Semantic Web Interest Group IRC Chat Logs for 2002-07-12

12:27:53 <rreck> /msg NickServ IDENTIFY secret1

12:28:04 <rreck> oops

12:32:06 <rreck> i wonder how to *change* my pw

12:32:23 <larsbot> try /msg NickServ HELP

12:32:32 <rreck> yeah it doesnt say

12:32:45 <rreck> rreck has changed the topic to: RDF/Semantic Web 24x7 chat - http://rdfig.xmlhack.com/ blogit does

12:32:50 <rreck> ouch

12:33:35 <rreck> there we go

12:33:53 <rreck> now how do i fix the topic

12:34:23 <larsbot> just change it again

12:34:31 <larsbot> larsbot has changed the topic to: RDF/Semantic Web 24x7 chat - http://rdfig.xmlhack.com/ blogs it

12:34:37 <rreck> i dont know how i did it the first time

12:34:40 <rreck> heh

12:36:11 <larsbot> in XChat the topic appears in an input field at the top of the page, which I just change, and then press enter

12:37:27 <rreck> rreck has changed the topic to: RDF/Semantic Web 24x7 chat - http://rdfig.xmlhack.com/

12:52:39 <MorbusIff> MorbusIff is now known as Morbus

13:00:49 <Morbus> aaaah. danbri: i thought the perl thingies were simplistic FOAF parsers, not the ::RDFWeb stuff.

13:08:59 <jang> jang is now known as jang_jang

13:09:22 <jang_jang> jang_jang is now known as jang

13:16:49 <jordan> hey

13:17:25 <jordan> rreck: stayed up till 5am doing that

13:17:39 <jordan> rreck: finally got debian 2.2 running, then upgraded to woody

13:18:08 <jordan> rreck: yes but I'm still installing all the XML/lisp/XSLT/RDF related stuff and some of it decided to give me trouble :(

13:18:39 <jordan> .google xsl validator

13:18:40 <datum> xsl validator: http://www.redrice.com/ci/generatingXslValidators.html

13:27:14 <jordan> aaronl: where's that XSLT validator you once showed me? I can't find it today

13:29:17 <jordan> s/aaronl/aaronsw

13:45:29 <danbri> morbus, yup... the prob is that FOAF aggregators need to do some fancy-ish merging, so we need more than just a parser

13:47:42 <Morbus> what sort of merging?

13:48:01 <Morbus> (mind you: i've never been *much* of a fan of RDF, but I like FOAF a lot).

13:48:56 <danbri> merging that makes it possible for a.rdf, b.rdf and c.rdf to all mention someone (eg. you), yet do so in different ways. a.rdf might use your homepage, b.rdf might mention an old mailbox of yours, c.rdf might use both.

13:49:09 <Morbus> ah, gotcha.

13:49:17 <danbri> the merge would 'defragment' the rdf store so it could answer questions about you drawing on all three files

13:49:20 <Morbus> that should be relatively easy to do.

13:49:25 <Morbus> i do merging now of OPML files in AmpehtaDesk.

13:56:25 <larsbot> anyone here familiar with RDQL?

13:56:33 * larsbot has a non-working RDQL query

13:56:54 <Morbus> Ridicule? oh man.

13:57:23 * larsbot feels pretty ridiculed already, thank you

13:57:25 <larsbot> :)

14:01:48 <mhgrove> lars, i am a little

14:01:59 <mhgrove> emphasis on little though

14:03:47 <larsbot> a little may well be enough

14:03:50 <larsbot> here's my query: SELECT ?child, ?parent WHERE (?child, <tmann:parent_id>, ?parent) AND (?child, <rdf:type>, <tmann:type#3>) AND (?parent, <rdf:type>, <tmann:type#1>) USING tmann FOR <http://psi.ontopia.net/tmann/>, rdf FOR <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

14:04:17 <larsbot> basically, I want the child-parent relationships where child is of type 3 and parent of type 1

14:04:20 <larsbot> that's all I want

14:04:55 <mhgrove> seems reasonable, what happens when you run it?

14:05:12 <larsbot> I get no results

14:05:31 <mhgrove> are you sure there are children and parents with that relationship?

14:06:03 <larsbot> I'll check

14:06:41 <mhgrove> also, I dont think you need the AND's between the where clauses, i thought they are only comma separated

14:07:03 <mhgrove> which means i guess they are implicitly anded...*shrug*

14:08:10 <larsbot> actually, it wouldn't aprse with the ANDs, so I replaced them with commas

14:08:40 <mhgrove> oh, so the actual query doesnt have the ands?

14:09:02 <larsbot> no, I was showing an old version (sorry)

14:09:08 <larsbot> but the ANDs are the only difference

14:09:14 <mhgrove> oh its ok, i was just hoping the solution would be that easy ;)

14:10:13 <larsbot> but the query means what I thought it means, right?

14:10:42 <mhgrove> yeah...does that page exist? http://psi.ontopia.net/tmann/ ?

14:10:47 <larsbot> no, it doesn't

14:11:04 <mhgrove> that might be why you dont get back results?

14:11:20 <larsbot> as long as Jena doesn't try to download it all should be fine

14:11:25 <larsbot> because the URIs do exist in the model

14:11:48 <mhgrove> *shrug* i'm not a fan of Jena, so I'm not sure about how it works

14:13:12 <larsbot> web server log shows no accesses to that URI (except for yours :)

14:13:17 <larsbot> so apparently it's not doing that

14:14:29 <larsbot> hmmmm. taking out the last two ()s in the WHERE clause makes it work

14:17:30 <mhgrove> so just having where (?child, <tmann:parent_id>, ?parent) works?

14:17:41 <larsbot> yes, it does

14:18:12 <mhgrove> what happens if you add either of the other two clauses?

14:18:32 <larsbot> adding one worked

14:18:34 <larsbot> adding two did not

14:18:41 <larsbot> so apparently the problem is in the data

14:18:48 <mhgrove> so having two of the three works, but not all three?

14:19:02 <larsbot> yeah

14:19:15 <ambient> is the data that you're querying online?

14:19:33 <mhgrove> yeah i would tend to agree, the data doesnt seem correct

14:19:47 <mhgrove> cause your query is definately correct

14:20:14 <larsbot> the stuff is being autogenerated from some CSV files, which is what makes things tough

14:20:21 <larsbot> I'll keep digging

14:20:25 <larsbot> (thanks for your help, though :)

14:21:08 <mhgrove> lars, you are autogenerating rdf from csv files?

14:21:46 <larsbot> yeah (and XML, and JDBC, and all kinds of stuff)

14:21:53 <mhgrove> what are you using to do that?

14:22:06 <larsbot> the Ontopia Autogeneration Toolkit :)

14:22:30 <mhgrove> lol, i was just trying to work in a shameless plug for my own csv -> rdf tool =)

14:23:00 <larsbot> we have a framework for autogeneration projects, actually

14:23:12 <larsbot> it's going to become a commercial product, but isn't there just yet

14:23:31 <larsbot> it's becoming quite powerful, actually

14:24:34 <mhgrove> oh wow, that sounds pretty slick

14:24:44 <larsbot> we're becoming quite pleased with it

14:24:55 <mhgrove> my stuff is just hacked together for someone on my team to use, yours sounds much nicer =)

14:25:06 <larsbot> you write an XML file that sets up a chain of modules which then does an entire processing

14:25:20 <larsbot> well, mine is the result of several man-months of effort :)

14:25:43 <mhgrove> hehe

14:26:23 <larsbot> so it's not really apples and apples :)

14:27:47 <mhgrove> true true

14:28:36 <mhgrove> i imagine you can scrap a lot of RDF using that stuff?

14:30:11 <larsbot> oh yeah

14:30:20 <larsbot> I've done all kinds of stuff with it, which is quite nice

14:31:37 <larsbot> but I go obsolete-junk -> RDF -> topic maps, rather than junk -> RDF :)

14:32:35 <mhgrove> that's cool

14:33:03 <mhgrove> we have a db we're trying to convert to using RDF, so we've been trying to populate it with lots of triples so we can test

14:33:03 <larsbot> yeah. turns out RDF -> topic maps is not very hard

14:33:51 <larsbot> yup. it was the data. now it works just fine :)

14:35:17 <mhgrove> cool!

15:48:44 * dajobe stumbles upon an rdf/xml example in SOAP 1.2

16:29:36 <Morbus> hey, cool.

16:29:42 <Morbus> amphetadesk is listed as "current awareness" on http://www.dartmouth.edu/~library/search_internet.shtml

16:30:18 <sandro> So I've been thinking about how to webize FOL. I'm used to otter/prolog syntax, not LISP/KIF, but I don't think that matters here much. The obvious thing is to just use URI-Refs as your logical constants, but...

16:31:17 <sandro> but that truly hides the text of the URI-Ref from the system, which is not actually what you want. Looking inside URI-Refs is necessary -- without it, we could just use UUIDs for everything.

16:31:59 <DanCon> I was wondering if you'd get near stuff like log:uri (and log:semantics)

16:32:31 <DanCon> looking inside uri-refs is unnecessary, for most things, though

16:32:58 <DanCon> but we could look at tag issue 15, though... which asks:

16:33:00 <sandro> So the right thing is to define a small set of constants, like: zero, integerSuccessor, unicode, strFirst, strRest, emptyStr === that lets you build strings; then define log:uri mapping between those and objects.

16:33:20 <sandro> Then you can look inside the strings when you want to, if you want to, and make axioms about the contents.

16:34:27 <DanCon> re unicode, strFirst, etc.: I highly recommend doing that formalization while reviewing the charmod spec.

16:34:35 <sandro> It also lets us look inside RDF literals, and give them semantics if we want to, at no additional cost.

16:35:03 <sandro> I have read charmod enough to know I don't understand it. :-/

16:35:22 <DanCon> I started to do it in larch: http://www.w3.org/Architecture/theory/Character.lsl

16:35:39 <DanCon> ouch!

16:35:39 <sandro> logger, here

16:35:39 <sandro> See http://ilrt.org/discovery/chatlogs/rdfig/2002-07-12#T16-35-39-1

16:35:48 <sandro> ouch?

16:36:04 <DanCon> ouch: you don't find the charmod spec straightfoward

16:36:43 <sandro> Oh, it was more that it's just something that will take brain cells that I'm not sure I want to spend on it.

16:36:43 <DanCon> [I think the charmod spec should be split in 3; I think the TAG agreed, and asked them to do that]

16:37:18 <DanCon> yes, well, if you're gonna formalize stuff like strFirst and such, it's a shame to do so without reading the W3C Recommendation that (supposedly) says the same thing in English.

16:38:09 <sandro> Yeah.

16:38:10 <DanCon> I'm hoping the TAG is willing to do stuff like that... like saying "we think this 20 line formalization captures the formalizable aspects of that 70 page charmod spec"

16:38:23 * danbri grins at http://jibbering.com/rdfsvg/search.1?noun=Beer

16:38:24 <sandro> Interesting.

16:38:55 <sandro> This last two weeks of playing with FOL has made me increasingly suspecious of getting formal models right.

16:39:14 <xover> DanCon: Isn't that going a bit far towards Micro Management?

16:40:47 <DanCon> well, it's a good question, xover. One I spend my carreer persuing. "His research interest is investigating the value of formal descriptions of chaotic systems like the Web, especially in the consensus-building process." -- http://www.w3.org/People/all#connolly

16:41:45 <DanCon> my research thesis is, basically, if you can take a natural language spec that's supposed to be very widely deployed, and communicate it to a machine, you win big.

16:42:23 <DanCon> the machine can then check various properties of implementations, generate test cases, etc.

16:42:37 <DanCon> look at the benefits of BNF, for example.

16:43:06 <sandro> So the basic, underlying principles of URIs, it seems to me, includes (1) they mapped from strings, using one common/shared map (log:uri); (2) the strings are constructed such that different agents can get at the peices they need to (the different axioms used to peer into the log:uri string); (3) and different agents can construct them, with refined semantics, in a vaguely hierarchical fashion. (some thing, the other direction).

16:43:11 <DanCon> back to tag issue 15 and axioms involving peeking into URIs...

16:43:46 * sandro has to step away for a couple minutes; sorry. back soon I hope.

16:44:01 <DanCon> { :RES log:uri [ str:startsWith "http:"] } log:means { :RES a :HTTPRes }. <= allow me to define :HTTPRes thusly, ok?

16:46:13 <DanCon> then, recall a conjecture I made in discussion of tag issue 15: { <http://dm93.org/y2002/myCar-232> a :Car }

16:46:39 <DanCon> and timbl's conjecture: { :HTTPRes ont:disjointWith :Car }

16:46:51 <DanCon> those can't all be right.

16:48:47 * DanCon feels an obligation to make this into a real test case for the tag

16:49:41 * DanCon feels conflicted: I advised sandro to look at formalization of RDF literals and the charmod spec at the same time, but when the TAG reviewed the charmod spec, I didn't dust off my formalization stuff.

16:51:01 * DanCon feels really conflicted about using N3 in TAG discussions. It's a high-bandwidth mechanism for talking with TimBL, but we'd have to educate the rest of the crowd, not to mention the fact that N3's status as a formal notation is pretty suspect.

16:52:21 <DanCon> hmm... I could do it in rdf-logic

16:53:12 <xover> Am I totally out of line if I also suggest that the TAG has an obligation to not be needlessly opaque to the "web community" trying to follow the TAG's public discussion? (i.e. n3 is greek to me) :-)

16:53:38 <DanCon> no, I totally agree the TAG has an obligation to keep the community engaged.

16:53:47 <DanCon> as I said, "we'd have to educate the rest of the crowd"

16:54:06 * JibberJim thinks if TAG talk Greek every now and then it will help a portion of the "web community" who only speak greek.

16:54:15 <xover> Ok, I thought "rest of crowd" == "TAG members".

16:56:50 <mdupont> hello

17:19:21 <sandro> Yes, DanC, your { :HTTPRes ont:disjointWith :Car } test case looks like a reasonable formalization of the old debate.

17:21:35 <sandro> For myself, I'm trying to escape the debate by being ready to use multiple mappings from URI to objects, eg one mapping to digital artifacts (documents in broad sense), another to objects, another to communication end points, and more if necessary. Not as simple as we'd like, but a lot cleaner than confounding the 3+ mappings, I suspect.

17:24:47 <sandro> That is entityDescribedBy("http://www.w3.org/") is a Consortium, abstractDocumentAlwaysAt("http://www.w3.org") has several concrete forms in different languages and at different times, entityAnswering("http://www.w3.org/") is maybe a file, an apache server process, ... I'm not sure.

17:27:29 <sandro> But the model I most like is informationAbout("http://www.w3.org/#Consortium") which gives you back information in a content-negotiated lanugage about the consortium. informationAbout("http://www.w3.org/") would get you info about the document, but since the document varies so much due to cont-neg, it's kind of useless.

17:27:42 <sandro> ... this may all be off-topic.

18:53:28 * DanCon returns from running an errand...

19:08:40 <DanCon> more recent work on formalizing charmod: http://www.w3.org/2001/03swell/charmod.n3

19:08:50 <DanCon> v 1.3 2001/03/30

19:13:45 <DanCon> WOT use case: a travel admin just asked me for a credit card number, to reserve a hotel room for me. She actually doesn't need to know the number, so...

19:14:07 <DanCon> ... ideally, she'd give me a public key of the merchant, signed by visa

19:14:49 <DanCon> I'd check that sig, knowing a public key for visa, then encrypt my cc # for the merchant, and give that to the admin

19:15:54 <DanCon> or better yet: credit card numbers would just be public keys (or fingerprints) maybe

19:15:57 * xover wishes his bank would PGP sign/encrypt my bank statements and just email them...

19:17:02 <DanCon> but they'll only do that when a significant portion of their users grok PGP. I wonder if expecting consumers to manage public keys is too much.

19:17:29 <DanCon> I think little USB hardware tokens might do the trick.

19:17:47 <DanCon> meanwhile, smartcards are pretty widely deployed in europe, and the U.S. market just doesn't seem to grok.

19:18:04 <xover> With current tech, maybe. But who says we've reached the pinnackle of UI evolution?

19:18:39 <xover> I keep my SSH keys on a USB keychain and plug it into my keyborad for "single-signon" to all the SSH servers at work.

19:18:44 <DanCon> well, if you keep your secret key on your PC, you have to manage your PC securely. (e.g. guard against keystroke-stealing viruses). That's pretty close to impossible, no?

19:18:57 <xover> Impresses the hell out of the Windoze folk. :-)

19:19:33 <DanCon> the USB thing you use... does it just act like a disk? i.e. does the secret key go into the PC?

19:19:47 <xover> It's just a disk, yes.

19:20:12 <DanCon> ideally, the little gizmo would do the encryption stuff.

19:20:27 <DanCon> which raises the question: how does it know you've really told it to sign something?

19:21:12 <DanCon> maybe like this: it's a little USB dongle, about the size of a regular house key. When a request to sign something comes in, it lights up. you squeeze it to say "yes"

19:21:34 <xover> Current dogma on security is that authentication should require "Something you know, something you have, and something you are".

19:21:37 <DanCon> of course, then how do you know what you're signing? the light would have to include some hash of what you're signing or some such. sigh.

19:21:56 <xover> IOW, passwd + dongle + biometrics (retina/fingerprint).

19:22:16 <DanCon> yeah, well, how many passwords do I have to memorize?

19:22:42 <xover> Hopefully, just the passphrase for the USB thingy. :-)

19:22:45 <DanCon> a reasonable number would be: one per trusted hardware device.

19:23:59 <DanCon> that 3-way dongle says that a trusted device needs enough of a user interface to input a password, and enough hardware to do fingerprint/retina stuff.

19:24:03 <DanCon> sigh.

19:24:46 <xover> Given that these things are available in sizes up to several hundred megabytes...

19:25:07 <DanCon> the cost of such devices is still in the hundreds of dollars. It's got to get down to <$1 to take the place of, say, credit cards.

19:25:09 <xover> ...and that "convenient size for not loosing it" coincides with what you need for a fingerprint scanner...

19:25:32 <xover> Sure. $1 is a bit extreme, but in general, yes.

19:26:20 <DanCon> hm... what's the most expensive thing that "everybody" carries around? i.e. the most expensive thing that a bank could afford to give you when you set up an account with them?

19:26:34 <DanCon> I suppose paper checks cost around $10 per order.

19:27:12 <xover> A VISA costs me ~$10 per month...

19:27:35 <xover> (yes, I really should switch to a sane bank)

19:27:36 <DanCon> yeah... what I really want is something to replace the carbon copy feature of my checkbook: I want digital records of my credit card transaction right there inside the device as soon as I've executed the transaction; no waiting 5 days to look it up via http

19:28:47 <xover> Now we're talking smart-cards here!

19:28:57 <DanCon> yup

19:29:14 <xover> And pretty smart at that if it's to provide a transaction history that can be used for anything.

19:29:36 <DanCon> well, clearly no smarter than the USB disks.

19:29:48 <xover> (Intergrate with Quicken or GNUCash, say)

19:30:12 <xover> But the USB disks are much larger and don't double as a credit card.

19:30:27 * DanCon makes a link to this discussion from http://dm93.org/z2001/FractalAccounting

19:30:31 <DanCon> logger, pointer?

19:30:31 <DanCon> See http://ilrt.org/discovery/chatlogs/rdfig/2002-07-12#T19-30-31

19:31:49 <DanCon> I'm just saying megabyte storage technology in the palm of your hand is very feasible these days.

19:32:31 <xover> Sure, my 32MB cost ~400NOK (~$50).

19:33:27 <xover> Any one "killer app" would quickly drop them down to the price range you mentioned from the increased volume.

19:34:46 <xover> At work we're looking at buying a smart card system to replace our current key cards.

19:35:30 <xover> That project quickly led to the theory that we could use a combination of the smart card and a fingerprint scanner to authenticate users when they log in to their computers.

19:35:31 <DanCon> I'm starting to think that the market is somewhat unstable here... i.e. the most efficient solution doesn't benefit the entrenched parties. VISA is sitting pretty right now: they're almost the only game in town when it comes to convieneint transactions at a distance. Why would they deploy a totally decentralized solution, which would allow anybody to compete with them?

19:35:56 <xover> Very _very_ good point!

19:36:28 <xover> On-line terminals are expensive, and use a lot of weird tech that I suspect of being a way to maintain a monopoly.

19:38:06 <DanCon> of course, to be fair, VISA does perform a valuable service: they accept a little bit of risk for each transaction, and they sorta normalize a lot of policies, so that you don't have to have a protracted negotiation with each merchant.

19:38:40 <DanCon> (er... VISA actually just markets the results; the banks accept the risk)

19:39:56 <xover> Sure, and without the likes of VISA I would have had to carry around cash all the time. :-)

19:40:41 <xover> But so long as they have no incentive to change the status quo, you can't rely on them to drive future development. It just isn't in their best interest.

20:57:16 <Morbus> [[[

20:57:17 <Morbus> if RDF is about the "simple 3-part sentence", then why is the XML so verbose? I know someone murmured something about "the problem with RDF is XML", where else should be looking? N3? What's N3? more info?

20:57:20 <Morbus> ]]]

20:57:22 <Morbus> er, dammit.

20:57:27 <Morbus> alright, well, that's my question.

20:58:29 <deltab> N3 is RDF without the XML

20:59:02 <Morbus> and that's the stuff like ":deltab :memberOf :rdfig" I see every so often?

20:59:09 <deltab> right

20:59:29 <sbp>http://www.w3.org/2000/10/swap/Primer

20:59:29 <dc_rdfig> A: http://www.w3.org/2000/10/swap/Primer from sbp

20:59:39 <sbp> whoops. wrong channel...

20:59:45 <sbp> A:|Notation3 Primer

20:59:45 <dc_rdfig> titled item A

21:00:20 <Morbus> sbp, thanks. i've seen stuff also like :deltab :memberOf { :rdfig :channelIn :opn }.

21:00:27 <Morbus> is that a shortened form of two seperate statements?

21:00:40 <sbp> the {} bit is a formula

21:00:49 <sbp> it just allows you to talk about triples

21:02:24 <Morbus> is that explained in Primer?

21:02:38 <sbp> yep

21:02:55 <sbp> { :Morbus :getsBought [ :item :BritneysDanceBeat; :by :Sean ] } :seanThinksIs :RatherUnlikely .

21:03:27 <Morbus> why is everything prefaced by a :?

21:03:40 <Morbus> i can see :getsBought, but not Morbus, Sean, etc.

21:03:50 <sbp> the :blargh thing is just short for (usually) <#blargh>

21:04:04 <sbp> what it really means is "these terms aren't officially defined anywhere yet"

21:04:21 <Morbus> and if they were, they'd be rdfs:resource, foaf:name, ?

21:04:25 <sbp> right

21:04:29 <Morbus> k.

21:04:32 <deltab> do you know about XML namespaces?

21:04:36 <Morbus> mmhmm.

21:04:42 <deltab> same thing

21:04:42 <dfg_olin> sbp: Is there a builtin in cwm that gives the length of a list of arbititrary elements?

21:04:53 <sbp> dfg: lemme check

21:05:17 <Morbus> so... .. if N3 is so great, why is the XML so wordy?

21:05:23 <dfg_olin> The math builtin doesn't seem to work for anything except numbers.

21:05:45 <sbp> werid. math:memberCount should work

21:05:49 <deltab> N3 was invented for expressing RDF; XML wasn't

21:05:59 <sbp> yeah. XML is verbose for anything

21:06:07 <Morbus> what's wrong with such innocence as <morbus> <getsbought>Britney Dance Beat</getsbought> </morbus>

21:06:18 <dfg_olin> sbp: I sent you a mail about it.

21:10:43 <sbp> [[[

21:10:44 <sbp> $ n3s -s '{("x" "y" "z") math:memberCount ?x} => {?x a :Answer}' | cwm --think

21:10:45 <sbp> [...]

21:10:49 <sbp> "3" a <#Answer> .

21:10:53 <sbp> ]]]

21:11:03 <sbp> Morbus, so you prefer: [[[

21:11:03 <sbp> <rdf:Description>

21:11:03 <sbp> <log:is rdf:parseType="log:Quote">

21:11:03 <sbp> <rdf:Description rdf:about="#Morbus">

21:11:05 <sbp> <getsBought rdf:parseType="Resource">

21:11:06 <sbp> <by rdf:resource="#Sean"/>

21:11:08 <sbp> <item rdf:resource="#BritneysDanceBeat"/>

21:11:08 <sbp> </getsBought>

21:11:11 <sbp> </rdf:Description>

21:11:11 <sbp> </log:is>

21:11:15 <sbp> <seanThinksIs rdf:resource="#RatherUnlikely"/>

21:11:16 <sbp> </rdf:Description>

21:11:18 <sbp> ]]]?

21:11:24 <Morbus> no, I find that flipping satanic.

21:11:26 <sbp> admittedly, there's some CWM stuff in there (log:is?). the reified version would not be much better

21:11:28 <sbp> heh, heh

21:11:32 <Morbus> that's the exact reason I never looked into RDF in the first place.

21:11:43 <Morbus> because I was under the impression the XML was it's only output (not N3).

21:11:44 * sbp bows to the Morbus

21:11:55 <Morbus> and thought it was wordy, verbose, and made my code shui meter go off the deep end.

21:12:12 <deltab> that's why N3 was invented - XML was just too awkward

21:12:14 <Morbus> part of RDF/XML I didn't understand was the need for rdf:resource and all the URIs.

21:12:21 <sbp> it's nuts

21:12:28 <Morbus> why did it get nuts?

21:12:33 <sbp> welcome to the ever-growing list of people that hate it

21:12:35 <Morbus> what's wrong with my stupid innocent example?

21:12:46 <Morbus> welcome? i've been here for a year now. you know that ;)

21:12:56 <sbp> well, you have to look at the history of RDF, I suppose. it has a really long and twisted history, starting from PICS

21:13:03 <sbp> there have been other serializations proposed

21:13:10 <sbp> but the main fault lies in XML, I think

21:13:18 <deltab> Morbus: nothing defines what those tags mean

21:13:18 <sbp> XML just isn't good for encoding triples. too verbose

21:13:39 <Morbus> deltab: and what defines that in N3? an equiv of a doctype/schema?

21:13:57 <sbp> you use prefixes to bind namespaces to the little x:blargh deelies

21:14:06 <sbp> in other words, giving them all URIs

21:14:30 <sbp> @prefix rss: <http://purl.org/rss/1.0/> . is a bit like xmlns:rss="http://purl.org/rss/1.0/"

21:15:00 <deltab> and RDF statements describe the terms, relate them to others, etc.

21:15:14 <Morbus> ok. and the reason you can't <demo:morbus> ... innocent example ... </demo> is due to historical evolution?

21:16:26 <sbp> and that XML is overly verbose whatever you do

21:16:50 <Morbus> alright. so RDF and I got off on the wrong foot.

21:17:01 <sbp> whatever way you think of serializing RDF in XML, you always run afoul somewhere

21:17:02 <sbp> yeah

21:17:05 <Morbus> these other serializations - has anything come of them?

21:17:06 <sbp> RDF is just triples

21:17:15 <sbp> the other XML serializations? nope

21:17:22 <sbp> I've proposed a couple myself

21:17:23 * sbp rummages

21:17:40 <sbp> http://lists.w3.org/Archives/Public/www-rdf-interest/2002Mar/0128

21:17:44 <Morbus> so... .. why is RDFWeb focusing on FOAF in XML and not an N3 parser?

21:18:08 <deltab> I saw this - maybe it'll help: http://www.oasis-open.org/cover/xmlAndSemantics.html

21:18:24 <Morbus> yeah, sbp, that works for me.

21:18:30 <Morbus> it's along the same line i was thinking of.

21:18:39 <Morbus> where did it shoot itself in the foot?

21:18:50 <Morbus> deltab: thanks. reading.

21:20:26 <sbp> why FOAF uses XML RDF: because DanBri likes XML RDF. some people really do prefer it over Notation3

21:20:28 <sbp> I think it's just because they've been used to XML for so long

21:20:30 <sbp> honestly, I don't know why anyone would choose the XML serialization over N3 otherwise

21:20:53 <Morbus> are there n3 parsers? any in perl?

21:20:59 <sbp> yes, there is a Perl N3 parser

21:21:04 <sbp> hang on, there's a list somewhere...

21:21:25 <Morbus> how does n3 describe the "meaningless" data?

21:21:55 <Seth> morbus, if your looking for a straight forward searilization grammar, you might check out quads

21:22:12 <Morbus> url?

21:22:13 <sbp> on http://infomesh.net/2001/05/notation3/

21:22:26 <sbp> => http://www.cpan.org/authors/id/P/PC/PCIMPRICH/RDF-Notation3-0.50.tar.gz

21:22:36 <sbp> I'm not sure how good that parser is

21:22:38 <Morbus> 404.

21:22:52 <Morbus> it's .97 now.

21:22:58 <Morbus> er, .70.

21:23:04 <Seth>http://robustai.net/sailor/grammar/Quads.html

21:23:04 <dc_rdfig> B: http://robustai.net/sailor/grammar/Quads.html from Seth

21:23:35 <Seth> B:| Quads and alternative searilization for labeled directed graphs

21:23:35 <dc_rdfig> titled item B

21:24:04 <sbp> ah, cheers

21:24:17 <deltab> B:|Quads and alternative serialization for labeled directed graphs

21:24:17 <dc_rdfig> titled item B

21:24:17 <Morbus> you may just want to link to the dir.

21:24:21 <Morbus> it's pema.

21:24:23 <Morbus> perma.

21:24:52 <sbp> will do. /2001/05/notation3/ is a little out-of-date: I need to update it

21:26:03 <Seth> B: Quads are to semenglish as N-Triples are to rdf/xml

21:26:03 <dc_rdfig> added comment B1

21:28:22 <Seth> B: [http://robustai.net/mentography/semenglish.html|semenglish] easily written by humans, easily understood by computers

21:28:22 <dc_rdfig> added comment B2

21:31:20 <dfg_olin> sbp: Ok on your example, but why does ("x" "y" "z") a :List . blow up in math-test.n3?

21:33:15 <Seth> B: you could think of semenglish as n3 without all those dumb colons

21:33:15 <dc_rdfig> added comment B3

21:33:22 <sbp> [[[

21:33:22 <sbp> ( "x"

21:33:22 <sbp> "y"

21:33:22 <sbp> "z" )

21:33:22 <sbp> a <#List>;

21:33:22 <sbp> <#memberCount> "3" .

21:33:24 <sbp> ]]]

21:33:26 <sbp> works fine for me

21:33:34 <sbp> try checking out the latest version of CWM

21:34:15 <sbp> (that was the result from math-test.n3 unmodified: $ GET http://lists.w3.org/Archives/Public/www-archive/2001Nov/att-0095/03-math-test | cwm --think)

21:37:14 <dfg_olin> Did that. puzzled. does the math-test.n3 in the test directory with the latest cwm run for you?

21:39:57 <sbp> not, that math-test.n3 fails. "ValueError: invalid literal for float(): y"

21:41:19 <sbp> s/not/no/

21:41:59 <sbp> I'll hazard a guess and say that a variable isn't being bound somewhere, and that "y" is the frag of a URI-ref, not a literal

21:43:34 <sbp> cool tests in that file

21:46:12 <dfg_olin> Yes, I got the 'invalid literal error. Maybe you could put the good math-test in the CVS archive?

21:49:04 <Seth> B: RDF gives us no way to talk about a set of arrows; but quads, n3 and semenglish does. A set of arrows is alternatively called a context, a graph, a model, or a formula.

21:49:04 <dc_rdfig> added comment B4

21:50:12 <Seth> B: But without context, we'll never be able to do the semantic web .... well imho of course.

21:50:12 <dc_rdfig> added comment B5

21:50:55 <Seth> just wanted to get all that on the record and off my chest

22:01:07 <dfg_olin> sbp: thanks for the pointer . bye

22:28:41 <danbri> <sbp> why FOAF uses XML RDF: because DanBri likes XML RDF. some people really do prefer it over Notation3

22:29:31 <danbri> Because XML/RDF is the W3C recommended syntax for shipping RDF data around, and N3 is just something neat that TimBL made up and keeps changing and isn't intended as a mass market data format.

22:29:37 <danbri> No mystery.

22:34:10 <sbp> so, if N3 was a W3C rec and XML RDF was something that Tim hacked up, you'd go for N3?

22:34:18 <sbp> what if they were both recs?

22:35:23 <sbp> (good points about N3. I should have noted that I can't understand why people prefer the *syntax*, not the stability and status of the language)

22:35:50 * DanCon works on a rule ala: "if my team page says I'm travelling to WHERE WHEN, but my evolution calendar doesn't, complain"

22:36:10 <danbri> N3's cute, but its a jumble of all sorts of things. A nice plain text RDF syntax. An RDF representation of an abstract structure that isn't RDF (another richer FOL++ subset), and a random grab-bag of useful and not-so-useful predicates and hacks. It isn't current form REC material. So the question won't arise.

22:37:25 <DanCon> I want to reply to mcdermott: there are mostly-text languages (HTML, SGML, TeX, ...) and mostly-markup languages (C, LISP, etc.). the datatypes stuff looks funny because we're writing logical formulas, which are usually written in mostly-markup languages, in a mostly-text language.

22:38:03 <danbri> If XLink, XML Schema, XML Namespaces, XHTML, SVG, XSLT all also were ported over to the N3 universe, I'd take it more seriously. But we say RDF is part of the XML family of specs, and we shouldn't say it lightly. Offering a consistent message to the Web community re data on the Web is pretty important.

22:38:28 <DanCon> meanwhile, my ical2rdf.pl script and my team-page-schedule-scraper are working well enough that I might be able to propagate my Montreal trip from my team page to my evo calendar by machine...

22:38:35 <danbri> interestign angle

22:38:49 <danbri> thats a good sign :)

22:39:10 <DanCon> team-page-scraper is XSLT, btw, a benefit of living in the mostly-text world.

22:39:19 * danbri wants to go bug hunting in MaxF's xslt parser#5, but things he'll take a break

22:39:49 <DanCon> wohoo! check it out!

22:39:51 <DanCon>

22:39:51 <DanCon> <,evocal.rdf> :_needs [

22:39:51 <DanCon> ical:DESCRIPTION :_gs14;

22:39:51 <DanCon> ical:DTEND [

22:39:52 <DanCon> ical:DATE "2002-08-09" ];

22:39:54 <DanCon> ical:DTSTART [

22:39:56 <DanCon> ical:DATE "2002-08-04" ];

22:39:58 <DanCon> ical:LOCATION "Montreal";

22:40:00 <DanCon> ical:SUMMARY "Extreme Markup Languages, 2002" ],

22:40:26 * DanCon fixes :_gs14 bug...

22:41:54 <DanCon> there...

22:41:56 <DanCon> <,evocal.rdf> :_needs [

22:41:56 <DanCon> ical:DESCRIPTION "file:/home/connolly/w3ccvs/WWW/2002/08dc-ymx/status-formal.rdf#L3846";

22:41:56 <DanCon> ical:DTEND [

22:41:56 <DanCon> ical:DATE "2002-08-09" ];

22:41:56 <DanCon> ical:DTSTART [

22:41:57 <DanCon> ical:DATE "2002-08-04" ];

22:41:59 <DanCon> ical:LOCATION "Montreal";

22:42:01 <DanCon> ical:SUMMARY "Extreme Markup Languages, 2002" ],

22:42:28 <deltab> <,evocal.rdf> ?

22:43:07 <DanCon> ,evocal.rdf: $(HOME)/evolution/local/Calendar/calendar.ics $(PALMAGENT)/vcal2xml.pl

22:43:08 <DanCon> $(PERL) $(PALMAGENT)/vcal2xml.pl $(HOME)/evolution/local/Calendar/calendar.ics >$@

22:43:35 <DanCon> i.e. it's generated by vcal2xml.pl from http://dev.w3.org/cvsweb/2001/palmagent/

22:43:56 <DanCon> it's my evolution calendar.

22:43:58 <deltab> I just don't often see filenames starting with a comma

22:44:14 <DanCon> start-with-comma is a temp-filename pattern I picked up from the MH tools

22:44:24 <deltab> ah

22:44:29 <DanCon> it means "don't check this file in and share it with the world!" to me

22:45:52 <DanCon> danbri, have you seen the semweb plan page schedScrape.xsl thingy? i used that on my team page to create status-formal.rdf; then I made ,evocal.rdf as above, then I wrote rules ala "if a trip is in sched-formal but not in evocal, conclude evocal needs it"

22:46:21 <danbri> ooh. No, I've not.

22:46:38 <danbri> I think I need some new desktop tools. Maybe evolution, or see how Mozilla calendar coming along.

22:46:49 <danbri> can you generate a public view of semweb plan?

22:46:52 <DanCon> I started to show it (schedScrape.xsl) to you a while ago

22:47:13 <DanCon> public view: that's a policy question. Given the relevant policy rules, sure, I can do it.

22:47:15 <danbri> Rings a bell. You've shown me a lot of stuff though :)

22:47:57 <danbri> That'd be useful. Lots of people are listing RDF/SW stuff. Would be good to have a reasonably comprehensive public event feed.

22:48:16 * danbri makes note to look at this, heads off for the night

22:48:31 * DanCon chacls schedscrape...

22:49:08 <DanCon> the stuff on the semweb plan is not intended to be comprehensive. And it's pretty sensitive stuff. like who'se travelling where when.

22:49:27 <DanCon>http://www.w3.org/2001/sw/schedScrape.xsl

22:49:27 <dc_rdfig> C: http://www.w3.org/2001/sw/schedScrape.xsl from DanCon

22:49:56 <DanCon> C:schedScrape.xsl -- scrape/formalize schedules from HTML to RDF

22:49:56 <dc_rdfig> added comment C1

22:50:10 <DanCon> C:needs a public example. hmm... maybe my homepage

22:50:10 <dc_rdfig> added comment C2

22:50:57 <DanCon> C:maybe could be adapted to do [W3C upcoming appearances|http://www.w3.org/Promotion/Appearances/]

22:50:57 <dc_rdfig> added comment C3

22:51:15 <danbri>http://www.mozilla.org/projects/calendar/

22:51:16 <dc_rdfig> D: http://www.mozilla.org/projects/calendar/ from danbri

22:51:20 <DanCon> huh... " * 8 August - Liam Quin presents XML Reduced"

22:51:41 <danbri> D:|Mozilla Calendar project, seems to have life in it again!

22:51:41 <dc_rdfig> titled item D

22:52:03 <DanCon> "The features that should be removed include NOTATION, ENTITY, CDATA marked sections, DOCTYPE and processing instructions." go Liam!

22:52:04 <danbri> D:Includes some calendar examples in downlaod format (us holidays etc).

22:52:04 <dc_rdfig> added comment D1

22:53:22 * xover bets aray would have a thing or two to say on /that/ subject... :-)

22:53:56 * DanCon has trouble appreciating aray's worldview/motivation

22:54:38 <DanCon> C|schedScrape.xsl -- scrape/formalize schedules from HTML to RDF

22:54:43 <DanCon> C:|schedScrape.xsl -- scrape/formalize schedules from HTML to RDF

22:54:43 <dc_rdfig> titled item C

22:55:37 <DanCon> danbri, would RDF versions of those holiday files be interesting/useful?

22:56:02 <danbri> f'sure :)

22:56:24 * danbri trying to install mozilla calendar in 5 mins

22:56:28 * DanCon crosses fingers, tries nasty perl script...

22:56:59 <deltab> "downlaod format"? is that meant to be "iCal format" or "for download"?

22:57:11 <danbri> ah, tip: Mozilla needs to run as root when installing .xpi extension bundles

22:57:40 <danbri> ...but root linux cfg of mozilla has javascript turned off. Which is exactly what pages like http://www.mozilla.org/projects/calendar/# don't seem to expect

22:58:23 <DanCon> those holiday files look bogus

22:58:50 <DanCon> X

22:58:50 <DanCon> ;MEMBER=AllDay

22:58:50 <DanCon> :TRUE

22:58:58 <DanCon> wtf is that?

23:00:27 <xover> DanCon: You don't happebn to have an URL handy for "something" that discusses why Liam's Axe List of features are "Considered Harmfull"?

23:00:33 <danbri> bogus per ical?

23:01:03 <deltab> presumably part of the CLASS

23:01:25 <deltab> or maybe that means something different

23:02:35 <DanCon> no, xover, but look at XSLT, RELAX-NG, etc; all they care about are elements, attributes, and characters. And namespaces. Nothing else is really worth keeping around, to me.

23:02:57 <DanCon> ah... tim bray's skunkworks xml...

23:03:14 <DanCon> XML-SW, a thought experiment

23:03:14 <DanCon> From: Tim Bray (tbray@textuality.com)

23:03:14 <DanCon> Date: Wed, Feb 06 2002

23:03:19 <DanCon> -- http://lists.w3.org/Archives/Public/www-tag/2002Feb/0031.html

23:03:44 <DanCon> that should do, xover

23:04:13 * danbri runs into problems running mozilla calendar :(

23:04:17 <danbri> giving up for now

23:04:27 <DanCon> paste diagnostics?

23:04:30 <xover> If that's the kind of thread I think it is, then, yes, it should do more then nicely! :-)

23:04:36 <xover> Thanks!

23:04:55 <DanCon> welcome.

23:05:20 * DanCon consults ical RFC to see who's right: my perl script or the mozilla holiday data

23:06:20 * DanCon thinks the bug is in the way I folded lines, maybe...

23:09:52 <DanCon> no, it's in the way I handle attributes; I need a certain amount of schema-ish knowledge about each field, and I didn't have it for 'X'

23:10:26 <DanCon> phpht. they put N calendars in one file

23:10:38 <DanCon> my code exits when it sees END:CALENDAR

23:10:42 * DanCon checks the RFC again...

23:11:58 <DanCon> "However, multiple

23:11:58 <DanCon> iCalendar objects can be sequentially grouped together."

23:12:10 <DanCon> -- 4.4 iCalendar Object http://www.imc.org/rfc2445

23:12:41 <DanCon> no semantics is given to the syntactic concatenation of calendars. harummph

23:15:06 * DanCon tries the obvious fix, loses

23:20:06 <DanCon> "# iCal2xcs.xsl convert xml to xcs" -- http://www.jsoft.com/Gary/calendar/ hmm...

23:24:33 * danbri gets mozilla calendar running

23:24:39 <danbri> quite pretty, seems to work

23:24:49 <DanCon> "Using RDF with the mozilla calendar" 04/21/02 09:27 in news://news.mozilla.org/netscape.public.mozilla.calendar

23:27:35 <danbri> oh

23:27:44 * danbri subscribes to that newsgroup

23:27:50 <danbri> I'm starting to be convinced by mozilla

23:27:55 <DanCon>news://news.mozilla.org/netscape.public.mozilla.calendar

23:27:58 <DanCon> phpht

23:28:07 <danbri> news, mail, www, irc work

23:28:14 <DanCon> dc_rdfig, news://news.mozilla.org/netscape.public.mozilla.calendar is a perfectly good URI. pls chump it

23:28:14 <danbri> chump doesn't like all uris

23:28:43 <danbri> D:See also [news://news.mozilla.org/netscape.public.mozilla.calendar|mozilla newsgroup]

23:28:43 <dc_rdfig> added comment D2

23:28:46 <danbri> that might do it

23:28:47 <DanCon> aha.

23:28:49 <DanCon> well played

23:28:58 * DanCon is neglecting family obligations...

23:29:18 <xover> "In a sense, we've already stabbed [the SGML community] in the back[...]" - Norman Walsh

23:29:35 <xover> Well, so long as we're in agreement on that much, I'm happy. :-)

23:30:16 * DanCon figures this is worth chumping...

23:30:45 <DanCon>http://www.extrememarkup.com/extreme/2002/thursday.asp

23:30:45 <dc_rdfig> E: http://www.extrememarkup.com/extreme/2002/thursday.asp from DanCon

23:30:54 <DanCon> E:|Daily Polemic: XMLr (XML Reduced)

23:30:54 <dc_rdfig> titled item E

23:31:13 <DanCon> E:an [upcoming appearance|http://www.w3.org/Promotion/Appearances/] by Liam Quin

23:31:14 <dc_rdfig> added comment E1

23:32:04 <DanCon> E:see also: [XML-SW, a thought experiment|http://lists.w3.org/Archives/Public/www-tag/2002Feb/0031.html] by Tim Bray in www-tag, Feb 2002, and the following thread.

23:32:04 <dc_rdfig> added comment E2

23:32:24 <DanCon> there, xover, now you can add your back-stab comment

23:32:33 <xover> :-)

23:32:46 <xover> What does chump do with [] that aren't significant?

23:33:16 <DanCon> dunno

23:33:46 <xover> E:"In a sense, we've already stabbed [the SGML community] in the back[...]" - [Norman Walsh|http://lists.w3.org/Archives/Public/www-tag/2002Jan/0035.html]

23:33:46 <dc_rdfig> added comment E3

23:33:50 <xover> Let's try it...

23:34:18 <xover> Hmm. Not very good...

23:34:50 <xover> E3:"In a sense, we've already stabbed \[the SGML community\] in the back\[...\]" - [Norman Walsh|http://lists.w3.org/Archives/Public/www-tag/2002Jan/0035.html]

23:34:50 <dc_rdfig> replaced comment E3

23:35:41 <DanCon> nice try ;-)

23:35:53 <xover> E3:"In a sense, we've already stabbed (the SGML community) in the back(...)" - [Norman Walsh|http://lists.w3.org/Archives/Public/www-tag/2002Jan/0035.html]

23:35:53 <dc_rdfig> replaced comment E3

23:35:59 <xover> Oh well... :-(

23:36:16 <xover> E:Well, so long as we're in agreement on that much, I'm happy. :-)

23:36:16 <dc_rdfig> added comment E4

23:39:17 <DanCon> hmm... they're talking about http://www.ietf.org/internet-drafts/draft-ietf-calsch-many-xcal-01.txt in the mozilla cal project... maybe an opportunity to introduce RDF? hmm...

23:39:58 <xover> Isn't that Hixie's domain?

23:41:06 <xover> Hmm. No. "Dan Brickley" was the name that rung a bell on http://www.mozilla.org/rdf/doc/.

23:41:48 <DanCon> E:hmm... they're talking about [iCalendar DTD Document (xCal)|http://www.ietf.org/internet-drafts/draft-ietf-calsch-many-xcal-01.txt]

23:41:49 <dc_rdfig> added comment E5

23:42:16 <DanCon> E:ugh... NOTATIONs and such

23:42:17 <dc_rdfig> added comment E6

23:43:06 * xover still hasn't figured out what's wrong with NOTATION...

23:43:11 * Hixie hears his name spoken and looks up

23:43:13 <xover> "...and such..."

23:43:28 * xover apologizes for the disruption!

23:43:46 <DanCon> E:hmm... wonder what RELAX-NG DTD import tools, or [XML Schema tools|http://www.w3.org/XML/Schema#Tools] would make of that DTD

23:43:46 <dc_rdfig> added comment E7

23:43:48 <xover> Apparently, it was danbri I was really trying to summon.

23:44:19 <DanCon> notation lives in its own world, outside elements/attributes/characters/namespaces, for no good reason

23:47:05 <xover> Accepting for the moment that DTDs and similar SGML baggage er evil[tm], has anyone made an attempt at "normative"-quality Schema for, say, XHTML?

23:47:21 <DanCon> yup...

23:47:42 <DanCon> Modularization of XHTML? in XML Schema

23:47:42 <DanCon> 19 December 2001, Daniel Austin, Shane McCarron

23:47:43 * danbri hears his name spoken and runs away

23:47:47 <DanCon> -- http://www.w3.org/TR/xhtml-m12n-schema/

23:47:56 <danbri> yup, i help maintain the Mozilla RDF pages

23:48:05 <danbri> even updated the faq (slightly) last week

23:48:36 <xover> Is that WD dead, or is another update due soon?

23:48:40 <DanCon> I've seen relax-NG schemas for XHTML too, but I'm not sure how fare they went

23:48:48 <xover> (published on my birthday, no less!) :-)

23:49:01 <danbri> I had an idea for introducing RDF into the iCal/mozilla discussions without being pushy: show some Event-augmented RSS feeds being consumed in Mozilla (with XUL)

23:49:19 <danbri> if we can get vanilla RSS one showing, the event properties shoudl be easy to add

23:49:29 <danbri> ...and once in mozilla, slurping them over into the cal can't be _that_ hard

23:49:32 <danbri> he says naively

23:49:46 <DanCon> sound great, danbri

23:49:52 <xover> Aren't the mail/news bits pretty RDFified?

23:50:14 <DanCon> that WD shows little signs of life. dunno.

23:50:23 <danbri> yeah, I thought it sounded great too. Dunno if I'll get to do it myself.

23:50:35 <xover> Who are good people to ping about current status?

23:50:38 <danbri> mail/news -- they were more RDFy than they ended up, I think.

23:51:08 <danbri> not sure. mail to mozilla-rdf often goes unanswered. Tracking down waterson et al in #mozilla over there might be worthwhile.

23:51:09 <DanCon> ah... it's scheduled for last call Jul 2002, per http://www.w3.org/MarkUp/xhtml-roadmap/ . I doubt that particular date, but that's a sign of life

23:51:12 * danbri lurks there

23:51:41 <xover> DanCon: Cool, Thanks!

23:52:11 * DanCon really must go now.

23:52:13 <xover> danbri: #mozilla on...? Here (i.e. OPN)?

23:52:34 * xover waves goodbye at DanCon...

23:54:09 <danbri> irc://irc.mozilla.org/mozilla

23:54:28 <danbri> also irc://irc.mozilla.org/mozillazine is friendly

23:54:36 <danbri> ...better for questions, I believe

23:56:08 <xover> Ah, thanks.