00:00:10 sorry, 2.59KB 00:00:28 Neat. 00:03:37 O.K., now I need some really crud HTML to test it on 00:03:56 it doens't have tidy built-in? 00:04:10 It doesn't need it 00:04:26 I hope :-) That's why I need some crud HTML, to test... 00:07:01 aha:- 00:07:02 [[[ 00:07:02 THE SUBARU WEBSITE IS BUILT FOR INTERNET EXPLORER 00:07:02 VERSION 4 OR HIGHER AND NETSCAPE VERSION 4 OR HIGHER. SCREE 00:07:02 N RESOLUTION MUST BE 800x600 OR HIGHER. CLICK ON THE APPROPR 00:07:02 IATE LINK BELOW TO DOWNLOAD THE MOST CURRENT BROWSER VERSION. 00:07:04 < HERE> TO ENTER 00:07:06 ]]] 00:07:08 - http://www.subaru.com.au/ 00:07:33 of course, I can change the browser ID :-) 00:11:27 tansaku (~sam@h131-044.tokyu-net.catv.ne.jp) has joined #swhack 00:11:43 Hmm... somehow, I need to support frames 00:15:27 Perhaps you can convince deltab to switch from Links to browser.py 00:16:23 lol 00:36:47 * sbp adds better formatting rules 00:36:53 now it doesn't do st 00:36:57 uff like this wh 00:36:59 ich is good 00:37:08 s/this/this,/ 01:06:46 Hmm... it's putting line breaks in
elements for no apparent reason 01:16:11 Hmm... sometimes it adds a /n after
  • , and sometimes it adds two 01:22:29 Aha, solved both problems 01:29:27 sbp has quit (Killed (NickServ (Ghost: SeanP!~sean@m573-mp1-cvx3b.pop.ntl.com))) 01:29:45 sbp (~sean@m573-mp1-cvx3b.pop.ntl.com) has joined #swhack 01:33:39 tansaku has quit (Read error: 110 (Connection timed out)) 02:11:10 Hmm, the DAML/RDF layering discussion is interesting. 02:19:48 * AaronSw sends a message to Pat/www-archive about it. 02:21:21 Pat Hayes on Pat Hayes: "He also restores antique mechanical clocks, remodels old houses and enjoys arguing with anyone about almost anything." 02:21:25 - http://www.coginst.uwf.edu/~phayes/ 02:24:07 Roy Fielding on why he's in the TAG: "[To] Figure out what W3C WGs have been working on while I've been out of touch..." 02:27:55 Ian says the IRC logs will be public, but they're member-confidential right now... 02:28:29 my browser has a first name | it's browser dot pee why | my browser has a second name | it's browser dot pee why 02:29:03 Yeah, but it's nice that the minutes are public 02:29:40 That's true. 02:31:50 * AaronSw wonders who daveo100 is in the IRC log 02:31:56 must be David Orchard 02:37:43 lol: Dan Connolly: If the entire web disappears behind "POST" due to web services, I will be annoyed. 02:37:47 That's an understatement. 02:37:51 heh, heh, heh 02:43:06 lol: http://lists.w3.org/Archives/Public/www-archive/2002Jan/0014 02:43:13 Go Joseph Reagle, go! 02:44:46 lol, neat 02:49:08 The custom Google setup for the W3C is pretty cheap, IMO. I mean, passing HTML thru the URL? 02:49:46 Come again? 02:49:52 deus_x has quit (forward.openprojects.net irc.openprojects.net) 02:49:57 like http://www.google.com/custom?cof=T%3Ablack%3BLW%3A72%3BALC%3A%23ff3300%3BL%3Ahttp%3A%2F%2Fwww.w3.org%2FIcons%2Fw3c_home%3BLC%3A%23000099%3BLH%3A48%3BBGC%3Awhite%3BAH%3Aleft%3BVLC%3A%23660066%3BGL%3A0%3BAWFID%3A0b9847e42caf283e%3B&sitesearch=w3.org&domains=w3.org 02:50:25 heh, yeah... 02:50:35 Funny when you take that bit out 02:50:40 or even better, replace it 02:50:46 Heh. 02:51:15 Hmm... nothing happens when you replace it. That sucks 02:51:47 Hmm, that's really weird. 02:52:02 >>> import urllib 02:52:02 >>> urllib.unquote("T%3Ablack%3BLW%3A72%3BALC%3A%23ff3300%3BL%3Ahttp%3A%2F%2Fwww.w3.org%2FIcons%2Fw3c_home%3BLC%3A%23000099%3BLH%3A48%3BBGC%3Awhite%3BAH%3Aleft%3BVLC%3A%23660066%3BGL%3A0%3BAWFID%3A0b9847e42caf283e%3B") 02:52:02 'T:black;LW:72;ALC:#ff3300;L:http://www.w3.org/Icons/w3c_home;LC:#000099;LH:48;BGC:white;AH:left;VLC:#660066;GL:0;AWFID:0b9847e42caf283e;' 02:52:03 >>> 02:52:19 Ah. 02:52:22 now *that's* really weird 02:52:33 It looks like they just signed up for the Free SiteSearch service. 02:52:37 Yeah 02:52:59 Good idea, too 02:53:28 You should do that for Infomesh or something. 02:54:06 heh: http://www.google.com/custom?cof=&domains=w3.org&q=blargh&sitesearch=w3.org 02:54:13 infomesh? Why bother? 02:54:28 I'll probably do it for mysterylights or something 02:54:29 deus_x (~deusx@bgp993973bgs.nanarb01.mi.comcast.net) has joined #swhack 02:55:26 W3C should upgrade to Google Gold. 02:55:41 Google Gold? What do you get with that? 02:55:47 See http://www.google.com/services/silver_gold.html 02:55:56 You get to put any sort of HTML in. 02:56:18 Neat 02:56:51 ooh: http://www.google.com/custom?cof=&domains=w3.org&q=blargh&sitesearch=w3.org 02:57:33 what's the difference? 02:57:48 Note who all of the search results are by 02:58:05 of course 02:58:09 Ooh: http://www.google.com/custom?cof=&domains=infomesh.net&q=blargh&sitesearch=infomesh.net 02:58:31 Heh, what? 02:58:41 Heh heh: http://www.google.com/custom?q=blargh&sitesearch=logicerror.com 02:59:00 Ah, they're still down to me :-) 02:59:21 and all of those in: http://www.google.com/custom?q=blargh&sitesearch=blogspace.com 02:59:22 hmm, custom looks nicer than the default 03:01:29 Hmm, I find blargh more common than ping, contrary to what the Jargon File says. 03:01:50 I find that the meaning of "blargh" according to the Jargon File is wrong 03:02:06 Really? 03:02:07 It is now quite clearly a metasyntactic variable first and foremost. Get with the times 03:02:27 Only in your phenomic little world. ;-) 03:02:42 blargh 03:03:49 Cool, Roger was accepted. http://www.crypto.com/papers/fc02accepted.txt 03:04:58 ooh, someone needs to buy semanticweb.ai 03:05:10 even better: semanticweb.is.ai 03:05:57 i wonder what country .ai is 03:07:03 or thesw.is.ai 03:07:07 Hmm... no 03:07:15 http://the.semantic.web.is.ai/ 03:07:31 http://the.semantic.web.is.ai/so/f**k/you 03:07:36 heh heh 03:07:39 :-) 03:07:54 and its rival site: http://the.semantic.web.is.not.ai/so/f**k/you 03:09:38 Oh feckle, I can't get preformatted text to work properly 03:11:20 ah, I think I've got it 03:11:47 What an odd error: 03:11:48 dyld: /Users/aaronsw/Sputnix.app/Contents/MacOS/Sputnix can't get realpath of executable: /Users/aaronsw/Sputnix.app/Contents/MacOS/Sputnix (No such file or directory, errno = 2) 03:11:51 I need to make it spew the text out earlier so that it can recognize that it's still part of a preformatted block 03:11:58 Ooh, that is odd 03:12:00 dyld? 03:12:14 dynamic loader 03:12:36 i'm having problems like that all over. 03:12:38 Ooh, browser.py is now 3.86KB 03:12:44 like i can't even hopen the dyld man page! 03:12:52 perhaps I should reboot. 03:13:43 sbp: where is that? 03:14:10 Where is what? If you mean browser.py, it's not on the Web yet 03:14:26 shucks 03:14:47 Don't worry, it's no contest for Links 03:15:18 Aha! Preformatting works 03:16:20 Hmm... urllib is being really slow in getting the page 03:17:11 I'll probably switch to httplib at some point, anyway 03:18:48 Heh, forgot to add "span" to the list of inline elements... 03:20:16 actually, I guess that's a stupid way to do it 03:20:37 Most browsers render anything that they don't know, and just exclude the ones in the etc. 03:21:43 of course, I also want people to be able to configure it, and perhaps to eventually recognize stylesheets 03:23:10 Ooh, funny: it barfs on the news item headers on the W3C homepage 03:23:22

    Some text

    03:23:49 why? 03:23:54 I'm not sure 03:24:13 It just gives "@ " as the output, instead of "@ Some text" 03:26:11 Hmm... double line break in front of items. Not good 03:27:53 Wow, making a browser is very easy and very difficult at the same time 03:29:14 sbp has quit (Killed (NickServ (Ghost: SeanP!~sean@m641-mp1-cvx3b.pop.ntl.com))) 03:29:32 sbp (~sean@m641-mp1-cvx3b.pop.ntl.com) has joined #swhack 03:33:54 ugh:

    An Aaron Swartz Project

    03:33:57 small? small? 03:34:08 on blogspace.com 03:34:24 I only noticed because my browser doesn't recognize crap like 03:34:59 tonight you're going to hear music you never heard before! 03:35:07 come again? 03:39:44 There we go: I changed it so that it spews out all data now except in the (and except for , which is also spewed) 03:39:55 <sbp> s/and // 03:40:26 <sbp> Hmm... at the moment, we get stuff like:- 03:40:27 <sbp> [[[ 03:40:27 <sbp> Also, we host the <<swhack/> #swhack Website>, complete with 03:40:28 <sbp> <<swhack/chatlogs/> chat logs> and <<swhack/weblog/> weblog>. 03:40:33 <sbp> ]]] - http://blogspace.com/ 03:40:49 <sbp> the <<URI> title> syntax lets you know that there's a link there 03:41:15 <sbp> I'm wondering if a) I should change the style b) I should put the base URI onto any relative links 03:41:37 <deltab> not while displaying them like that 03:41:50 <sbp> no... I suppose it would be a bit unweildy 03:50:51 <AaronSw> * AaronSw is watching a fascinating show about deaf culture on PBS. 03:51:12 <sbp> Hmm... I need to comply to "1.9 When a Web resource includes metadata that may be recognized by the user agent, allow the user to view that metadata." in CUAP 03:51:28 <sbp> Ooh, I could have a -meta flag 03:51:56 <sbp> it could get the usual <link> and <meta> stuff, but also follow a link to some RDF :-) 03:52:24 <sbp> Hmph... if the W3C's RDF Validator used GET, I could run it through that 03:54:03 <sbp> BUG: always misses the space before the first inline element in a block level element 03:54:45 <sbp> Hmph, why is it doing that? 03:56:04 <sbp> ah, of course. Duh 03:56:52 <sbp> I used string.strip instead of string.lstrip 03:57:05 <atariBed> atariBed is now known as atariboy 03:58:34 <AaronSw> Wow, thay was quite a show. 04:02:16 <AaronSw> @ http://www.pbs.org/wnet/soundandfury/ 04:02:29 <chumpster> A: Sound and Fury - Home from AaronSw 04:02:50 <AaronSw> A:|Sound and Fury 04:02:51 <chumpster> titled item A 04:03:14 <AaronSw> A::An absolutely wonderful documentary. Fascinating and informative, it explained a world to me that I never knew existed. 04:03:14 <chumpster> commented item A 04:03:25 <AaronSw> A::I want to learn sign language some day. 04:03:26 <chumpster> commented item A 04:03:43 <AaronSw> AaronSw has changed the topic to: Canada: Our Eskimo Neighbors to the South 04:20:41 <deus_x> deus_x has quit (forward.openprojects.net irc.openprojects.net) 04:21:42 <deus_x> deus_x (~deusx@bgp993973bgs.nanarb01.mi.comcast.net) has joined #swhack 04:22:44 <sbp> Aaron (or anyone), do you know who said """If you're going to give out a press release in PDF, you may as well bury it in your back yard."""? 04:22:55 <AaronSw> Yeah. 04:23:02 <AaronSw> Geoff in an email to me. why? 04:23:03 <sbp> Can you tell me? 04:23:15 <sbp> Ah, right, thanks. I was just writing an email to Stu 04:23:36 <sbp> Er... Geoff... Geoff Freed? 04:23:56 <sbp> [Stu sent me a PDF file, so I thought I'd rant a bit about PDF] 04:24:44 <AaronSw> No, Geoff from TidBITS. 04:24:54 <AaronSw> Geoff Duncan 04:25:13 <sbp> Cool. Many thanks 04:25:42 <AaronSw> lol: Geoff: "An Apple operating system like a Britney Spears album: slick, but wrong on *so* many levels. (Note to Apple: I want Unix on my desktop the same way Wayne Newton wants me in his backup band. Think about it.) " 04:25:55 <AaronSw> - http://www.quibble.com/geoff/ 04:26:13 <sbp> heh, heh, heh 04:28:56 <AaronSw> lol: "[I've been] all but one of the following: dishwasher, lab assistant, technical writer, software test lead, floor washer, story editor, musical arranger, statistical consultant, illustrator, book-vacuumer, and guest lecturer. Guess which one I wasn't." 04:28:58 <AaronSw> - ibid. 04:29:27 <sbp> ah, it can now get text/plain too:- 04:29:28 <sbp> [[[ 04:29:29 <sbp> $ python browser.py http://blogspace.com/swhack/chatlogs/2002-01-09.txt 04:29:29 <sbp> Getting http://blogspace.com/swhack/chatlogs/2002-01-09.txt... 04:29:29 <sbp> Got it: text/plain; charset=iso-8859-1 04:29:29 <sbp> 00:00:10 <sbp> sorry, 2.59KB 04:29:30 <sbp> 00:00:28 <AaronSw> Neat. 04:29:32 <sbp> [...] 04:29:34 <sbp> ]]] 04:30:12 <sbp> Guess which one he wasn't? Eek 04:33:38 <sbp> I'll go with guest lecturer, and overlook the fact that there are only three book-vacuumers in the world 04:33:45 <AaronSw> But wait! It's more than just a lamp: http://sakots.pekori.jp/imgboard/imgs/img20020107162803.jpg 04:34:43 <sbp> heh! 04:35:08 <deltab> haha 04:35:15 <AaronSw> deltab, wmf is getting: <wmf> Host unknown (Name server: espians.com: no data known) 04:36:13 <deltab> when doing what? 04:36:20 <AaronSw> when sending mail to me 04:38:23 <deltab> well, I don't know why, it's there 04:38:29 <AaronSw> yeah 04:38:41 <sbp> Problem on wmf's end, perhaps 04:39:16 <sbp> Ugh, talk about bloat... browser.py is now 4.33KB! 04:39:53 <tav> almost doubled! 04:40:41 <sbp> Oh dear, you 04:40:48 <sbp> you're right. That's terrible 04:42:22 <sbp> Hmm... I wonder how we get the response code through urllib2 04:42:56 <sbp> Pff, it would also be nice if it didn't barf when it came across a 401 04:44:18 <AaronSw> you get the response code from errors, i think 04:44:23 <sbp> Argh, trailing whitespace in pre... 04:44:32 <sbp> But 200 doesn't give an error 04:45:19 <AaronSw> well that's why it's 200 04:45:46 <sbp> Phew, that's better... 04:46:04 <sbp> Yeah, but I want the code anyway. 20x shouldn't give an error 04:46:11 <sbp> neither should 30x 04:47:21 <AaronSw> it's like object.status or something, isn't it? 04:48:17 <sbp> dunno - I'll look it up properly when I get back to it (I'm back on the HTML parser again) 04:49:17 <AaronSw> Have you seen my HTML->Text deeley? 04:49:38 <AaronSw> It's in TCL, but I like it a lot. 04:49:40 <sbp> textify.tcl? Yeah. Not seen the source, though 04:49:52 <AaronSw> Yeah, http://developer.arsdigita.com/api-doc/proc-view?proc=ad_html_to_text&source_p=1 04:49:55 <sbp> But it's in TCL anyway, not Python... 04:50:05 <sbp> It is cool though, I'll give you that 04:50:22 <AaronSw> Ooh, I'm on the WGBH website! http://www.wgbh.org/api-doc/proc-view?proc=ad_html_to_text 04:50:23 <AaronSw> :-) 04:51:27 <sbp> Hmm... I prefer my link syntax 04:51:51 <sbp> A matter of preference, I guess. I should give a choice 04:52:29 <deus_x> deus_x has quit (forward.openprojects.net irc.openprojects.net) 04:52:49 <deus_x> deus_x (~deusx@bgp993973bgs.nanarb01.mi.comcast.net) has joined #swhack 04:54:12 <sbp> Hmm... I'm not quite sure how to do block vs. flow 04:54:50 <sbp> Roughly, I have a set of elements that print '/n/n' after themselves (block), a set that print '\n' (flow)m and a set that print nothing (inline) 04:55:35 <sbp> But that means that lists are <ul>(block)<li/>(flow)</> and so they get \n\n\n printed after them, which is a line break too many 04:58:01 <sbp> oh, heh, unless I make ul flow 04:58:21 <sbp> Ugh. See how terrible my HTML parser is? 04:59:12 <sbp> Ooh, that didn't work 04:59:27 <AaronSw> yeah, i like the way we did it in ad_html_to_text 04:59:33 <AaronSw> it solves those problems nicely 04:59:56 <sbp> I won't ask 05:00:16 <AaronSw> I don't remember how we solved them, so you'll have to figure it out yourself. 05:00:25 <AaronSw> I think we had a linebreak buffer... 05:00:39 <AaronSw> so if it had just spit out a line break, it wouldn't do it again. 05:00:57 <sbp> Ugh. That's the kind of hacking that I want to avoid 05:01:30 <tansaku> tansaku (~sam@h134-063.tokyu-net.catv.ne.jp) has joined #swhack 05:02:38 <sbp> Hmm... doesn't make sense: when I have ul as block, it appends two new lines. As flow, none 05:04:37 <AaronSw> hacking? I thought it was reasonably elegant. The element would set linebreak = 1, and then another piece of code would take care of it. 05:04:50 <AaronSw> linebreaks wouldn't pile up since they'd always be 0 or 1 05:05:20 <sbp> I do something a bit like that 05:06:25 <deltab> that's how margin collapsing is done in CSS 05:07:16 <deltab> the total space is the larger of the two 05:07:35 <deltab> or something like that 05:10:32 <AaronSw> Interesting: http://ali.apple.com/events/aliqttv/ 05:10:55 <sbp> aha, I think I've worked it out 05:12:02 <AaronSw> I should probably go to sleep now. nite 05:13:50 <sbp> Fixed 05:13:52 <sbp> 'night 05:14:26 <AaronSw> Goals for tomorrow: grok Chord, find best way to exercise. 05:14:31 <sbp> argh, no, not fixed... 05:15:32 <AaronSw> sleep 05:16:33 <sbp> ah, I think that the whitespace in the end of the <ul> is being pciked up 05:28:54 <sbp> sbp has quit (Killed (NickServ (Ghost: SeanP!~sean@m674-mp1-cvx3b.pop.ntl.com))) 05:29:13 <sbp> sbp (~sean@m674-mp1-cvx3b.pop.ntl.com) has joined #swhack 05:38:02 <sbp> There, fixed with a lot of hacking 05:38:06 <sbp> Gotta run 05:58:24 <sbp> sbp has quit (Read error: 110 (Connection timed out)) 06:25:56 <GabeW> GabeW (~gwachob@12-236-92-153.client.attbi.com) has joined #swhack 06:27:17 <GabeW> GabeW has left #swhack 06:28:34 <GabeW> GabeW (~gwachob@12-236-92-153.client.attbi.com) has joined #swhack 07:00:50 <GabeW> .seen AaronSw 07:00:50 <xena> AaronSw seen in #swhack saying: [ sleep ] ~ 1 hr(s) 45 min(s) 17 sec(s) ago 07:01:11 <GabeW> .seen #rdfi 07:01:11 <xena> GabeW: no match found: #rdfi 07:01:13 <GabeW> .seen #rdfig 07:01:14 <xena> GabeW: no match found: #rdfig 07:25:04 <GabeW> GabeW has quit ("Client Exiting") 08:22:58 <tansaku> tansaku has quit (forward.openprojects.net irc.openprojects.net) 08:22:58 <BenSw|away> BenSw|away has quit (forward.openprojects.net irc.openprojects.net) 08:45:46 <tansaku> tansaku (~sam@h134-063.tokyu-net.catv.ne.jp) has joined #swhack 08:46:28 <BenSW> BenSW (~Snak@12-249-96-16.client.attbi.com) has joined #swhack 09:24:03 <tansaku> tansaku has quit (Read error: 60 (Operation timed out)) 09:24:03 <tansaku> tansaku (~sam@h134-211.tokyu-net.catv.ne.jp) has joined #swhack 10:41:14 <tansaku2> tansaku2 (~sam@h134-211.tokyu-net.catv.ne.jp) has joined #swhack 10:41:18 <tansaku> tansaku has quit (Read error: 110 (Connection timed out)) 10:41:42 <tansaku2> tansaku2 is now known as tansaku 10:51:27 <atariboy> atariboy is now known as atariMovie 13:31:33 <BenSw`> BenSw` (nobody@un.impressive.net) has joined #swhack 13:31:51 <BenSw`> Hello 13:50:08 <AaronSw> hi 13:50:09 <BenSw`> BenSw` has quit (Remote closed the connection) 13:58:18 <atariMovie> atariMovie is now known as atariboy 14:06:23 <AaronSw> heh! http://www.scripting.com/images/2002/01/08/pixar.jpg 14:32:52 <em-mit> em-mit (~em@24-6-152.wireless.lcs.mit.edu) has joined #swhack 14:34:51 <AaronSw> hey em 14:43:09 <em-mit> howdy AaronSw :) how was the trip? 14:43:23 <AaronSw> it was great, thanks 14:43:56 <AaronSw> how was your vacation? 14:44:52 <em-mit> very short (albiet enjoyable)... www2002 consumed a chuck of it, but i did manage to do some much needed sysadmining, and coding 15:00:11 <deus_x> deus_x has quit (forward.openprojects.net irc.openprojects.net) 15:03:58 <deus_x> deus_x (~deusx@bgp993973bgs.nanarb01.mi.comcast.net) has joined #swhack 17:24:05 <sbp> sbp (~sean@m393-mp1-cvx3b.pop.ntl.com) has joined #swhack 17:30:24 <sbp> * sbp hacks on his browser a bit 17:44:49 <sbp> * sbp adds FTP capabilities 18:12:10 <sbp> cool, done 18:12:55 <sbp> Argh, browser size is a lofty 5.25KB 18:28:09 <AaronSw> @ http://cyberlaw.stanford.edu/lessig/content/audio/I_Thought_We_Knew_That.mp3 18:28:13 <chumpster> B: http://cyberlaw.stanford.edu/lessig/content/audio/I_Thought_We_Knew_That.mp3 from AaronSw 18:28:22 <AaronSw> B:|I Thought We Knew That.mp3 18:28:22 <chumpster> titled item B 18:29:00 <AaronSw> B::Larry Lessig MP3s! Rocking... 18:29:01 <chumpster> commented item B 18:29:15 <AaronSw> B::Check out some [more Lessig audio/video|http://cyberlaw.stanford.edu/lessig/content/audio/] 18:29:16 <chumpster> commented item B 18:30:01 <AaronSw> Wow, this browser is going to be quite something when you're done, isn't it? 18:30:27 <AaronSw> B::via Zooko in #nowplaying 18:30:27 <chumpster> commented item B 18:30:34 <AaronSw> join #nowplaying and tell us what you're listening to 18:33:38 <AaronSw> B::lol: Gates: "We did not attempt to monopolize the browser market." Lessig: "Only a moron would think of saying that." 18:33:39 <chumpster> commented item B 18:35:32 <sbp> quite something: heh, heh 18:38:54 <sbp> I should learn TKInter, really, and go a GUI browser 18:39:00 <sbp> s/go/do/ 18:39:47 <sbp> But can you imagine programming a GUI? Aaaaaargh! 18:40:19 <sbp> The problem with that is that you have to cater for every little mouse click, tabbing, window geometry, and so on 18:42:58 <sbp> Well, this can do FTP, -head, -text, HTML rendering, and pipe in a URI, so that'll probably do 18:43:07 <sbp> it's not as if I'm actually going to use it for anything 18:43:28 <sbp> although I might use it for HEAD: I can never remember the cmd line for lynx 18:43:56 <sbp> and it 18:44:14 <sbp> it's good for getting a page that you can paste into IRC or email: it formats it, putting the links inline 18:57:52 <AaronSw> I use curl for HEAD. 18:57:58 <AaronSw> curl -I http://url.foo 18:58:10 <AaronSw> and textify.tcl for pasting into an email. 18:58:25 <AaronSw> It'd be cool to see you do a GUI -- I saw a lot of RDF graph visualization GUIs when I was at ILRT. 19:14:28 <AaronSw> .seen anubis3 19:14:40 <AaronSw> * AaronSw pokes xena 19:14:45 <AaronSw> xena is losing her edge 19:14:46 <xena> anubis3 seen joining #infoanarchy ~ 30 day(s) 15 hr(s) 26 min(s) 45 sec(s) ago 19:15:00 <AaronSw> 30 days... eek 19:17:17 <sbp> sbp has quit (Read error: 110 (Connection timed out)) 19:49:41 <atariboy> atariboy has quit ("*gone*") 20:05:53 <sbp> sbp (~sean@m319-mp1-cvx3b.pop.ntl.com) has joined #swhack 20:06:15 <AaronSw> sbp, did you hear about #nowplaying? 20:06:26 <sbp> no? 20:06:44 <AaronSw> it's where we share what we're listening to 20:07:31 <sbp> Heh, neat 20:08:22 <AaronSw> Perfect for you. ;-) 20:08:57 <sbp> damn straight 20:14:43 <AaronSw> * AaronSw wonders how the icon for his copy of Acrobat Reader got to be a folder. 20:15:34 <AaronSw> a 20:19:24 <AaronSw> test 20:30:00 <AaronSw> ouch, looks like Bernstein got a patent letter 20:30:01 <AaronSw> http://markBernstein.org/Jan0201.html#note_3404 20:31:17 <sbp> heh: """[...] the patent is written in esoterically amateur language that seems tantamount to gibberish. I doubt if anyone knows what this patent is trying to say.""" 20:32:46 <sbp> Ooh, it uses httplib now, but is still only 5.43KB 20:34:04 <sbp> argh, but it doesn't handle redirects 20:36:35 <sbp> Which is good actually - more flexibility 20:45:46 <sbp> Hmm... since I started using httplib, it's slowed down again. urllib was slow too. Only urllib2 opened HTTP URIs with any speed 20:46:56 <sbp> THe FTP part kicks ass - it streams the file out *as* its downloading it! 20:47:59 <AaronSw> Heh, heh, heh: 20:48:01 <AaronSw> .google sha512 20:48:03 <xena> sha512: http://blogspace.com/swhack/weblog 20:48:07 <AaronSw> :-) 20:48:10 <sbp> hooray! 20:48:30 <AaronSw> but I clearly need to link to it more, since we've not been indexed in a while 20:49:05 <sbp> YEah 20:49:16 <sbp> ARgh, DAmn IT 20:49:39 <AaronSw> he 20:49:42 <AaronSw> HEh 20:50:02 <sbp> MUST... RELEASE... SHIFT... KEy... 20:50:21 <AaronSw> odd... it doesn't recognize a backlink from aaronsw.com 20:51:42 <sbp> weird 20:53:20 <AaronSw> yeah, it is. especially since i was counting on those links to give us major google power 20:57:43 <sbp> heh, heh 20:58:00 <sbp> ask jillium, perhaps 20:59:54 <AaronSw> i clicked the "report a link you wanted to see here" thing and filled it out 21:00:09 <AaronSw> how many bits are in a sha512 hash? 21:00:51 <sbp> Hmm... I wonder 21:00:59 <AaronSw> i would have guessed 512, but my calculations return 2048 21:01:23 <sbp> Your calculations therefore suck 21:01:42 <AaronSw> my calculation: len(sha512.hexdigest())*16 21:01:46 <AaronSw> what's wrong with that? 21:02:04 <AaronSw> even if i'm off by two that leaves me with 1024 21:02:19 <sbp> try multiplying by 4 21:02:26 <sbp> instead of 16, that is 21:02:30 <AaronSw> why 4? 21:02:35 <sbp> just try it 21:02:41 <sbp> what answer do you get? 21:02:43 <AaronSw> it obviously works 21:02:51 <AaronSw> 128*4=512 21:02:59 <sbp> There you go, then. Now let's say no more about it 21:03:11 <AaronSw> That'd be cheating. 21:03:19 <sbp> Nah 21:04:05 <AaronSw> let's see... if it was in binary i'd multiply by 1 21:04:40 <sbp> can you do print sha512.hexdigest() 21:04:46 <AaronSw> 'f7fbba6e0636f890e56fbbf3283e524c6fa3204ae298382d624741d0dc6638326e282c41be5e4254d8820772c5518a2c5a8c0c7f7eda19594a7eb539453e1ed7' 21:07:36 <AaronSw> it appears your *4 metric is correct (it works on sha1) but i don't know why 21:08:45 <sbp> * sbp neither 21:11:20 <sbp> The lack of SHA 512 in Python is quite annoying 21:11:36 <sbp> s/quite/very/ 21:12:31 <AaronSw> Heh. 21:12:54 <sbp> and sha512.exe won't take a pipe 21:13:02 <sbp> ooh, I could write a Python wrapper for it 21:13:12 <sbp> what methods does sha512 have? 21:13:20 <AaronSw> same as usual: 21:13:25 <AaronSw> ['__del__', '__doc__', '__init__', '__module__', 'blocksize', 'copy', 'digest', 'digestsize', 'hexdigest', 'shaContext', 'update'] 21:13:25 <AaronSw> >> 21:13:38 <sbp> Mmmkay 21:16:26 <AaronSw> oh, i figured it out 21:16:40 <AaronSw> you need to multiply by the number of bits per character, not the number of characters per character. 21:16:51 <AaronSw> i.e. the square root of the latter. 21:16:56 <sbp> ah 21:24:03 <AaronSw> what's the keyword for assert in python? 21:25:05 <AaronSw> hmm, it is assert 21:25:15 <sbp> heh, heh 21:25:32 <sbp> Python: making everything too easy 21:25:57 <sbp> does sha512 have 'new'? 21:26:07 <sbp> x = sha512.new() 21:40:46 <sbp> getting there:- 21:40:47 <sbp> [[[ 21:40:47 <sbp> $ python 21:40:47 <sbp> Python 2.2 (#1, Dec 31 2001, 15:21:18) 21:40:47 <sbp> [GCC 2.95.3-5 (cygwin special)] on cygwin 21:40:47 <sbp> Type "help", "copyright", "credits" or "license" for more information. 21:40:49 <sbp> >>> import sha512 21:40:51 <sbp> >>> x = sha512.new() 21:40:53 <sbp> >>> x.update('x') 21:40:55 <sbp> >>> x.update('blargh') 21:40:57 <sbp> >>> x.hexdigest() 21:40:59 <sbp> 'SHA512 : sometempfile\r\n\r\n 52e9282e113609adb255c4d797d9c450642f94cf580e0353a4cb9481d53fff58\r\n 524ce1433ed62743e14ae421da539e94fc416c4c49d53c1bf27afa0db6dda748\r\n' 21:41:02 <sbp> >>> 21:41:06 <sbp> ]]] 21:42:31 <sbp> aha:- 21:42:32 <sbp> [[[ 21:42:32 <sbp> >>> x.hexdigest() 21:42:32 <sbp> '4a59e80b6a5bfb4c8d8a592086a290b347c5d2b62418abe878cfa8037063ca268edeadd184bc3d54bf683584f294caf7a3109c90458b45f93d25587618070fa1' 21:42:33 <sbp> ]]] 21:43:09 <AaronSw> cool 21:43:20 <AaronSw> can you just do x.new('foo').hexdigest() 21:43:21 <AaronSw> ? 21:43:27 <sbp> nope, not yet 21:43:56 <AaronSw> * AaronSw is trying to translate the Chord paper into Python 21:43:59 <AaronSw> it's slow going 21:44:39 <AaronSw> I understand how Chord works, but I don't understand the paper so well. 21:44:48 <AaronSw> Maybe i'm missing some math... 21:44:49 <sbp> [[[ 21:44:49 <sbp> >>> sha512.new('blargh').hexdigest() 21:44:50 <sbp> '4a59e80b6a5bfb4c8d8a592086a290b347c5d2b62418abe878cfa8037063ca268edeadd184bc3d54bf683584f294caf7a3109c90458b45f93d25587618070fa1' 21:44:50 <sbp> ]]] 21:45:58 <AaronSw> what's sha512.new().__class__? 21:46:35 <Morbus> Morbus (~Morbus@morbus.totalnetnh.net) has joined #swhack 21:46:42 <AaronSw> hey Morbus 21:46:43 <Morbus> AaronSw, you said Python was NOT installed by default in OS X, right? 21:47:12 <AaronSw> I don't recall saying that. 21:47:22 <Morbus> ok. is python installed by default on OS X? 21:47:39 <AaronSw> i believe so... either that or with the devtools. 21:47:44 <AaronSw> is perl installed by default? 21:48:02 <Morbus> yeah, but you can't use CPAN without the devtools. 21:48:14 <sbp> [[[ 21:48:15 <sbp> >>> sha512.new().__class__ 21:48:15 <sbp> <class sha512.SHA512 at 0x10145158> 21:48:16 <sbp> ]]] 21:48:19 <Morbus> that's what i've been told though. i dunno for sure. 21:48:31 <sbp> Morbus?! 21:48:43 <sbp> Heh, I thought I was in the wrong channel for a second, there 21:48:50 <AaronSw> Hmm... maybe it isn't installed by default. 21:48:53 <Morbus> don't worry, i'm not staying long. just pumping aaron for info, then i'm outta here faster than a d00d at a 0day. 21:49:15 <Morbus> yeah, i was under that impression too. i can't check though, of course. 21:49:15 <sbp> Mmmkay 21:49:54 <AaronSw> Python 2.0 got in /usr/bin somehow 21:50:18 <AaronSw> let me check other machine ... brb 21:51:36 <BenSW> * AaronSw twiddles thumbs 21:51:44 <Morbus> heh, heh. 21:52:05 <BenSW> <AaronSw> Hmm, terminal isn't opening 21:53:56 <sbp> heh, heh 21:54:11 <sbp> Gotta run 21:54:32 <BenSW> hmm, it's not on here 21:54:51 <BenSW> <AaronSw> and this has the developer tools 21:55:12 <Morbus> ok. that's what i needed to know. thank you kindly. 21:55:23 <AaronSw> np 21:55:27 <Morbus> Morbus has left #swhack 21:55:43 <AaronSw> Ooh: http://www.apple.com/downloads/macosx/unix_apps_utilities/python21.html 21:57:01 <AaronSw> i'm taking bets on how long it takes tav to join #nowplaying 22:08:45 <AaronSw> Hmm, Wolfgang Nejdl and Stefan Decker are on the Edutella project. 22:10:22 <AaronSw> Edutella components include: 22:10:29 <AaronSw> * Query Service: Standardized query and retrieval of RDF metadata. 22:10:29 <AaronSw> * Replication Service: Provide data persistence / availability and workload balancing while maintaining data integrity and consistency. 22:10:29 <AaronSw> * Mapping Service: Translate between different metadata vocabularies to enable interoperability between different peers. 22:10:29 <AaronSw> * Annotation Service: Annotate materials stored anywhere in the Edutella Network. 22:11:53 <AaronSw> heh: http://www.kbs.uni-hannover.de/Diverses/edutella-archive/discussion/msg00147.html 22:14:19 <sbp> sbp has quit (Read error: 110 (Connection timed out)) 22:47:20 <AaronSw> Hmm, edutella is inventing their own query language, rdf conversion mechanism, etc. 22:48:46 <AaronSw> I think that's the wrong place for something like that. 22:51:06 <AaronSw> I like how Bit Torrent went from version one-point-oh to one-point-oh shit, in the words of Bram. ;-) 23:02:14 <sbp> sbp (~sean@m336-mp1-cvx3b.pop.ntl.com) has joined #swhack 23:11:02 <sbp> ta da: http://lists.w3.org/Archives/Public/www-archive/2002Jan/att-0025/01-sha512.py 23:16:38 <sbp> * sbp lengthens an old song that he wrote 23:16:47 <sbp> let's time it 23:16:48 <sbp> .time 23:16:50 <xena> 2002/01/09 23:18:27.6259 Universal 23:19:07 <sbp> .time 23:19:08 <xena> 2002/01/09 23:20:46.2678 Universal 23:19:18 <sbp> Hmm... pretty good. Just about long enough 23:25:06 <sbp> Perhaps I can make an arpeggio of it, although that might make it a bit weak 23:25:21 <sbp> I did used to do that to AXV when I'd play it to people, but gave in after a while 23:26:31 <sbp> Nope, sounds better strummed 23:32:10 <sbp> * sbp updates songs.html 23:34:37 <sbp> Pff, even Stu commented the other day that I can't write songs like AXV anymore 23:34:50 <sbp> He's right, and it's annoying 23:36:54 <hazmat> hazmat (~ender@adsl-66-123-57-58.dsl.lsan03.pacbell.net) has joined #swhack 23:37:24 <AaronSw> songs.html? 23:37:26 <AaronSw> AXV? 23:37:42 <sbp> uh huh 23:38:09 <AaronSw> full uri for songs.html? 23:38:20 <sbp> logster, grep <sbp> .*A.*x.*v 23:39:05 <logster> I'm logging. I found 30 answers for '<sbp> .*A.*x.*v' (showing 0...4) 23:39:06 <logster> 0) 2002-01-09 23:38:20 <sbp> logster, grep <sbp> .*A.*x.*v 23:39:07 <logster> 1) 2002-01-08 04:16:17 <sbp> Aaron, can you be canonical about links to http://www.plexdev.org/ ? 23:39:08 <logster> 2) 2001-12-31 23:33:21 <sbp> Well, it's counter intuitive. Archiving everything has only recently become possible - even feasable on the level that the Plex is being designed for. And there are the complexities of searching billions of triples for really simple search patterns. If it wasn't for things like Google, you wouldn't have thought that such searches would be possible 23:39:09 <logster> 3) 2001-12-31 23:25:26 <sbp> Ah. But wasn't the problem that xena wasn't relaying the text over? 23:39:11 <logster> 4) 2001-12-30 01:26:01 <sbp> When you think about it, if Alexa and Google can store all that content, there's not much need for people paying out for server space. Memory is so cheap these days, so the situation is getting better 23:39:14 <sbp> full URI is something like file://c:/[...]/songs.html 23:39:22 <sbp> well, that was good 23:39:27 <AaronSw> You know that never works. 23:39:30 <sbp> Heh, heh 23:39:34 <AaronSw> you need to do [^ ] not . 23:39:52 <sbp> AXV is the filename abbreviation that I use for Alexandr'a Views 23:40:32 <AaronSw> aha 23:43:40 <sbp> The mixed reacations that I've got upon playing it to people are pretty fun 23:43:57 <AaronSw> You didn't play it for me, did you? 23:44:05 <sbp> nope 23:44:16 <AaronSw> i've somee comments on http://lists.w3.org/Archives/Public/www-archive/2002Jan/att-0025/01-sha512.py 23:44:24 <AaronSw> like what's up with """Copyright (C) 2001 Sean B. Palmer. GNU GPL 2"? 23:44:32 <sbp> As far as I recall, I played The Rain Song 23:44:37 <AaronSw> yeah, that was cool 23:44:39 <sbp> What about it? 23:44:46 <AaronSw> having a bout of quote indecison? 23:44:54 <AaronSw> triple quotes, double quotes? 23:44:54 <sbp> oh, heh! 23:45:00 <sbp> forgot to strip that off 23:45:03 <sbp> oh well, who cares 23:45:21 <AaronSw> and you should use the tempfile module 23:45:23 <hazmat> regarding plex, how is it going to efficiently search through the mass of triples? 23:45:24 <sbp> me playing the Rain Song cool: I hit two wrong notes. Two! Shoews how nervous I was 23:45:41 <sbp> tempfile - I like my little times based solution. I use that a lot now 23:45:43 <AaronSw> Heh, heh. I wish I could play that well. 23:45:47 <hazmat> i haven't seen any indexing technologies in python up to the task. 23:46:01 <AaronSw> hazmat, on a network basis or a local one? 23:46:23 <hazmat> local 23:46:30 <AaronSw> for the local one, we're investigating building our own lightweight rdf database and storage system, possibly based on irondoc 23:47:00 <hazmat> .google irondoc 23:47:02 <xena> irondoc: http://www.irondoc.net/home.html 23:47:16 <AaronSw> .google mccusker irondoc 23:47:17 <xena> mccusker irondoc: http://www.best.com/~mccusker/irondoc/irondoc.htm 23:47:48 <AaronSw> Initially, however, we're just going to use anydbm. 23:47:59 <AaronSw> (which is how everyone else I know does it) 23:48:12 <hazmat> which is not scalable. 23:48:40 <hazmat> berkelydb would be a better choice 23:48:45 <tansaku> tansaku has quit (Read error: 110 (Connection timed out)) 23:48:52 <AaronSw> by anydbm i meant berkeleydb (ideally) 23:49:51 <hazmat> the functionality offered by the raw python bindings to bdb is much more advanced than a simple anydbm interface. it includes additional functionality like transactions and customization of the indexing structure (hash, btree, heap, etc.). 23:50:11 <sbp> sbp has quit (Killed (NickServ (Ghost: SeanP!~sean@m56-mp1-cvx4c.pop.ntl.com))) 23:50:30 <sbp> sbp (~sean@m56-mp1-cvx4c.pop.ntl.com) has joined #swhack 23:50:46 <AaronSw> I know, but I looked thru it and didn't see anything I needed. 23:50:52 <hazmat> ok. 23:50:54 <AaronSw> Is there something I missed? 23:51:20 <hazmat> well if you think its adequate, than no. 23:51:36 <AaronSw> obviously, this is not a long-term solution 23:51:39 <hazmat> i would just be concerned about managing potentially gigabytes of info via that interface. 23:51:54 <hazmat> esp. without transactions to maintain integrity. 23:51:58 <AaronSw> yeah, me too, which is why i'm working my way thru "Managing Gigabytes". 23:52:03 <AaronSw> :-) 23:52:08 <sbp> heh, heh 23:52:16 <sbp> .google "Managing Gigabytes" 23:52:17 <xena> "Managing Gigabytes": http://www.cs.mu.oz.au/mg 23:52:19 <hazmat> and the customization of the indexing storages to make searches timely 23:52:24 <sbp> .google "Managing Gigabytes" now with a forward by Aaron Swartz 23:52:25 <xena> no results found. 23:52:28 <sbp> blargh 23:52:29 <AaronSw> heh heh 23:52:34 <hazmat> check out some of the papers from doug cutting. 23:52:36 <hazmat> and lucene. 23:52:41 <hazmat> .google lucene 23:52:42 <xena> lucene: http://jakarta.apache.org/lucene 23:52:47 <AaronSw> Lucene looks cool... I hadn't thought about it in this context, tho. 23:53:02 <AaronSw> I keep meaning to add it to Blogspace. 23:54:09 <AaronSw> Do you think Lucene fields would be good for storing/querying RDF? 23:54:14 <hazmat> the core functionality are the same, rapidly being able to search through vast amounts of information. 23:54:38 <hazmat> AaronSw: it would require some customization to create the index to do rdf associations, i would think. 23:54:46 <AaronSw> I guess the ideal thing might be PostgreSQL. 23:54:56 <AaronSw> or maybe even (gasp!) MySQL. 23:55:47 <hazmat> an rdbms does really sound tempting, since its basically relationship matching. but is it appropiate? 23:55:54 <sbp> Just remember, the simpler (less dependencies) the better. Then again, as long as it can be packed, I suppose it doesn't matter 23:56:01 <sbp> s/packed/packaged/ 23:56:22 <AaronSw> Exactly. That's why I've shied away from full SQL DBs. 23:56:43 <AaronSw> A lot of the more advanced RDF QLs have been built atop full RDBMS 23:56:53 <sbp> Yeah 23:57:05 <sbp> But by "advanced", do you mean "quick"? 23:57:24 <sbp> Has anyone does a decent benchmark of the best RDF APIs? 23:57:29 <hazmat> i think alot of that derives from the lack of good indexing systems that are flexibile and scalable enough . 23:57:30 <AaronSw> No, but it's probably faster than a simple lightweight db 23:57:37 <sbp> er... RDFQ APIs, that is 23:57:54 <AaronSw> Well, RDFQs are really young! 23:58:01 <AaronSw> I'm interested in how RDQL turns out. 23:58:13 <AaronSw> And of course we'll support RDAP. 23:58:34 <hazmat> i'm still one the beg. of the rdf learning curve. a novice still. sigh.. so much to learn. 23:58:39 <hazmat> s/one/on 23:59:02 <sbp> Nah, RDF's just a big feck of a Webized database format 23:59:06 <AaronSw> You seem to have picked up a surprising amount! 23:59:13 <AaronSw> Yeah, sbp's got it. 23:59:47 <hazmat> is irondoc usable? 23:59:55 <sbp> .google irondoc 23:59:56 <xena> irondoc: http://www.irondoc.net/home.html