Semantic Web Interest Group IRC Chat Logs for 2003-09-12

00:03:46 <Aredridel> pipe it through OD, though there are no illegal characters in 8859-1

00:28:44 <zacker> greetings~!

00:41:05 <aharth> ndw: if have a java program that removes illegal control characters from XML

00:41:36 <aharth> s/if/i

00:45:03 <tav|offline> tav|offline is now known as tav

03:29:01 <adamhill> adamhill is now known as _adam_wk

07:07:46 <mdupont> mdupont is now known as mdupont-work

08:21:28 <arnarl> hi

09:26:20 <JimJibber> JimJibber is now known as JibberJim

09:44:20 <sapi> any one know how to change a base map in a street pilot 3?

10:37:53 <cmjg> danbri, danc: you awake?

10:41:29 <danbri> sure

10:44:37 <Wack> hmm, webserver on http://rdfig.xmlhack.com/ is down (?)

10:47:14 <dajobe> upgrades

10:47:29 <dajobe> 200M of upgrades actually

10:53:20 <dajobe> you can always just read the logs

10:53:25 <dajobe> see the topic

11:04:54 <dajobe> website back

13:04:13 <DanC>http://weblogs.java.net/pub/wlg/445

13:04:14 <dc_rdfig> A: http://weblogs.java.net/pub/wlg/445 from DanC

13:04:32 <DanC> A:|SAML V1.1 is final

13:04:32 <dc_rdfig> Titled item A.

13:04:38 <DanC> A:from Eve Maler's weblog

13:04:38 <dc_rdfig> Added comment A1.

13:04:52 <DanC> A:"One especially useful scenario for SAML is single sign-on (SSO), where a user can log in to one website but then proceed to use resources at a website in a different domain"

13:04:53 <dc_rdfig> Added comment A2.

13:33:40 <zoyd> does anyone know if chefmoz can be queried or is there any regional feed .. other than the big .gz blob?

13:48:53 <ndw_> ndw_ is now known as ndw

13:51:12 * DanC notes nearly a quorum of W3C URI CG folk frequent this channel...

13:52:29 * DanC considers scheduled CG chats... I'm getting that feeling like we need a synchronous event to deal with some outstanding email, but telcons are kinda a drag

13:56:36 * ndw is willing to give it a whirl

13:57:23 <DanC> on the other hand, most of the outstanding email is stuff that *I* owe responses to, I think...

13:57:48 <DanC> the URI CG is supposed to be a mechanism for me to distibute that load... but I haven't even spent the time necessary to get that working.

13:58:41 * DanC looks at other TODO input to see if URI foo can get some real bandwidth today...

13:58:48 <DanC> is the TAG extensibility thing public yet?

13:58:58 <DanC> I need to prep for the RDDL discussion too.

13:59:10 <ndw> extensibility=versioning?

13:59:17 <DanC> ppht. yes, versioning.

13:59:41 <ndw> No. Dave and I said we'd make the next one public, now that we have some TAG feedback. Next week sometime is the plan.

14:00:04 * DanC grumbles at a whole bunch of tag telcon time spent on something that's not public

14:00:33 * ndw decides not to listen.

14:01:10 * DanC declares stalemate

14:01:39 * ndw sighs.

14:02:02 <ndw> I'm sympathetic, but I'm a coward. I wanted some private feedback before public.

14:02:41 <DanC> I can appreciate getting private feedback. But tag telcons aren't private.

14:02:53 <ndw> Fair point.

14:23:45 <ndw> What's the URI for an XML Schema simple type?

14:25:48 <ndw> Oops. The namespace URI for xdt: is http://www.w3.org/2003/05/xpath-datathttp://www.w3.org/2003/05/xpath-datatypesuntypedAtomic> which seems a bit unfortunateypes which makes xdt:untypedAtomic <

14:28:39 <ndw> Nevermind

15:13:44 * DanC waves

15:13:58 <ndw> Can cwm test for the namespace name of a resource?

15:14:08 <ndw> I can wrote { :p :hasProp :q } log:implies ...

15:14:19 <ndw> can I write { :p isinthenamespace xxx } log:implies ... ?

15:14:26 <DanC> yes...

15:14:54 <DanC> { :p.log:uri str:startWith "http:..//nsuri..." } => {... }

15:15:11 <ndw> thx

15:15:18 * chaalsMEL waves

15:15:34 <DanC> BLURB: WebOnt test results, EARL, and provenance

15:15:34 <dc_rdfig> B: WebOnt test results, EARL, and provenance from DanC

15:16:34 <DanC> B:cf [http://lists.w3.org/Archives/Public/w3c-wai-er-ig/2003Sep/0003.html|browsable test results] thread

15:16:34 <dc_rdfig> Added comment B1.

15:16:43 <chaalsMEL> B: [EARL|http://www.w3.org/TR/EARL10] is a W3C draft for an RDF vocabulary that expresses conformance results

15:16:44 <dc_rdfig> Added comment B2.

15:16:53 * DanC wonders if sandro is handy

15:17:18 <chaalsMEL> B: developed by the [http://www.w3.org/WAI/ER|Evaluation and Repair Tool Group] at w3c.

15:17:18 <dc_rdfig> Added comment B3.

15:18:12 * sandro is not very handy

15:18:15 <sandro> but a little

15:20:13 <DanC> sandro, it seems our test results ontology overlaps substantially with EARL

15:20:27 <DanC> seems like an interesting real-world case of ontology mapping/migration

15:20:42 <chaalsMEL> B: EARL has been implemented in several tools for testing accessibility (in the sense of conformance to [WCAG|http://www.w3.org/TR/WCAG10]) -

15:20:42 <dc_rdfig> Added comment B4.

15:21:16 <chaalsMEL> B: AccVerify and AccessValet are two commercial tools that spring to mind, WAINu is an open source tool.

15:21:17 <dc_rdfig> Added comment B5.

15:21:28 * sandro acks, but doesnt have the cycles to spare on it right this minute.

15:21:31 <DanC> B:"The important difference is that you have no notion of provenance in your model." writes charles. Indeed, we figured we could leave that out for now.

15:21:31 <dc_rdfig> Added comment B6.

15:22:06 <sandro> indeed, although I'm interested in trying it. Tempted to do it outside the RDF, though.

15:22:28 <DanC> yeah, it's interesting to think about using {} for provenance.

15:22:39 <sandro> (ugh. finally got my mail working again I think. It's been bouncing for ~24 hrs.)

15:22:55 <chaalsMEL> chaals believes there are good reasons for doing it inside the RDF (see the discussions that went into getting an EARL spec out)

15:23:08 <chaalsMEL> {} ??

15:23:32 <sandro> actually, I'm inclined to try keeping provenence entirely outside the set of formulas being manipulated, much like one would normally keep proof information. (provenence seems to be about the same as proof.)

15:23:50 <sandro> of course there can be an RDF view of that info -- but it's in a different KB.

15:23:56 <danbri> what's the latest on replacement reification vocab for RDF? ie. something like rdf:predicate that ranged over URI terms, not properties...

15:24:02 <danbri> ...is there a well worked out design anywhere?

15:24:06 <sandro> otherwise the reasoning is just too... crazy.

15:24:32 <sandro> My latest is that my LX design from about a year ago is fine for RDF reification.

15:24:51 <sandro> (the one written up in my rejected www conference paper.)

15:24:58 <DanC> yes, provenance looks like proof "i.e. 'oh yeah?'" to me too.

15:25:03 * chaalsMEL doesn't feel sure he undertstands what sandro is saying :(

15:25:14 <danbri> sandro, how does that compare to the stuff ericp's been doing lately?

15:26:09 <sandro> I'm saying that if you try to track the provenence of all the data in your database, and keep that tracking information in the SAME database, and you hook up some automated reasoning system, it'll be aweful. It should be fine to keep the provenance off to the side, in another database.

15:26:17 * DanC wonder whether "LX design from about a year ago" is available via http in a form that charles and I could review... surfs...

15:27:13 <chaalsMEL> Hmm. It needs to be clos enough taht when you generate some reasoned results you can find out where they came from...

15:27:16 <DanC> well, not necessarily another database; cwm can manipulate multiple sources of data in the same store

15:27:34 <danbri> there's a foaf use case for it too: division of labour between web crawlers and aggregators

15:27:49 <sandro> I'd call that another KB or database. The fact that there's some common structure is kinda irrelevant.

15:27:50 * DanC wishes for about 8 more hours in the day to work on proof stuff

15:27:54 <chaalsMEL> one of the assumptions behind EARL is that a given result assertion is not necessarily correct.

15:28:06 <danbri> you scutter up 2.5k URIs into a big local stash of RDF, chewing up time and computer resources... how then to share that with other rdf tools who don't want to do their own web-crawl...?

15:28:13 <sandro> That's okay, chaals.

15:28:23 <danbri> one answer would be an rdfdump of the db... but how to keep the who-said-what intact

15:28:33 * danbri doesn't think RDF reification up to the job, so looking to borrow something else

15:28:33 <sandro> LX stuff

15:28:34 <DanC> danbri: package it up as a wwwoffle cache

15:28:37 <sandro> LX stuff: http://www.w3.org/2002/08/LX/v3/

15:29:10 <DanC> B:hmm... consider [http://www.w3.org/2002/08/LX/v3/|LX stuff] as another angle on provenance?

15:29:11 <dc_rdfig> Added comment B7.

15:29:14 <danbri> yep, a .tar.gz of the harvested RDF (plus base URI meta info) was one thing I was considering... maybe wwwoffle makes that simpler?

15:29:27 <sandro> Yeah -- I don't have a vocab for "who said what". the LX vocab is just for the "what" part.

15:29:51 <danbri> that's ok... i'm really just after a drop-in for predicate/subject/object/Statement

15:29:53 * DanC wishes for another 8 hours a day to work on standardized freeze-dried caches; what some folks call "packaging"

15:30:29 * DanC hunts for postponed RDF core issue on quoting...

15:31:15 <danbri> B:See also [http://www.w3.org/2000/03/rdf-tracking/#rdfms-quoting|postponed rdf core issue on quoting].

15:31:16 <dc_rdfig> Added comment B8.

15:31:21 <chaalsMEL> earl has an assertor property for a collection of stuff...

15:31:49 <chaalsMEL> typically an earl:Assertion

15:32:41 * DanC wonders what RDF Core WG is doing to say "see? we have lots of tools passing all our tests"

15:32:54 <chaalsMEL> Hmm. Or is it an assertedBy property in the latest spec.

15:43:09 <sandro> danbri, simple example of using LX- this is "<a> <b> _:X" reified: http://www.w3.org/2000/10/swap/test/ref/flatten-1e.n3

15:44:27 <WillWare> I work for a cellphone firmware company, where I have been pitching the idea that we would be smart to start putting some semweb code into our product. If anybody is curious, I'd like to briefly describe my strategy and hear any feedback you folks might have.

15:46:45 <ericP> sandro, http://www.w3.org/2002/08/LX/v3/chunky.html is a 404. i was hoping to see how you did existential/universal in RDF

15:47:22 * WillWare hears no objections, and blunders on...

15:47:38 <WillWare> The internal marketing involves finding pre-deployment internal apps. A few of these are PIM data, putting timestamps and text annotations on photos and sound bites, and metadata descriptors for internal applets. There is a lot of interest in tracking user preferences.

15:47:49 <WillWare> Given my place in the org char, my job is to provide libraries/services for people writing apps, so my thoughts are (1) parsers for RDF/XML and for N3, (2) some kind of persistent database for triples, and (3) some kind of inference engine.

15:48:18 * chaalsMEL listening

15:48:43 <WillWare> Does this sound reasonable? Anything glaringly missing or bogus?

15:49:08 <WillWare> s/char/chart/

15:50:10 * chaalsMEL thinks the only value of n3 is that cwm does inferencing. You are making sense to me.

15:50:32 <sandro> ericP, chucky is obsolete. See http://www.w3.org/2000/10/swap/test/ref/flatten-1e.n3 for existential, change the e to a 'u' for universal. or try 'em yourself in cwm.

15:50:39 <danbri> sandro, thanks for the example.

15:50:48 <ericP> tx

15:50:51 * danbri catches up

15:50:56 <danbri> willware, yes definitely interested

15:50:58 <WillWare> Having done my most successful tinkering with cwm, I might have an over-developed fondness for N3.

15:52:06 <ndw> DanC, following your earlier example with str:startsWith, should I expect this rule to assign the URI to the :bar property:{ :p a :Foo } log:implies { :p :bar :p.log:uri } .

15:52:09 <sandro> WillWare, do you need to re-write it all yourself because of the platform, or can you borrow heavily?

15:52:39 <JibberJim> danbri want's foaf to take over the mobile world, so of course he's interested!

15:52:41 <danbri> will, can you say a bit more about the constraints of the platform? I've been reading up a bit on cellphone stuff, Symbian, MIDP especially as I have a P800...

15:52:48 <WillWare> I can borrow heavily as long as it's a business friendly license. BSD is OK, GPL is bad, I think LGPL is OK.

15:53:02 <sandro> python, java, C, C++, ...?

15:53:15 * sandro laughs, imagining perl on a cell phone.

15:53:16 * chaalsMEL thinks (hopes) sandro is talking about another chucky.

15:53:35 <WillWare> Java and C. Alas, no Python, and _no_ Perl, thanks.

15:53:35 <sandro> "chunky" we briefly a name for a logic language.

15:53:42 <sandro> "chunky" WAS briefly a name for a logic language.

15:53:43 <danbri> I started looking around for a SAX2 XML parser that'll run on my phone, to plug into Sesame/Rio parser...

15:53:44 * chaalsMEL was looking at his cell phone and wondering what it had in teh way of programming languages available.

15:53:55 * ericP zakim, please mute me

15:53:57 <danbri> Java/C: would that be midp, personal java, ...

15:54:01 * danbri mutes ericp

15:54:08 * sandro thinks you'll need to justify not just using Jena2

15:54:09 <ericP> oops

15:54:23 <sandro> ericP is now muted.

15:54:30 <dajobe> no he's not

15:54:32 <sandro> ericP is now moot.

15:54:41 <danbri> sandro: Jena2/ARP is rather large, due to it's comprehensive I18N/Unicode support

15:54:45 <chaalsMEL> danbri, zakim is temporarily sandro

15:54:45 <ericP> curses, mooted again

15:54:49 <danbri> Rio is skinnier. Single threaded too...

15:54:58 <danbri> alrighty, enough with the w3t in jokes :)

15:55:00 <WillWare> Platform is an XScale processor. We run J2SE with tweaks to run on a mobile/small platform. I think we have midp, I hear a lot about midlets.

15:55:15 <danbri> ooh, j2se...

15:55:39 <danbri> problem i've run into is that interesting stuff needs to be able to talk to phone facilities -- gps, bluetooth, addressbook, camera.

15:55:39 <sandro> How about jython?

15:55:50 <danbri> which means Symbian, or else the ageing PersonalJava profile

15:55:59 <WillWare> I have to think a little, I want to make sure I don't say stuff I'm not supposed to.

15:56:07 <danbri> MIDP1 appears to be boringly sandboxed; and MIDP2 isn't on many phones yet.

15:56:10 <WillWare> Nope, not Symbian, http://www.savaje.com

15:56:21 <danbri> sorry, i meant 'which on my phone means...'

15:56:26 <chaalsMEL> danbri, you want to talk to the triple store, right?

15:56:37 <danbri> i do?

15:56:49 <chaalsMEL> ... in willware's outlined world.

15:57:08 <WillWare> Jython: there is some interest but nobody's done anything with it yet. I'd love to see it, maybe if I can find some time...

15:57:13 <danbri> sure, most likely to web services

15:57:25 <danbri> willware, some mobile folks hang out in #mobitopia

15:57:54 <danbri> my notes-as-i-learn are scribbled in http://rdfweb.org/topic/FoafMobile

15:58:41 <chaalsMEL> er, to the triple store and to the phone's interfaces (speech recogniser - another use case for a triple store perhaps?)

15:59:31 * DanC sees ndw's question...

15:59:43 <DanC> no, norm, you can't "assign" a URI.

16:00:17 <DanC> you relate two URIs by equality... => { ?P = ?Q }

16:00:40 <DanC> timbl tells me cwm has some built-in support for equality now.

16:00:43 <DanC> I haven't tried it.

16:01:30 * chaalsMEL noodles off into making the phone more accessible (to people with disabilities) by exposing the controllers through rdf, allowing people to make personal profiles that let them drive the phone through bluetooth etc.

16:02:11 * DanC notes WillWare's questions... oh for more hours in the day...

16:03:02 <WillWare> danbri: the foafmobile stuff looks very interesting, thanks.

16:03:08 * chaalsMEL suspects that's past where willware's going, but he may be interested in talking to people about accessibility. in which case INCITS V2 is a group that have done a lot of work on this.

16:03:52 <ndw> ok, thanks Danc, i'll dig a little deeper

16:04:23 <danbri> talking on incits v2, i just tried to get a replaement remote control for my tivo. 25 ukpounds! Crooks! roll on open device controller interfaces...

16:04:28 <DanC> log:uri is a rather pointy instrument. use with care.

16:05:02 <WillWare> There are people to whom I can pass along the accessibility thoughts. I'm planning to focus on getting RDF into the phone at all, gotta pick my battles.

16:08:49 <chaalsMEL> picking battles: sure.

16:09:14 <chaalsMEL> danbri, are you using the p800SDK on a real linux box or on a mac?

16:09:21 * chaalsMEL imagines the former :(

16:09:28 <danbri> linux

16:09:39 <danbri> needs Wine. Not sure about Wine under macosx.

16:09:47 <danbri> the pure Java stuff works on macosx.

16:09:55 * chaalsMEL owns a copy of virtualPC 5.

16:09:56 <danbri> see http://wireless.java.sun.com/midp/articles/osx/

16:10:02 <danbri> not same...

16:10:13 <chaalsMEL> correct...

16:10:32 <danbri> you do thinks within linux/perl like run 'wine ../sdk/tools/foo.exe'

16:10:43 <danbri> having a PC emulator in a separate window is wrong granularity...

16:10:51 <mdupont-work> mdupont-work is now known as mdupont

16:11:01 <danbri> though you could buy a windows set of C dev tools and do it 100% in the emu

16:11:36 <deltab> Wine on a Mac would need x86 emulation

16:13:32 <chaalsMEL> right.

16:13:44 <chaalsMEL> anyway, thanks for the sun.com pointer

16:14:02 * chaalsMEL goes to see if that's in the wiki page before going to bed.

16:14:40 <danbri> it is

16:15:22 <chaalsMEL> yep.

16:15:26 <chaalsMEL> Night...

16:17:12 <dmiles_afk> dmiles_afk is now known as dmiles_tmp

16:28:18 <sandro> anyone familiar with ISCW accommodations? How far the South Seas on Captiva is from the Sundial?

16:28:44 <sandro> er ISWC

16:46:02 <mdupont> anyone see bijan?

16:47:13 <DanC> A:so... is SAML kerberos with pointy brackets? or does SAML also include public key auth? I'll have to study it some more.

16:47:14 <dc_rdfig> Added comment A3.

17:00:27 <dmiles_tmp> dmiles_tmp is now known as dmiles_afk

17:03:56 <MikeM> which wiki engine is esw running?

17:04:17 <sbp> moin moin

17:05:06 * sbp flicks through his server via FTP, finds there's a bizzare number of copies of cwm...

17:05:36 <MikeM> hmm... wikibot is moin moin compatible.... any reason why we're using chump bot instead of wikibot?

17:05:47 <MikeM> just cause that's what's been used previously?

17:06:12 <sbp> yeah. rdfig.xmlhack.com outdates ESW by years

17:06:20 <sbp> predates, rather

17:07:20 <MikeM> that's true... I'm installing one for a blog I run and I was just wondering if there was some other compelling reason to use one over the other....

17:08:14 <sbp> I know quite a few channels that have a chump bot, but I don't think I've come across one with a wikibot yet... so I can't really pass judgement on how they compare, though I can tell you that chump works rather well

17:08:52 <DanC>http://www.w3.org/2002/09/Lists2RSS?ml=public-tag-announce

17:08:52 <dc_rdfig> C: http://www.w3.org/2002/09/Lists2RSS?ml=public-tag-announce from DanC

17:09:24 <DanC> C:|RSS 1.0 feed for TAG announcements

17:09:24 <dc_rdfig> Titled item C.

17:10:03 <DanC> C:among [http://www.w3.org/2002/09/Lists2RSS|RSS Feed from W3C Public Mailing Lists]

17:10:03 <dc_rdfig> Added comment C1.

17:38:50 <DanC> grumble... xchat's logging is great, until you want to find the friggin log file and mail it to somebody.

17:39:09 <DanC> in the context menu for a channel, I want a "copy the log filename to the clipboard" option

17:39:38 <DanC> better yet: if I drag from the channel tab to a mail compose window, figure out that I want to attach the log

17:39:55 <DanC> .google RemotePresence

17:39:56 <datum> RemotePresence: http://dm93.org/z2001/RemotePresence

17:40:03 <DanC> logger, pointer?

17:40:03 <DanC> See http://ilrt.org/discovery/chatlogs/rdfig/2003-09-12#T17-40-03

17:41:42 <JibberJim> DanC, I've been stunningly pleased with my Bluetooth headset for PC voip and phone work, works well, and lets me wonder away from PC whilst still talking.

17:42:01 <Wack> how far is the range?

17:42:02 <DanC> price?

17:42:12 <JibberJim> 42 UKP

17:42:35 <DanC> what bridges bluetooth to voip?

17:42:45 <JibberJim> range around 6M without problem, not tried further, but can go through walls for about 3

17:43:22 <JibberJim> this is just using through PC H323 soln., it just provides mic/headphone same as other.

17:43:47 <DanC> hmm... RemotePresence is out of date... I don't use the BE-3850BK any more; it has this annoying click. I use a cheapo 900mhz deal I got from a drugstore now.

17:44:20 <DanC> wow... PC H323 works? I got the impression it introduced a lot of latency. (.5sec when I last tried)

17:44:50 <JibberJim> it does a fair bit!

17:45:08 <DanC> fair bit... works a fair bit? or fair bit of latency?

17:45:33 <JibberJim> a fair bit of latency, too much for conversation, but I use it for demos where I'm mostly just speaking one way

17:46:20 <DanC> ok, that's consistent with my observations.

17:47:34 * DanC sees cheapo 900mhz deal is in RemotePresence after all, but only to note that I was trying it out.

17:49:52 <DanC> anybody know how to delete a table row/cell in evolution's editor?

18:20:59 <WillWare>http://willware.net:8080/semwebcd.iso

18:21:00 <dc_rdfig> D: http://willware.net:8080/semwebcd.iso from WillWare

18:21:25 <WillWare> D:|ISO file for a semweb CDROM I've been putting together, 360 MB

18:21:25 <dc_rdfig> Titled item D.

18:22:18 <WillWare> D: Includes lotsa open-source software, assorted tutorials, and various related educational material (pdf lecture notes and the like) that I could find on related topics.

18:22:18 <dc_rdfig> Added comment D1.

18:23:06 <WillWare> D: Part of my internal sales pitch for putting RDF into cellphones

18:23:06 <dc_rdfig> Added comment D2.

18:23:12 <mdupont> cool

18:23:40 <WillWare> thanks

18:24:26 <mdupont> i am struggling with a large database of rdf

18:24:35 <mdupont> i wish there was a fast way to deal with this data

18:24:54 <WillWare> That big 60000-statement thing from Introspector?

18:25:15 <zacker> hey there - anyone alive?

18:25:30 <mdupont> yes, WillWare, it is

18:25:43 <mdupont> that is the core of a c-based prolog engine in n3

18:25:47 <mdupont> just one function

18:25:59 <karlcow>http://scriptingnews.userland.com/backissues/2002/09/12#romanticWeb

18:25:59 <dc_rdfig> E: http://scriptingnews.userland.com/backissues/2002/09/12#romanticWeb from karlcow

18:26:03 <mdupont> the c function is 20k lines long

18:26:21 <karlcow> E:|"Why be Semantic when you can be Romantic?

18:26:21 <dc_rdfig> Titled item E.

18:26:21 <karlcow> "

18:26:45 <karlcow> E: You can be both... or at least I'm trying in my daily life :p

18:26:46 <dc_rdfig> Added comment E1.

18:27:35 <zacker> if someone is feeling especially generous with there time, im trying to figure out a sensible way to scrap this doc: http://www.blogforamerica.com/rightNav_syndicate.html and turn it into a RSS 2.0 feed, not sure how to approach this

18:27:51 * _joshua looks

18:28:09 <_joshua> which part?

18:28:12 <wkearney99> how about just asking them to make a feed of it?

18:28:16 <zacker> the whole thing

18:28:18 <zacker> well i am them

18:28:26 <_joshua> there's gifs and things there

18:28:30 <_joshua> what are you expecting?

18:28:31 <zacker> and thats what im trying to do :) but i dont think MT can export the right side nav to RSS

18:28:43 <wkearney99> so what're you making the page from now?

18:28:51 <dajobe> WillWare: could you make a listing of the iso so we can see the contents? (ls -lR if unix)

18:28:57 <wkearney99> it's in MovableType already, innit?

18:28:58 <zacker> MT generates it, but cant as RSS - atleast i havent found a way to get it to do that

18:29:14 <wkearney99> generates /what/ as rss?

18:29:20 <zacker> this doc: http://www.blogforamerica.com/rightNav_syndicate.html

18:29:34 <wkearney99> you want to turn the sidebar into an rss file?

18:29:37 <zacker> yep

18:29:40 <zacker> thats the plan

18:29:40 <wkearney99> why?

18:29:59 <wkearney99> it's not the sort of thing that "qualifies" as a feed.

18:30:05 <_joshua> i see a bunch of images. you want the link to be the link and the alt tag to be the title?

18:30:18 <zacker> so it can be syndicated / imported by the CMS's on the state / other campaign websites

18:30:18 <GabeW> wow, zacker, you are throwing the net wide!

18:30:27 <zacker> haha yo gabe!

18:30:54 <zacker> josh i just sat down and started reading RSS specs this morning, so im not really sure quite how to approach it

18:30:59 <_joshua> zacker: why not tell them to just include the file

18:31:08 <wkearney99> well, what are you expected to be providing here?

18:31:11 <_joshua> rss is just a list of links, with title and a slightly longer description

18:31:13 <zacker> because XML is better suited

18:31:16 <_joshua> this isn't a good mapping

18:31:24 <zacker> yeah, we will end up with some what of a custom schema i think

18:31:27 <_joshua> I think you'ev got a case of XY problem

18:31:28 <wkearney99> better suited to what?

18:31:30 <zacker> with HTML in it

18:31:37 <_joshua> Then that's got nothing to do with XML

18:31:41 <_joshua> er, nothing to do with RSS

18:32:04 <zacker> well, will RSS 2.0 with custom modules be a valid way to do this?

18:32:30 <_joshua> I think it'd be easier to just set your hair on fire.

18:32:33 <wkearney99> valid way to do /what/? reader programs don't typically do much with something like what you've given us as an example.

18:32:41 * wkearney99 runs and gets the matches...

18:32:42 <_joshua> why not let them include the file?

18:33:01 <_joshua> the problem with custom XML modules or whatever is that mostly readers just throw them out

18:33:02 <zacker> because XML is better suited for the job, gives them more flexibility

18:33:11 <_joshua> because in reality it's unworkably complex

18:33:12 <zacker> we dont really care about readers, we are making our own

18:33:23 <zacker> i mean - this isnt meant to be put in a news reader

18:33:23 <_joshua> zacker: then there's no need to use RSS either

18:33:35 <zacker> it is meant to be picked up by the readers on our CMS

18:33:50 <_joshua> Then write very simple HTML, put all the formatting in CSS

18:33:53 <zacker> well there is no reason it cant be basd on RSS right?

18:34:02 <wkearney99> sure, then just use a movabletype MTInclude directive and put the HTML into a template file.

18:34:02 <_joshua> aside from the fact that it makes no sense?

18:34:23 <wkearney99> Then make sure that same template file is written into a part of the filesystem (or web) that can be downloaded by the CMS.

18:34:41 <_joshua> it's just a bunch of pictures with some links

18:34:45 <_joshua> RSS isn't content markup

18:34:47 <_joshua> it's content metadata

18:35:12 <wkearney99> as in, your MT system gets setup with a template, that template writes itself to a disk file. Then your HTML page template uses an MT include to pull it into it.

18:35:15 <zacker> right... so this seems like a perfect application for metadata. each image is an image, and each image correlates with a link

18:36:11 <karlcow> E:I wonder if it's time to build an ontology about love

18:36:12 <dc_rdfig> Added comment E2.

18:36:21 <wkearney99> err, you could call it that (it would be an odd use of the term) but one could.

18:36:57 <wkearney99> so let's get this straight, you want to have your sidebar be something other sites can import into themselves, correct? And you're using MT to make this site, right?

18:37:08 <zacker> yes thats right wkearney

18:37:23 <zacker> the plan is to make a script that scrapes the html and dumps it wrapped in usable metadata

18:37:34 <zacker> that the CMS can then import and do fancy things with

18:37:42 <wkearney99> well, that's nothing to do with rss and it certainly doesn't need scraping.

18:37:51 <_joshua> Right

18:38:00 <WillWare> http://ilrt.org/discovery/chatlogs/rdfig/2003-09-12.html#T15-44-27|Discussion of RDF in cellphones

18:38:01 <dc_rdfig> F: http://ilrt.org/discovery/chatlogs/rdfig/2003-09-12.html#T15-44-27|Discussion from WillWare

18:38:11 <_joshua> Set up a second MT blog. Write things in that blog. Use SSI includes to include the output of the first in the second.

18:38:15 <zacker> ok well your right, i dont care what schema it is formatted in

18:38:41 <sethl> howdy all

18:38:56 <WillWare> Oops, chump blunder: how do I change the URI for link F?

18:39:17 <WillWare> F:|Discussion of RDF in cellphones

18:39:17 <dc_rdfig> Titled item F.

18:41:14 <WillWare> F: oops, URI broken, try again

18:41:14 <dc_rdfig> Added comment F1.

18:41:49 <JibberJim> F:=http://ilrt.org/discovery/chatlogs/rdfig/2003-09-12.html#T15-44-27

18:41:49 <dc_rdfig> Replaced URL of F.

18:42:47 <WillWare> Thanks, JibberJim

18:43:40 <WillWare> F1:""

18:43:40 <dc_rdfig> Deleted comment F1.

18:49:30 <GNUPredator> GNUPredator is now known as mdupont

18:49:48 <mdupont> msg: dc_rdfig ?

19:28:59 <mdupont> http://gollem.swi.psy.uva.nl/cgi-bin/pl-cvsweb/~checkout~/pl/packages/semweb/semweb.html?rev=1.3&content-type=text/html

19:29:16 <mdupont> I guess this is old news, the SWI prolog semantic web toolkit

19:29:31 * mdupont is trying to get the rdf editor running

20:06:40 <mdupont> mdupont is now known as mdupont-afk

20:38:47 <DanC> ah! nifty... I (re-)discovered the unregistered chrome: URI scheme, noted it in the wiki, and realized, with frustration, that I don't know where to look it up...

20:39:13 <DanC> ... then I realized the FollowYourNose pattern is justified by this frustration, and made the connection.

20:39:46 <DanC> so I'm already seeing benefits of synergy from putting this stuff in the wiki.

20:43:23 <DanC> anybody know where mozilla firebird sqirrels away client certificates?

20:43:40 <DanC> I just went thru a pile of trouble to get an MIT certificate; I want to print it out or something.

20:45:30 <DanC> aha! .phoenix, per http://texturizer.net/firebird/edit.html#profile

21:06:02 <DanC> tim, I just went thru the trouble of setting up an MIT SSL client certificate; now I can't figure out how to move it from one browser to another...

21:06:16 <timbl> You're way ahead of me.

21:06:26 <DanC> ... reminds me of my "how many private keys will folks be expected to manage?" question...

21:07:15 <DanC> ... when I first understood your proof-based-acl design, it seemed like a person would need lots of keys... one for each relationship/role/whatever...

21:07:27 <DanC> ... then recently I heard you say more like just one.

21:08:12 <timbl> Well, the point is you can chose how much trouble you want to go to to have different identiies

21:08:18 <DanC> (recently was in a discussion of spam and how public key crypto solves the problem, in theory)

21:08:52 <timbl> If you want to use the same identity to get an MIT parking permit and buy thing from Amzon, then so be it .

21:08:54 <DanC> doesn't the service provider determine how much trouble you have to go thru to delegate between keys?

21:10:00 <timbl> Well, we could design the system either way.

21:10:40 <timbl> Alan could requir Tanya to register her certificate wfrom Kari with Alan, to make change less easy.

21:10:54 <timbl> Or he could allow the web server to check the sert.

21:10:58 <timbl> cert.

21:11:24 <DanC> but if alan makes that requirement, there's nothing Tanya can do about it.

21:11:35 <timbl> What you really wanted to MIT to do was authenticate your normal SSH key as good for business withg MIT.

21:12:14 <timbl> Liken I did with you yesterday.

21:12:32 <timbl> I gather MIT only lets you certify a new key?

21:12:58 <DanC> they certainly make that easier than any other option

21:13:39 <timbl> Yes. Maybe the way the software is done.

21:13:57 <DanC> the way ssh, SSL, PGP, and S/MIME use public/private keys is different, of course. i.e. the keys are used to sign different sorts of data.

21:14:33 <timbl> You only need diff keys if teh security is going to be doiffrent from them ... like you will take more care of one, or will only give one to your ssh-agent.

21:15:04 <DanC> or if the consequences of one of them being lost or compromised are very different

21:15:12 <DanC> (same thing, perhaps)

21:15:31 <timbl> But keys are keys, right, mathematically? It is only the random non-stnadradness which stops you useing a PGP key for SSH etc.

21:15:51 <DanC> umm... yes, random non-standardness

21:15:53 <timbl> (different sides of the same cion)

21:16:06 <timbl> I haven't even got a MIT ID card.

21:16:09 <DanC> well, rsa keys and dsa keys are really different

21:16:21 <timbl> yes.

21:16:26 <timbl> and keys of different lengths.

21:17:03 <DanC> I noted (in the chump) a weblog entry from Eve Maler that SAML is done (oasis standard). SAML seems quite close to proof-based access control, at least in the use cases.

21:18:12 <DanC> i.e. it may be that saml:proof-based-acl :: font-tag:style-sheets

21:18:30 <DanC> or ws-policy

21:19:12 <DanC> hi reagle... what brings you hear today? I was just noodling out loud about proof-based access--- oops.

21:21:40 <DanC> oh... tim, I did manage to port my list of URI schemes to the wiki... and it has paid off.

21:21:42 <DanC> [[[

21:21:45 <DanC> why shouldn't I create a new scheme for XYZ?

21:21:45 <DanC> * New URI schemes are the only thing you can't FollowYourNose to look them up.

21:21:50 <DanC> ]]] -- http://esw.w3.org/topic/UriSchemes

21:22:17 <DanC> I only made the connection out of frustration when trying to look up the chrome: URI scheme.

21:22:33 <DanC> (well... not only... I also heard you say that yesterday)

21:26:01 <timbl> So the pages like http://esw.w3.org/topic/UriSchemes_2fnntp are generated automatically? virtual?

21:34:26 * timbl isn't quite sure how it is made but assumes that DanC is passing editorship to the Community in some way

21:35:28 <DanC> they were generated automatically, once, by http://www.w3.org/Addressing/schmg.py

21:35:42 <DanC> and yes, I'm passing editorship to the community

22:22:10 <DanC> woohoo! "Cut, copy, and paste text in text fields." in the Sep 2003 sidekick update! https://www.t-mobile.com/help/products/sidekick_updates.asp

22:22:29 <DanC> "Persistent notifications give multiple reminders of new messages and pending events."

22:22:53 <DanC> " Load Web pages in the background and receive notification when a page is finished loading."

22:23:06 <DanC> "Assign priorities to tasks and sort by priorities."

22:23:11 <DanC> "Specify locations for meetings and other Calendar events."

23:12:50 <danbri>http://siliconvalley.internet.com/news/article.php/3076961

23:12:50 <dc_rdfig> G: http://siliconvalley.internet.com/news/article.php/3076961 from danbri

23:13:10 <danbri> G:|Semantic Web: Out of the Theory Realm, Michael Singer (Sept 12 2003)

23:13:10 <dc_rdfig> Titled item G.

23:13:21 <danbri> G:"Beyond the great wall of data on the Internet lies a goldmine for enterprises called the Semantic Web."

23:13:21 <dc_rdfig> Added comment G1.

23:13:41 <GabeW> it feels like there's hole in the ozone right over myhead

23:13:46 <GabeW> oh man

23:13:48 <GabeW> wrong channel

23:18:56 <dmiles_afk> dmiles_afk is now known as dmiles_tmp