings.

13:55:58 <AaronSw> Heh: "I have produced demonstrable GUI applications in a variety of languages with usable performance characteristics and that's good enough for me."

13:56:01 <bijan> Hmm. Now *this* a bit of hyperbole:

13:56:08 <bijan> """Worried about some vulpine criminal ripping off one of your digital jewels? YouÕre not alone: image theft far surpasses identity theft as the number one security concern on the Internet. """

13:56:20 <bijan> (From: http://www.apple.com/macosx/applications/photoshop/)

13:56:27 <AaronSw> Yeah, I *loved* that.

13:56:31 <bijan> Yes, I laughed at that line too by Watson.

13:56:35 <AaronSw> And the cute little lock they had to illustrate.

13:56:51 <bijan> vulpine criminality is the *worst*.

13:57:04 <bijan> Some english major/ma/phd is *way* out of control :)

13:57:14 * AaronSw giggles

13:57:55 <AaronSw> I can just imagine t he focus groups: "I mean, if it was just some stupid criminal stealing my image, I wouldn't mind. But *vulpine criminals*?! That's just going too far for me.

13:57:56 <AaronSw> "

13:58:02 <bijan> I worry about lupine criminal myself.

13:58:19 <bijan> Not to mention those pesky feline ones.

14:00:48 <bijan> Eek!:

14:00:50 <bijan> """So sweeping is this upgrade, in fact, that Photoshop 7.0 even addresses whatÕs traditionally been viewed as the weak link in the armor of most great graphics artists: their spelling. Finally, thereÕs even a built-in multilingual spell-checker that checks spelling in multiple languages within the same file. """

14:01:27 <bijan> Wow, rampant use of "even", even.

14:01:50 <bijan> Guess they need to add a "style and hyperbole checker" :)

14:07:09 <bijan>http://www.time.com/time/magazine/article/0,9171,1101020429-232589,00.html

14:07:10 <dc_rdfig> D: http://www.time.com/time/magazine/article/0,9171,1101020429-232589,00.html from bijan

14:07:34 <bijan> D:|Browser Wars Revisionism by the running dogs at Time Magazine

14:07:35 <dc_rdfig> titled item D

14:07:44 <bijan> D:Although they are boosting Mozilla.

14:07:44 <dc_rdfig> added comment D1

14:08:04 <bijan> D:"""In the beginning there was one Web browser. It was called Mosaic, and if you didn't like it you could go back to watching Murphy Brown, or whatever it was we did before we had the Web."""

14:08:05 <dc_rdfig> added comment D2

14:09:00 <bijan> D: I'll leave it to the dweebs to point out the errors in the fist clause. But clearly, what we did before the Web was read and send email (or netnews).

14:09:00 <dc_rdfig> added comment D3

14:09:12 <bijan> D: But the kicker:

14:09:12 <dc_rdfig> added comment D4

14:09:20 <bijan> D:"""Then Microsoft started giving away Internet Explorer, Mosaic turned into Netscape, and suddenly life was complicated."""

14:09:20 <dc_rdfig> added comment D5

15:16:29 <em> dajobe ?

15:16:58 <em> various sw talks ... http://www.w3.org/2001/sw/EO/talks

15:17:05 <em> re diglib

15:17:08 <em> the ECDL one

15:18:09 * chaals wonders how hard it is for AOL/Time/Warner to do research on strange things like Mozilla

15:18:43 <em> note that marja did an excelletn job building off this one for the fin semweb talk which may be more usefuly

15:19:24 <em> -> http://www.w3.org/Talks/2001/1102-semweb-fin/

15:22:25 * em wonders what the heck <ywbcreapq> I met so many more interesting people at http://www.afriend-finder.cjb.net spam is...

15:39:21 * JibberJim thinks it might be spam...

15:43:40 <grrrrr> I beleive I was instigating op wars on #hack and inventing icmp attacks before the web

15:49:47 <grrrrr> is there a quantifiable value to importing dmoz into parka?

15:56:50 <bijan> Er..I guess it depends on what you mean by "quantifiable"

15:57:42 <bijan> for parka folks, the initial value is getting as many Very Large Knowledge Bases into parka for testing, development, and timing purposes.

16:05:14 <sandro> Has the question of what <http://www.microsoft.com/> denotes ever been on the RDF Core issues list? The closest I see is http://www.w3.org/2000/03/rdf-tracking/#rdfms-fragments which is really different.

16:10:38 <sandro> Ah, got it. http://www.w3.org/2000/03/rdf-tracking/#rdfms-resource-semantics

16:11:44 <sandro> Ugh. Not Satisfactory.

16:18:28 <sandro> There's a process problem when someone raises an issue (a), it gets merged in with another issue(b), the issue (b) gets resolved, but the person who raised (a) never gets properly notified.

16:44:29 <AaronSw-lfc> sbp, a new approach is _always_ welcome?

16:46:53 <AaronSw-lfc> lol @ http://www.u.arizona.edu/~chalmers/misc/proofs.html

16:49:33 <sandro> That's wonderful. :-)

16:51:41 <AaronSw-lfc> neat: http://www.hpl.hp.com/shl/papers/smallworld/smallworldpaper.html

17:09:54 <AaronSw-lfc> simonstl: "HTTP was hackery, but had time to adjust before it reached a wide audience. Web Services is hackery, but its proponents insist on staying the Web Services course as previously charted.

17:13:32 <AaronSw-lfc> RoyF: " am not here to accommodate the

17:13:33 <AaronSw-lfc> requirements of mass hysteria.

17:13:40 <AaronSw-lfc> " - http://lists.w3.org/Archives/Public/www-tag/2002Apr/0235

17:14:38 <bijan> Behind the times Aaron :) Those quotes are *so* two days ago!

17:15:01 <bijan> Wait until you get to the insults hurled back at roy :)

17:16:04 <bijan> It must be Spring Fever.

17:16:06 <AaronSw-lfc> Heh.

17:16:17 <AaronSw-lfc> It's like there's Web Services craziness in the air. I'm getting sick already.

17:22:54 <bijan> Hmm. Patrick Stickler is inquiring about informal RDFCore F2Fs at WWW2002...given recent mailing list exchanges, would *you* show up? ;)

17:23:07 <AaronSw-lfc> I am so there.

17:23:28 <bijan> "For those of us who will be at WWW2002, do we want to arrange

17:23:28 <bijan> some particular time to get together (dinner, beer, assults, whatever)?"

17:23:30 * sandro :-( :-( (/me wishes he could be there.)

17:23:40 <bijan> Grr.

17:23:42 <bijan> assaults :)

17:23:48 <bijan> Oh well, the forgery was revealed.

17:24:03 <bijan> Should have gone with "knife fights"

17:25:11 * AaronSw-lfc dccs sandro magical MA-Hawaii transporter

17:25:20 <AaronSw-lfc> Oh, and a (cough) Home-MIT transporter (cough)

17:28:42 <sandro> Yeah, it's not the going there that's hard -- it's coming home in, say, under 3 hours (when my wife goes into labor).

17:30:08 <AaronSw-lfc> That's what the transporter is for.

17:33:01 <sandro> Ooh, yeah. I misread your (cough) bit as giving me a transporter back home from Hawaii.

17:49:49 <DanCon> sandro, my connection to mit is flakey allthesudden.

17:49:57 <DanCon> odd, cuz I can ping the relevant machine just fine

17:52:28 <sandro> The Internet works in strange and mysterious ways. Far more so these days.....

17:53:24 * DanCon wonders if chaals is around to help pick colors for WG schedule stuff

17:53:44 * chaals was about not to be around

17:54:07 <chaals> &*(^&) browser just crashed

17:55:07 <DanCon> somebody mailed me suggesting to use progressive colors of the spectrum to represent progress along the REC track.

17:55:11 <DanCon> the stylesheet is http://www.w3.org/2000/04/mem-news/wg-sched-style.css btw

17:55:15 <chaals> text is going to be blue, right? (since it is a link)

17:55:24 <DanCon> you're asking me?

17:55:48 * chaals got that - I plan to be a customer of this so it is bookmarked

17:56:08 <chaals> hmm. 5 things to be distinguished...

17:56:22 <DanCon> woohoo! the rules to integrate scheduling info into the chart worked!

17:56:42 <DanCon> well, now I have several more to distinguish: *planned* LC/CR/PR/REC

17:57:59 <DanCon> hm... I'd like to have the most recent publication and the next milestone in the same row of the chart, but I'm not sure I can do that without closed world reasoning.

17:58:22 <chaals> Approach 1: Use text LC CR etc on a background

17:58:33 <chaals> approach 2: use borders (weak)

17:59:16 * DanCon wanders off to an appointment...

18:14:25 <AaronSw>http://www.w3.org/2002/03/RRSAgent

18:14:25 <dc_rdfig> E: http://www.w3.org/2002/03/RRSAgent from AaronSw

18:15:50 <AaronSw> Oops, it's w3cacled

18:16:00 <AaronSw> E:|The RRSAgent IRC Bot

18:16:00 <dc_rdfig> titled item E

18:16:10 <AaronSw> E:Part of SWAD IRC fun

18:16:10 <dc_rdfig> added comment E1

18:16:39 <SethR> Sorry, Unauthorized.

18:17:04 <AaronSw> E:Oops, currently only for W3C members...

18:17:04 <dc_rdfig> added comment E2

18:23:34 * azaroth looks around for danbri

18:26:40 * chaals prodded the stylesheet and then wandered off to an appointment

18:44:57 <AaronSw> azaroth, bringer of all things MOOy

19:06:07 * azaroth grins to AaronSw.

19:06:28 * azaroth is looking for danbri for just that reason, yes

19:06:46 <AaronSw> Woo! I was trying to build a MOObot thing the other day that'd lett you do `look nick`.

19:15:08 <azaroth> A bot that would store decriptions of users and msg you with them in response?

19:15:50 <AaronSw> Yep.

19:22:53 <JibberJim> I was looking to add nick's to my foaf data and then FillyJonk can point you to an image of a person by nick.

19:24:01 <JibberJim> ie http://jibbering.com/rdf/foafwho2.asp?person=mailto%3Adanbri@w3.org points you too pictures of danbri

20:04:45 <azaroth> last danbri

20:05:03 <azaroth> Nope :) Is there a lastseen bot?

20:09:34 <AaronSw> .seen danbri

20:09:34 <xena> danbri seen in #rdfig saying: [ will be online later, fwiw ] ~ 9 hr(s) 7 min(s) 34 sec(s) ago

20:09:44 <AaronSw> .seen danbri 2

20:09:47 <azaroth> Ta

20:46:17 <dajobe> bijan: patrickS is quite nice IRL

20:46:28 <bijan> Oh I believe it!

20:46:31 <bijan> Most people are.

20:46:49 <bijan> Actually, the heat seems to be calming down.

20:46:53 <bijan> Er...

20:46:56 <bijan> cooling down.

20:47:00 <bijan> Well the *heat* isn't.

20:47:04 <bijan> I guess the discussions are :)

20:47:06 <bijan> Or something.

20:47:10 * dajobe looks forward to meeting bijan

20:47:12 <bijan> I think I may be a bit tired.

20:47:20 <bijan> Heh.

20:47:54 <AaronSw> We get to meet bijan?

20:48:01 <AaronSw> When? When?

20:48:03 <bijan> No.

20:48:03 <dajobe> hehe

20:48:06 <bijan> Not you aaron.

20:48:07 <AaronSw> Aw.

20:48:35 <dajobe> we just need to schedule a meeting at ~bijan, wherever that is

20:48:50 <AaronSw> UMD, I assume.

20:48:51 <bijan> But I'm always "somewhere else" :)

20:49:09 <bijan> Jim will prolly drag me along to something somtime.

20:49:39 <bijan> Heh. The GaTech Squeakers are finally going to see me in the flesh after *5* years :)

20:49:50 <dajobe> haha

20:49:57 <bijan> Unless I get sick or wimp out over the next week :)

20:50:07 <dajobe> is after your Squeak book chapter stuff?

20:50:15 <AaronSw> they better take a photo

20:50:25 <bijan> Actually, I knew mark guzdial before the Squeak book.

20:50:31 <bijan> That's how I got that gig, in fact.

20:50:44 <bijan> But I'm on their internal mailing list.

20:50:52 <bijan> So I'm always answering questions and having discussions.

20:51:07 <bijan> And inevitably, the first year grad students ask, "Who the hell are you? have I seen you around?"

20:51:20 <dajobe> bijan's chapter: http://coweb.cc.gatech.edu/squeakbook/19

21:02:01 <DanCon> AaronSw, have you played with the speech recognition stuff in OSX? it seems to work, at least the default vocabulary

21:02:14 <DanCon> my kids got a kick out of the "tell me a joke" feature

21:02:48 <DanCon> my wife wants me to configure it to handle "computer, get the weather report" but I haven't figured out how.

21:02:58 <bijan> DanCon, is that the same basic stuff that's been in MacOS for aeons?

21:03:03 <bijan> I.e., instead of viavoice or some such?

21:03:15 <DanCon> it's not viavoice. I dunno how long it's been around.

21:03:37 <bijan> It sounds like the stuff they put in when the powerpc first came out.

21:03:41 <bijan> I had it on my 6100

21:03:45 <bijan> Quite amusing.

21:03:52 <bijan> No training? Simple phrases?

21:04:09 <DanCon> no training. simple phrases.

21:04:26 <DanCon> do you know the UI well enough to get it to read web pages?

21:04:48 <DanCon> or at least: visit a web page from the favorites menu?

21:04:50 <bijan> Hmm.

21:04:53 <bijan> You me open them?

21:04:58 <bijan> Or do text to speech.

21:05:03 <bijan> It's been a while.

21:05:17 <DanCon> I'd prefer text to speech, but just visiting/opening the relevant page would be kinda cool

21:06:23 <bijan> Well, since you can use it to run applescripts, you can just wrap the url in a script.

21:06:30 <bijan> Hmm. I bet there's a generic "open foo"

21:06:46 <bijan> So if you had an alias to a web page, "computer, open google" should work.

21:07:06 <bijan> (where you made a easily recoged name for the alias.

21:07:43 <bijan> As for TTS, all I remember is that SimpleText had a menu for reading the document or a selection.

21:10:26 <AaronSw> hey DanCon, yeah i know quite a bit about it

21:10:38 <AaronSw> heh, tell me a joke is great

21:11:08 <AaronSw> to get it to visit a web page, just drag the @ thing from the website in the browser to the Speakable Items folder

21:11:17 <DanCon> "since you can use it to run applescripts"... do tell?

21:11:20 <AaronSw> and give it the name you want to say like "what's the weather" as the filename

21:11:23 <bijan> When you speak a built-in command or the name of a file in the Speakable Items folder, the computer acts as if you double-clicked the fileÕs icon. For example, if you have a file named ÒMy ResumeÓ in the Speakable Items folder, you can open it by saying ÒMy ResumeÓ.

21:11:33 <bijan> Here: http://www.apple.com/disability/easyaccess.html#text

21:11:42 <AaronSw> the tell me a joke thing is just an applescript

21:11:51 <bijan> Since an applescript is just a file, you can pop them into the speakable items folder too.

21:11:54 <DanCon> logger, bookmark?

21:11:54 <DanCon> See http://ilrt.org/discovery/chatlogs/rdfig/2002-04-26#T21-11-54

21:12:06 <DanCon> thanks for the clues.

21:12:52 <bijan> Switching voices is a lot of fun.

21:13:02 <bijan> The ones that sing always cracked me up.

21:13:04 <AaronSw> it'd be pretty easy to write an python applescript that regexped the page and said it

21:13:13 <AaronSw> the command to t2s is just "say", i think

21:13:38 <bijan> If your kids liked the jokes, open up simpletext, type in some text and try the different voices.

21:13:48 <bijan> I don't know why, but I did this for like two days :)

21:13:51 <bijan> As did all my friends.

21:14:18 <AaronSw> When I was in high school, some kids got a hold of my laptop and spent the entire afternoon making it say stupid things (that amuse high schoolers)

21:14:32 <DanCon> yeah, it's pretty cool stuff, for something that gets zero attention in the iMac marketing materials these days.

21:14:58 <AaronSw> ooh, this is neat. try this from the command line: echo 'say "hello world"' | osascript

21:15:06 <bijan> Alas, it's fairly old.

21:15:15 <bijan> IIRC, http://www.bell-labs.com/project/tts/voices.html is way smoother.

21:15:20 <edd> you kids and your toys, the amiga used to do that :)

21:15:27 <bijan> Squeak has some crude Klatt focused TTS.

21:15:41 <AaronSw> The speech recognition system that's included lets you write your own grammar-thingies if you want to do more advanced stuff

21:15:49 <bijan> With lip synching to cartoon heads.

21:15:58 <bijan> Which is fun.

22:20:49 <edd> dc_rdfig: foo

22:20:53 <edd> hmm

22:20:55 <edd> a bogus bot

22:22:20 <edd> dc_rdfig:view 1

22:22:20 <dc_rdfig> E: The RRSAgent IRC Bot (http://www.w3.org/2002/03/RRSAgent)

22:22:22 <edd> cool

23:48:35 <azaroth> Net splits, we hates them.