00:18:31 ndw has joined #swig 00:30:25 libby has quit 00:34:59 zwnj has joined #swig 00:35:12 lnxnt_ has joined #swig 00:42:06 lnxnt has quit 00:42:07 hi there 00:42:25 i'm confused with OWL and RSS 00:42:41 recently i read a bunch of OWL introductions 00:43:19 but i don't know how it's possible to put some data, like a web sites feed in OWL 00:43:31 i started work with Protege 00:43:39 and there's no problem there 00:43:59 but, what about the content of instances? 00:44:29 for example, how i can have a owl class, named Content 00:44:57 which can have Content children nodes 00:47:29 maybe the question is, is OWL completely disjoint from documents grammer (xml schema)? 00:55:16 kpreid_ has joined #swig 00:57:39 idealm has joined #swig 01:00:02 kpreid has quit 01:11:30 eikeon has quit 01:16:15 eikeon has joined #swig 01:19:09 kpreid_ has quit 01:22:50 kpreid has joined #swig 01:28:49 zwnj has quit 01:34:06 nemequ_ has joined #swig 01:35:28 zwnj has joined #swig 01:51:43 nemequ has quit 01:58:56 zwnj, OWL models not the documents structure but the relationships between the things the document i sabout. 02:04:24 timbl: thanks. so should i define the grammer is xml schema, and the logic relationship in owl? is there any good example around? 02:15:55 When you have figured out what your data is about, you can define those things in OWL - or RDFS - or not at all. But the RDF syntax defines the syntax for data. 02:16:04 You don't need any XML schema. 02:16:34 You can't use XML schema in fact. 02:21:21 timbl: hum, i think i got it now 02:21:45 timbl: btw, do you know any ontology for literature, poems, etc? 02:28:43 Libraries must have them... starting with Dublin Core 02:35:37 EliasT has joined #swig 02:36:26 EliasT has left #swig 02:42:50 ndw has quit 02:47:29 idealm has quit 02:49:47 idealm has joined #swig 03:03:02 zwnj has quit 03:10:05 danbri_ has joined #swig 03:21:01 danbri has quit 03:40:34 danbri_ has quit 04:16:03 nemequ_ has quit 04:16:14 bengtf__ has joined #swig 04:16:22 bengtf__ is now known as bengtf 04:46:58 zwnj has joined #swig 05:02:52 bengtf has quit 05:07:21 kpreid has quit 05:12:27 kpreid has joined #swig 05:17:41 zwnj has quit 05:44:42 idealm has quit 05:47:03 idealm has joined #swig 06:00:09 csarven has quit 06:56:28 monkeyiq has joined #swig 07:03:29 idealm_ has joined #swig 07:14:57 danja has joined #swig 07:22:21 idealm has quit 07:28:18 jargonjustin has joined #swig 07:33:04 http://www.intertwingly.net/blog/2006/10/28/XHTML-as-practiced 07:33:06 A: http://www.intertwingly.net/blog/2006/10/28/XHTML-as-practiced from danja 07:33:28 A:| XHTML as practiced 07:33:29 Titled item A. 07:35:17 A: Sam Ruby spotted a markup [error|http://validator.w3.org/check?uri=https%3A%2F%2Fround-lake.dustinice.workers.dev%3A443%2Fhttp%2Fdig.csail.mit.edu%2Fbreadcrumbs%2Fnode%2F153] on timbl's blog 07:35:18 Added comment A1. 07:35:20 every blogging engine should automagically show all referring links.. 07:35:26 ham is now known as c 07:36:05 * c thinks...it would be cool if ther was a magical 'escape' char for http URIs so that they wouldnt be so uggly when containing other URIs 07:36:48 c, i proposed universal metacharacters, one for commenting, one for escaping.... 07:36:56 interesting 07:36:56 to the w3c tag list 07:37:28 A: some good comments, e.g. Phil Ringnalda : "Seems to me like a fine example of how, if you want to render error-filled things, you have to specify error recovery, in all its trillions of possible permutations." 07:37:29 Added comment A2. 07:38:18 error recovery es evil 07:38:24 be strict to be cool 07:38:41 I am growing increasingly of the belief that one of the ramifications of driving towards “Radical Simplification” is that one must plan for, and accommodate, those that use simple text editors and templates. 07:39:18 imho, very much the ooposite. if we could code WYSIWYG *with* perfect code generartion, then all code would be valid, automatically 07:39:42 but who tries? amaya? nvu? 07:41:06 I prefer the idea of the CMS doing pre-publication validation/normalisation 07:41:34 whether the source is templated or text or wysiwyg or whatever 07:42:10 that would allow to take error correcting code outside browsers, which sounds good to me, danja 07:42:26 though a perfect wysiwyg would certainly help... 07:42:54 end users should never have to see code, exactly 07:44:50 hmm, but there's plenty of overlap between end users and developers (I thinking PHP) 07:45:17 some sort of scriptable perfect wysiwyg would be best, i assume 07:45:32 kind of mix between editor and cms 07:46:27 what's frustrating is that many of the wysiwygs are pretty close, but not perfect 07:46:35 exactly 07:46:55 but even more worrying, danja...is it even possible? 07:47:08 heh, good question 07:47:24 when was the last time someone checked that xhtml, as it stands today, is editable, perfectly, in a wysiwyg manner? 07:48:24 should be interesting to see where the TAG go with the soup issue 07:49:03 i think there are politics involved there, danja 07:49:13 i'm out of the loop 07:49:17 heh, quite possibly ;-) 07:49:20 but i very much suspect it 07:49:30 but some strong opinions too I imagine 07:50:07 well, danja...there seems to be no possible doubt whatsoever. stric is good, tagsoup is bad, i mean 07:50:30 and is so clear, that the only possibility i find for the tag to have even raised the issue...is politics 07:50:32 so the majority of the web is bad? 07:50:38 sure 07:50:46 we would be better off without it? 07:51:00 without tag soup and error correction? sure 07:51:39 so if overnight, the web was trimmed back to only the pages that passed validation, that would be a good thing? 07:52:17 danja...isn't it obvious? (i mean no disrespect, but, it stands cristal clñear to me, i mean) 07:53:15 the requisite for that, though, is wysiwyg edotirs, perfect 07:53:44 ok, here in Italy nearly every driver overtakes on corners 07:53:55 it's can be downright scary 07:54:36 one such perfect editor, would make perfect code, with zero (or almost) learning curve a given, automatically 07:54:55 to edit, would be way easier than by hand editing 07:55:16 and the tool *itself* would ensure perfect validity of code 07:55:34 but if you banned all those drivers, the road transport system would cease to exist 07:55:53 give the drivers robotic drivers 07:56:00 perfect robotic drivers 07:56:03 sure Biblio, but that depends on !. there being such a tool and 2. everyone using it 07:56:26 or perfect robotic drivers... 07:56:27 first, it depends on making sure it can even be done, at all, danja 07:56:33 of which i'm not positive 07:56:58 timbl's original idea, was wysiwyg 07:57:07 end users never saw code 07:57:15 that was the way to go 07:57:40 well, as far as I'm aware there aren't really hard and fast rules on how things should be rendered 07:58:11 which is one of the reasons why i'm asking timbl if w3c would do a browser 07:58:36 it would be nice to see amaya revived 07:58:52 but a bit better =P 07:58:58 and ready for semweb as well 07:58:58 yep 07:59:04 indeed 07:59:56 I'd better get on with some work (deadlines tomorrow), catch you later Biblio 08:00:06 later, a pleasure 08:01:35 monkeyiq has quit 08:02:16 xhtml is not being used cause vendors don't support it 08:03:56 "don't use xhtml, use html" <--*every* desing site/channel/forum says this, to the legion of new guys who would like to try it 08:07:07 A: See also http://chatlogs.planetrdf.com/swig/2006-10-29.html#T07-33-04 08:07:08 Added comment A3. 08:15:19 jargonjustin has quit 08:26:44 nemequ has joined #swig 08:42:26 gosh, i just see the timbl's blog post, i did not know it was there! 08:42:29 c has quit 08:42:52 i reaaly hopes he sees the former twenty lines here 08:43:42 xx has joined #swig 08:44:24 xhtml did not flop because too many small changes in syntax, those are even trivial 08:46:21 to go the strict way, is just about imposible if people handcodes, but perfectly doable if wysiwyg tools *with* perfect code generation exist 08:46:43 bblfish has joined #swig 08:47:03 the true problem is that nobody thought if html is ready for such a thing, for a long time 08:47:49 (daniel glazman, amaya people, probably the best for that) 08:48:32 changes *might* be needed, both in html, and in css, in order for it to be edited wysiwyg perfectly 08:49:02 example... 08:49:24 in a simple text file (.txt), i can copy and paste, perfectly 08:50:03 in html *if i request to keep styles as well while doing it*, i can *not* copy and paste 08:50:32 i do not know of even one public implementation able to do that, as of today 08:51:13 drrho has joined #swig 08:51:49 such a humoungously basic funtionality, and not possible 08:55:29 when css was being developed, probably nobody ever thought about how it would be to use it for wysiwyg editing 08:55:44 i'm no expert, but i doubt very much that that was the case 09:07:36 btw, the whole semweb is coming. do we want "semweb tagsoup" to appear as well? fault tolerance adding unnecessary complexity here as well? tremendous learning curves if people handcodes? 09:08:21 or do we want to go the wysiwyg way? the editing-tool-as-validator way? the visual, minimal complexity for users way? 09:08:32 garbage in, garbage out... 09:09:06 But nobody is arguing for the return of presentational markup. They're arguing for tags you don't have to close and no quotes around attributes. 09:09:38 they don't care about the stuff that's hard to implement in wysywig tools, they care about the stuff that's slightly less easy to write by hand. 09:09:59 but the problem is that nobody should edit by hand 09:10:50 )in which case quotes or not quotes, becomes just about irrelevant) 09:11:07 I agree. But that's what they're compaining about. 09:12:16 everyone hates presentational mark-up... but a lot of people still love tag soup. 09:13:06 do you know of even one wygiwys editor *with* perfect code generation today? coiuld they even choose? no such thing, is what i point 09:13:40 and handcoding means you can not go strict for the masses 09:14:20 (and my grandmother could never pull a valid page) 09:15:07 and before, that did not mean that much...but now, it means no interoperable, no machine processable content 09:15:15 so you think the w3c should ignore what they're asking for, and instead put together what the w3c things would be better? 09:16:11 I agree with you to an extent (they're asking for the wrong thing), but if the w3c delivered the right thing people would just complain about the w3c still not listening to them. 09:16:17 i think timbl's fierst idea was good. wysiwyg editor, users never saw code, output was as desired 09:16:53 i see this as more of a political move... just trying to appease people for now so eventually they'll follow over to a better web. 09:17:00 namequ...perhaps. unless it could be shown that a certain way of doing things, would provide solution for everyone 09:17:27 that would be great, but i don't think it's likely. 09:17:37 nemequ, yes, i see. and i undertand, and i agree, this seems politics. and agreed, i do not know all the details 09:18:09 but i also think that certain things were never thought in certains ways before (or at least for the last decade) 09:18:50 heh. yeah well with a decade of MS frontpage and friends, you can hardly blame people for not trusting wysiwyg ;) 09:19:01 exactly 09:19:50 i figure just ride out this crap, and keep working on the semweb. eventually, when we've got software to show instead of just (really good) ideas, they will come. 09:21:02 the semweb already runs the risk of going the tagsoup way, imho 09:21:24 i.e, microformats, for instance 09:23:42 yeah i'm not a fan of microformats... RDFa needs to hurry up. 09:24:09 rdfa, in any case, better, yes. or perhaps grddl. 09:24:35 (i question myself if embedding rdf into html is a good idea, at all, though) 09:26:45 perhaps, but until we get people to accept XHTML and namespaces... 09:27:25 I'm a fan of doing things right the first time, but all the people who are complaining about HTML right now want to do things gradually. 09:28:01 so it's either RDFa/GRDDL, microformats, or nothing. my vote is with the lesser of the three evils. 09:28:28 if necessary, seems sensible to me 09:28:39 JibberJim has joined #swig 09:29:19 but perhaps better to keep rdf appart, just linked, instead ef embedded 09:29:33 i'd like a good discussion on that, i mean 09:34:16 I dunno... data duplication means more opportunities for contradictory information, more difficulty to create and maintain, and generally more opportunities to mess things up. 09:34:44 would it need to be duplicated? 09:34:55 or would they be two different webs, linked? 09:35:00 JimJibber has joined #swig 09:35:53 my guess would be a lot of duplication. 09:36:37 providing an RDF vCard woudln't get you out of putting the data in a table on your web site, until you are confident everyone can use the RDF version. 09:37:06 tht's close to true, yes 09:37:07 look at how much duplication there is in weblogs that syndicate their content. 09:37:20 but there *are* some problems, tough, which is what worries me 09:38:10 i reckon people think that doing that, is "for free" (meaning, "it is an extra thing, it can cause no harm, if you do not want it, simply ignore it") 09:38:19 but that is not quite true, is not free 09:38:29 it adds complexity to processors 09:38:55 which means more code, more difficult to make a browser, more room for bugs, etc etc 09:43:51 nemequ, would you care for a fast, small example, to see what you think? 09:46:04 drrho has quit 09:47:57 kpreid has quit 09:48:34 * sbp reads Reinventing HTML, which worked its way to Slashdot 09:48:45 Wikier has joined #swig 09:48:59 sbp, slashdot already? 09:49:07 WikierOFF has joined #swig 09:49:10 WikierOFF has quit 09:49:48 yeah 09:50:04 not too surprising with stuff like "The perceived accountability of the HTML group has been an issue." 09:50:34 true 09:50:57 incremental improvements to HTML... wonderful 09:51:48 JibberJim has quit 09:52:48 http://dig.csail.mit.edu/breadcrumbs/node/166 09:52:50 B: http://dig.csail.mit.edu/breadcrumbs/node/166 from sbp 09:52:57 B:|Reinventing HTML 09:52:58 Titled item B. 09:53:41 B: In Slashdot: http://developers.slashdot.org/developers/06/10/28/131246.shtml 09:53:43 Added comment B1. 09:54:26 JimJibber is now known as JibberJim 09:54:41 B:TimBL describes a new, evolutionary future of HTML, with a hand up admission of issues over the accountability of the present HTML WG, and a roadmap describing two new groups: incrementally evolving HTML, and continuing work on XHTML 2.0. Damn good stuff. 09:54:42 Added comment B2. 09:55:28 B:Implies but doesn't specifically mention what will happen to the current HTML WG, nor what the balance of redistribution of its resources would be between the two new HTML and XHTML 2.0 working groups. 09:55:29 Added comment B3. 09:57:38 B:Lots of possible issues and so forth. Some that spring to mind: does evolutionary development mean a Doctype-a-Day? Will this mean, eventually, new improved facilities for RDF/Metadata in HTML? Will the enormous Errata for HTML 4.01 be folded in first? Will this end up in HTML and XHTML actually forking different ways, with two separate groups working on them? 09:57:39 Added comment B4. 10:00:43 nemequ has quit 10:01:02 B:The lack of focus on the validator is another crack that may develop into an issue: "The validator and other subjects cropped up too, but let's focus on HTML now." Given the popularity of test-based development outside and inside the W3C (*cough* SWAP tools *cough*), I'm not sure that a philosophy to the contrary in this context is tenable. 10:01:03 Added comment B5. 10:02:32 Wikier has quit 10:02:46 grove has joined #swig 10:02:57 B:On the other hand, the article later on does mention the future of the validator and is very positive about that, mentioning its current value, how to augment that, and also noting that there's some new hardware goodies available for it. Again, though, it may be that the coordination between specification and tests will need strengthening. 10:02:58 Added comment B6. 10:06:51 * sbp wonders if he should've commented here or on http://www.w3.org/QA/2006/10/reinventing_html_discuss.html - probably doesn't matter too much 10:07:54 sbp, i did worse: i commented here, in the channel 10:09:08 yeah. that is worse since these logs are ungooglable... :-) 10:09:15 I've thought about running my own logger in here 10:11:14 * sbp reviews the logs 10:14:12 kpreid has joined #swig 10:14:40 weird... the style seems to have disappeared on the TAG Issues List 10:15:28 stylesheets do dissapear sometimes from w3c site, don't they? 10:18:51 do they? kinda weird it hasn't been noticed 10:19:06 shouldn't whip them out without checking what pages use them 10:19:21 i find sometimes some pages with messages like "no xslt for this page", i mean 10:19:33 ah 10:19:36 I meant CSS 10:19:59 not sure here, but i think many pages there are xml+xslt 10:20:05 or xsl 10:20:36 ah, siteData-36 is still open... I have a bit of an interest in that lately 10:21:15 Biblio: don't think so, and not this one at least (http://www.w3.org/2001/tag/issues.html) 10:21:44 site data? i'm *very* interested in that too 10:22:01 but last i saw was a link to a log which can not be accessed 10:22:19 and former to that, some log with timbl asking if nobody was interested 10:22:36 (like if he was, but nobody else was) 10:23:40 is that the issue about what to do, or where to put metadata, or info about a whole site? or even how to define such thing as a site? 10:24:22 yeah, pretty much 10:24:43 it amazes me such thing does not exist, as of 2006 10:24:47 my renewed interest in it stems from the LinkedData Design Issue 10:24:53 it should exist since ages ago 10:25:04 it reminded me of the old Metadata: HTTP header proposals 10:25:07 baku has joined #swig 10:25:16 I don't think it should be a particularly complex thing, really 10:25:28 agreed 10:25:43 I mean, such a header solves siteData-36. game over 10:25:54 as to its specification... I've seen a few different ideas 10:26:06 I think it's important to be ecumenical, so I don't think a too-RDF-centric approach would be a good idea 10:26:22 I was thinking lately that it should just be a (role, object) thing 10:26:37 rdf, as opposed to what? 10:27:08 so, Metadata: , where is a URI resolved against some W3C base URI indicating the role, and is the URI of the actual metadata. perhaps there could be a Content-Type hint, optionally, at the end... 10:27:24 (so it's basically for HTTP) 10:27:48 RDF as opposed to robots.txt format, etc. 10:28:02 "the metadata for this site is all HERE" kind of thing, right? 10:28:10 I mean, it's been suggested that we could just have N-Triples in there or something 10:28:26 oh, yeah, there's that aspect too 10:28:59 * sbp shudders at the mere thought 10:30:56 I guess the problem is the practicality. it's a lot easier to deploy the Magic URI solution, the /robots.txt etc. thing, whereas adding headers to all of your responses... that's quite a bit harder 10:31:25 i'm no expert in the subject, i'm afraid 10:31:49 but "Metadata:" is cool from a Semantic Web point of view too 10:31:49 that said, i questioned myself here today if embedding rdf into html, at all, was a good idea 10:32:32 i think i rather have the for-humans and the for-machines data separated, and just linked 10:33:03 I mean, I was thinking upon reading LinkedData that the authoritativeness of data is rather important... timbl was talking about derferencing URIs as the primary way of finding out about something. you can do that for URIs that will return RDF or HTML (maybe), but not, say, GIF or some other binary format 10:33:18 yeah. I guess some should and some shouldn't 10:33:45 GRDDL is quite neat because you can hide data away in the , or present it however you want in the 10:34:13 but yeah, sometimes it's better to have it completely separate 10:34:45 the Tabulator slides I was reading yesterday mentioned the three following approaches for mining RDF from HTML: 10:34:46 * link rel=meta in HTML 10:34:51 * GRDDL profile 10:34:58 * RDF/a ?? 10:35:07 yes, i saw that too 10:35:23 which startled me a bit because I'd been so locked into being prescriptive about it, I forgot to think about how people are actually doing it 10:35:35 what does the first mean? 10:35:38 I note that RDFinXHTML-35 is still open, incidentally... 10:35:51 the first means that you do something like this: 10:36:07 10:36:41 you can consider that statement as meaning that the current document asserts all the statements in foaf.rdf. that's not formally specified though... it's a kind of ad hoc HTML extension 10:37:08 still, I think it has strength from usage. the RDF Syntax Specification informatively says to do things that way, too 10:37:21 but it means that all for-machines information is somewhere else, right? 10:37:27 right 10:37:33 thank you 10:37:44 note that it can be used in conjunction with GRDDL 10:37:58 indeed, all three of these approaches can be used in conjunction with one another 10:38:07 (RDFa is kinda wacky though) 10:38:38 rel="meta" is the least wacky. it's the ghetto way of doing it. it's simple, it works, it's not clever. it's like /robots.txt 10:39:35 i wrote a lot here today. and i think it *might* interest you 10:41:32 hmm: "The TAG has not resolved this issue since the loop has not been closed with the IETF." - uriMediaType-9 10:42:00 Biblio: a lot in the #swig logs? yeah, I've been looking at it. the whole WYSIWYG vs. Source thing is backgrounded... I'm thinking about it a bit 10:42:12 leobard has joined #swig 10:42:12 =P 10:42:41 I flip wildly back and forth on that issue 10:42:53 I wrote a WYSIWYG editor for HTML just a couple of months ago, for example 10:43:02 did you? really? 10:43:04 but now I'm working on a wiki-like syntax for editing HTML. just can't win 10:43:10 yeah. http://inamidst.com/mimulus/ 10:43:22 Firefox Extension 10:43:48 did you consider the issue of wysiwyg editor, gui included, *with* perfect code generation, and css? 10:44:17 i'm not sure it even can be done today, with current state of html 10:44:41 what do you mean by "perfect code generation"? I thought I knew, but now I'm reconsidering 10:45:00 a perfectly valid output, for starters 10:45:14 clean, simple, valid 10:45:33 (the opposite of word, let's say) 10:45:39 well that's pretty much what I tried to do 10:45:44 Wikier has joined #swig 10:45:49 to achieve it, I just concentrated on a bare subset of HTML facilities 10:46:09 so I just focussed on getting paragraphs, links, lists, headings, and preformatted sections correct 10:46:13 what about css? 10:46:26 would authors need to edit it by hand? 10:46:44 or could they edit content word-processor like, but with css? 10:46:53 well, for everything else it dips back into source edit mode, yes. but! one little trick is that you can convert a single element into source inline, and edit that way. but no, no perfect solution there yet 10:47:23 and, as I say, I've flipped the other way at the moment. working on a wiki-like PoorMan'sHypertext language 10:47:23 no surprise there. i'm not even sure it can be done 10:47:27 aye 10:47:55 but... that's likely why I flip-flop 10:48:02 likely, yes 10:48:07 there's not even no optimal solution, there's barely even a workable one 10:48:07 not trivial at all 10:48:23 at least at the two extremes of WYSIWYG and PoorMan'sHypertext 10:48:35 which rather makes one understand how HTML is so successful as it is... 10:48:38 ...unless we all think of a content authoring language able to be edited this way 10:48:55 oh? 10:49:07 during the last decade of html evolution, i doubt anyone considered even this problem 10:50:21 did anyone in this last decade check that html *was* able to be edited wysiwyg perfectly? 10:50:33 was css developed considering that? 10:52:11 yeah, there's been a constant thread of thought over it. as you said: " timbl's original idea, was wysiwyg". and you already know about Amaya and whatnot... the idea to do things WYSIWYG and to do them consistently has been around for a long time. DanC often laments the lack of decent WYSIWYG editing for HTML 10:52:30 but it hasn't been a huge consideration, probably because source editing is at least possible 10:52:34 even if it's a bit irritating 10:52:34 a tragedy! 10:52:44 two big problems: 10:53:05 1( grandma and grandpa can not edit perfect content unless they learn all the complexity 10:53:18 2) hand editing, you likely get tag soup 10:53:47 I find editing source only drives me over the edge when I'm editing a really deeply nested document. lots of nested lists in a huge document has even driven me to Amaya in the past, back in the day when it was utterly, utterly broken, rather than just a single utterly 10:54:08 well, for 1) you've got Frontpage and what-have-you 10:54:33 and yes... 2) is a problem. I get around that using nXML-mode in emacs and a nice little xhtml.rnc RELAX NG schema that I hacked up 10:54:46 true, but not if you request also "with perfect code generation". if you do, then you do not have even one app that does it 10:54:58 only I've been using HTML 4.01 for one project recently, because I'm sick of the useless and nope. there's no optimal solution 10:56:42 so we get tag soup because people handcodes 10:57:11 not because closing tags is difficult, but because we can not spit perfect code wysiwyg 10:57:31 ergo, xhtml flops 10:58:00 * sbp skims http://www.w3.org/2001/tag/doc/qnameids.html whilst thinking about an old RDF in XML proposal of his that used QNames in element content 10:58:20 Biblio: as I say, it is technically possible to have non-tag-soup when handcoding 10:58:27 if you have a tool like nxml-mode to hold your hand 10:58:34 very very difficult, imho 10:58:38 I've also been thinking about use of the W3C Validator in uploading 10:58:48 not impossible, perhaps, but...very difficult, imho 10:58:49 "don't let me upload this file unless the Validator says it's valid!" 10:59:22 very difficult for the non-emacs using populace, true 10:59:47 we should always think "grandma and grandpa" 11:01:38 heh, good common sense advice: [[[ 11:01:40 Specifications that use QNames to represent {URI, local-name} pairs MUST describe the algorithm that is used to map between them. 11:01:40 We observe also that there is an overlap in the lexical space of QNames and URIs. 11:01:40 Specifications that use QNames to represent {URI, local-name} pairs SHOULD NOT allow both forms in attribute values or element content where they would be indistinguishable. 11:01:48 ]]] - qnameids.html 11:02:12 (bit badly worded. had to read it a few times before I realised what it meant by "forms") 11:02:34 "It is simply not practical to suggest that this usage should be forbidden on architectural grounds." - yay 11:02:39 )i'm clueless on that, i'm afraid) 11:03:31 it pleases me since when I did this a while ago, I got flamed down because of it 11:03:31 I wonder how having small upgrades to html is going to help, without causing more confusion... 11:03:50 i reckon there is some politics in this, sbp 11:04:01 oh, bblfish, sorry 11:04:12 ah, kind of thinking that too 11:04:33 there are some good ideas regarding validator on the slashdot page 11:04:45 I like the idea of separating the bugs into groups 11:05:42 heh. use of XLink is only a (lowercase) "may". "XLink is not the only linking design that has been proposed for XML, nor is it universally accepted as a good design." - http://www.w3.org/TR/2004/REC-webarch-20041215/#xml-links 11:06:11 * sbp liked XLink at first, but its rampant behaviour-driven-design bugs me now 11:06:49 people are complaining that the missing "alt" tag bug report takes up too much space 11:07:47 wow: http://lists.w3.org/Archives/Public/www-tag/2003Jan/0418 11:07:56 that was back in 2003? I wonder how work on that is going now... 11:08:05 (subset of XML 1.1) 11:10:26 * sbp reaches siteData-36, confirms Biblio's observation that not much material on that is forthcoming at the moment... the only 2006 thing about it is Member Only 11:10:45 unaccesible, yes 11:11:03 and last to one, is timbl asking if anyone was interested, i think 11:11:28 got a ref for that? 11:11:48 mm, no, but i think is the tagmem link 11:12:03 last meeting cited in that issue 11:15:05 hmm. rdfURIMeaning-39 -> http://lists.w3.org/Archives/Public/www-tag/2003Jul/0127 *but* DanC doesn't mention siteData-36 as a related issue. what an odd oversight 11:15:23 perhaps that's a novel mentality 11:19:02 * sbp gets a bit lost in XMLVersioning-41 11:19:25 I hope that the finding for that is somewhat explicative and tutorial-like... 11:20:49 argh, XMLVersioning-41 is a timesink. avert your eyes! avert your eyes! 11:20:56 * sbp moves onwards 11:21:36 libby has joined #swig 11:22:31 (oops: I was speaking about the comments on slashdot further up) 11:23:23 bblfish, you mentioned, we undertood, is all fine 11:23:40 :-) 11:23:44 =P 11:24:15 "If anyone (including Roy) remembers what Roy was particularly concerned at here, please let me know." - http://www.w3.org/2001/tag/doc/URNsAndRegistries-50.xml 11:25:10 ooh, standardizedFieldValues-51 is awesome. DanC++ 11:28:56 there any browsers that support xhtml as compact-syntax relax-ng 11:30:14 xx: hmm? 11:32:15 the schema language is orthogonal to the content of the instance data. the only bearing it has on it at all is when you provide a hint as to the schema being used, as with !DOCTYPE or xsi:schemaLocation. but RELAX NG doesn't specify any such hints 11:32:55 so you can't even know when someone is using RELAX NG to validate a document; that process is transparent, and it can be applied by anyone, not just by the author 11:33:08 so, if I understand your question, the answer is: all browsers 11:33:25 hmm. i doubt that, im sure theyll just display it as plain text? 11:33:38 i just think it would be cool if 'view source' as pretty/readable as markaby 11:34:28 so far all i can find are random parsers out there, nothing like a firefox extension or anyting 11:34:34 hehe: "DC: it seems we have failed to convince at least one person" - http://www.w3.org/2001/tag/2005/04/19-minutes.html#item04 11:34:49 xx: um, I guess I still don't understand your question then. care to explain further? 11:36:15 briansuda has joined #swig 11:38:05 * sbp wonders where the *latest* CURIE work is 11:38:29 http://www.w3.org/2001/sw/BestPractices/HTML/2005-10-27-CURIE says that itself is, but I'm not sure I trust it 11:42:04 TagSoupIntegration-54 is scary. remember Hixie's Google-report on HTML in the wild? 11:44:10 tag soup /fault tolerance is plain evil 11:44:33 stricness, instead, is the way to go 11:44:45 why? 11:44:54 and then tell the TAG that :-) 11:45:20 this is my third day barely refraining from doing exactly that 11:45:43 oh. hehe. sorry 11:45:57 no probs =P 11:46:16 but they know it, sbp 11:46:29 xhtml was an attempt to go strict 11:46:54 this tag soup reversion now has got to be political 11:48:12 kpreid has quit 11:48:30 I'm not so sure 11:48:37 there's never any active harm in being a descriptivist 11:48:57 fault tolerance was perhaps understandable when the web started 11:49:12 at the very least, it might transpire that certain things about HTML in the wild can be fixed in ways that hadn't been seen before this issue was raised 11:49:18 cause nobody knew much back them, all was new, cognitive load for users was a lot 11:49:35 it seems pretty important now, still 11:49:53 not if you can edit wysiwyg perfectly 11:50:00 well, a billion documents sadly can't be wrong 11:50:07 that's why HTML 4.01 is so strong 11:50:21 old docs need to be converted to the strict format 11:50:22 there will always, always be tools to read it now. for a hundred years, maybe 11:50:38 but all new content could be done with wysiwyg perfect editors 11:50:46 could be. what's the motivation? 11:50:58 I mean at an individual level 11:51:34 clean, unambiguous, dependable, expectable markup, means cleaner, even trivial parsers 11:51:48 that means less code, less bugs 11:52:03 on top, it means better machine processability 11:52:13 evamen has joined #SWIG 11:52:23 that's a not a motivation for individuals. it's not even a motivation for the writers of UAs because, as I just noted, there are more than a billion tagsoup documents out there and they'll be around for a long time 11:52:41 (plus the wysiwyg bit means anyone can write content without even seeing the underlying complexity) 11:53:14 old docs can be tidied up, converted 11:53:18 kpreid has joined #swig 11:53:22 too much work? 11:53:25 but will they be? what's the motivation 11:53:26 exactly 11:53:37 where too much work is > 0 :-) 11:53:48 well, bvefore the web, there was *zero* content. we wrote it asll. we can do it again. but can be converted 11:54:18 but we won't. heh 11:54:56 no need, most tag soup worth it, can be converted 11:55:04 if you wrote bad code, now fix it 11:55:16 but it won't be 11:55:30 but at least let's make sure that all new content is correct 11:55:38 * sbp waves the pragmatism flag 11:56:46 (curiously, all argentinian politicians interested in keeping the cretin statu quo, describe themselves as "pragmatists") (no offense to you, of course) 11:57:24 I mean, the fact that we have TagSoupIntegration-54 and Reinventing HTML within a few days of one another is probably no accident, but I don't think it's all politics either. you have to look at and accept the way that things work before you can actually go about fixing them 11:57:33 otherwise you're only fixing them by luck, if at all 11:58:11 we do know that correct code is better than incorrect code 11:58:26 I don't *disagree* with you that content should be strict. Karl Dubost (W3C conformance manager, working on QA) has always had "Be strict to be cool." in his .sig, and I've always loved that 11:58:41 great motto, exactly 12:00:19 nor does the W3C disagree. it's absolutely startling the extent to which that policy is ingrained within its very fabric. that's why the recent development are so awesome: because they're a natural extension of that policy. you can't hammer away at that without going a bit further; you can't prescribe and prescribe and be ignored for long without realising that something isn't working 12:01:28 right. that's why i asked w3c if they would make a browser, for instance 12:01:40 the Reinventing HTML thing is entirely compatible with the Be strict to be cool. way of doing things; it will necessarily be so, because it's born at the W3C. the new Evolved HTML will have DTDs more than likely. there'll be conformance sections in the specification. the W3C Validator will chide you if you don't follow it 12:01:51 they have. Amaya. it sucks 12:02:10 it takes a lot of people and expertise to make a really, really good browser... 12:02:10 maya should be better 12:02:37 the Amaya folk are experts, but Amaya is still floundering pretty much. for a while it was going to be abandoned due to resource problems... I think they're past that now, at least 12:02:54 it is a big task, yes. but has happened before. 12:03:10 and amaya, never got to be a serious, for the masses option 12:03:31 nope. whereas Firefox did. that only had a handful of developers back when it was Phoenix 12:04:10 now that it has several, Firefox 2.0 kinda sucks. http://it.slashdot.org/it/06/10/28/2115202.shtml - Nine Reasons to Skip Firefox 2.0 12:04:28 firefox...don't get me started on them 12:04:34 heh, heh 12:06:44 (just to back up my assertion that Amaya sucks: http://inamidst.com/notes/amaya) 12:07:01 (something like a third of those disappeared in the latest release, and a few more got added...) 12:07:10 (I don't use Amaya now) 12:08:08 heh heh, i see you tried 12:08:52 still, bugs apart, is one of the two best choices currently existing 12:08:56 the other being nvu 12:09:06 coughmimuluscough 12:09:11 but yeah 12:09:12 =P 12:12:25 ironically, a friend is asking about the Poor Man's Hypertext I'm working on 12:12:29 so I'm going to hack on that for a bit... 12:12:44 =P 12:22:52 evamen has left #SWIG 12:25:22 bblfish_ has joined #swig 12:28:46 kpreid___ has joined #swig 12:28:48 idealm_ has quit 12:29:33 danbri_ has joined #swig 12:29:39 danbri_ is now known as danbri 12:31:11 kpreid has quit 12:42:08 bblfish has quit 12:43:16 zwnj has joined #swig 12:47:35 I think I understand: the idea is to standardise how browsers interpret tagsoup 12:51:07 It should work: pages written in tagsoup will always be slower to load, as they get transformed to a correct xml dom, which will make clean xhtml designed pages an advantage 12:54:20 that's one of the possibilities 12:54:26 and that already happens, just not standardisedly 12:55:22 well, wouldn't it mean that tagsoup-cleaning would have to go to a proper XML DOM+CSS instead of to whatever internal structure a particular browser uses? 13:00:10 bengee has joined #swig 13:00:14 what do you think all DOM implementing browsers do now? 13:01:01 if you have to implement it, you may as well use it 13:04:57 briansuda has quit 13:23:02 Wikier is now known as WikierOFF 13:29:05 exactly. I would love that to be standardised. Then people could write standard tag soup :-) 13:29:52 very clever really 13:32:39 Mikelodeon has joined #swig 13:32:51 Hello everybody 13:33:37 Mikelodeon has quit 13:46:35 idealm has joined #swig 13:55:15 leobard has quit 14:03:06 kpreid___ has quit 14:04:43 My interpretation: http://blogs.sun.com/bblfish/entry/standardising_tagsoup 14:05:16 ndw has joined #swig 14:08:12 kpreid has joined #swig 14:10:23 peepo has joined #swig 14:10:37 peepo has quit 14:15:47 bblfish_: note that what you describe there is only *one* of the topics in the issue statement for TagSoupIntegration-54. I rather like the second one, which touches on HTML revision a bit more: whether the mixed language phenomenon that XHTML allows can be retroactively applied to HTML somehow 14:17:11 ah. I need to read it again. 14:17:52 * sbp chuckles at http://lists.w3.org/Archives/Public/www-tag/2006Oct/0070 14:18:47 why the chuckle (sorry I am not so familiar with what is going on here) 14:18:50 ? 14:19:03 oh, Norm summarises *wonderfully* the point of view that I was trying to express to Biblio earlier about the benefits of XHTML having gone somewhat overlooked for practical reasons: http://lists.w3.org/Archives/Public/www-tag/2006Oct/0074 14:19:24 bblfish_: well, the issue of whether XHTML can be sent as text/html was quite a contentious one 14:19:32 ah. 14:19:52 and still remains so. but the specifications say you can, so great (for me; that's the side of the fence I come down on) 14:19:56 content negotiation should fix it 14:20:12 should fix what? 14:20:26 questions on whether to send XHTML or HTML 14:21:21 what question? if you have XHTML, you SHOULD send it as application/xhtml+xml but MAY send it as text/html if it meets the requirements in Appendix C of the XHTML 1.0 specification. if you have HTML, you MUST send it as text/html 14:21:38 the "question" is already answered by specification. nothing to do with conneg, either 14:24:30 the main consideration really is browser support, of course. plenty of people want to take advantage of XHTML (perhaps by using tools that produce it, or perhaps to use XML tools like GRDDL on it in various ways, or perhaps just to be chic) but also want people to actually be able to read it, and so they disobey the SHOULD and go for the MAY of text/html 14:24:50 that's why it's a MAY :-) 14:24:58 I suppose the point of backporting some of the xhtml to html is part of the same tagsoup enterprise: if you don't do it, people will write tagsoup xhtml and so you may as well get the tagsoup standard right 14:25:04 bblfish_ is now known as bblfish 14:26:41 not sure if it's pointedly part of the same enterprise... yet 14:26:41 I wonder though: now that IE7 is out, (how well does it support the latest standards?) is there not going to be more of a push towards using xhtml? 14:27:07 sbp: but if you do conneg you can give app/xhtml+xml to those that say that accept the new world, and text/html to the loosers. A fancy transformation tool can fix any glitches if needed. 14:27:09 I mean, tagsoup XHTML is going to be around whatever the circumstances, as long as there are broken tools—and as long as the human being is a tool for producing XHTML that's likely to be so :-) 14:27:43 s/push/advantage/ 14:28:23 stain: yeah, though actually I haven't come across many sites doing that. it's sort of a pain, and there's no point really. I mean, there's no point converting to HTML 4.01 if that's what you're implying; but there's also not much point publishing as application/xhtml+xml yet... unless you want to use XSLT, say, but then you're basically betting on people using a fairly modern browser anyway 14:29:31 bblfish: who knows... but Microsoft is a W3C member 14:30:32 (oh, and as far as I know you can use a text/html document in the XSLT document(...) function and XSLT processors will generally do the right thing. that's based on hazy experience; don't hold me to it) 14:30:54 (that is to say, if the text/html document turns out to be well-formed XHTML) 14:31:30 (actually, that's kinda an interesting case) 14:33:04 sbp: in software like TurboGears, XHTML is used in the templates, but when it's presented to the client, it is sent as HTML 4 14:33:12 simon has quit 14:33:22 easiest to note by all tags being uppercase 14:33:28 interesting. why is that so? 14:34:06 I think because they couldn't do the so a lot of this stuff is going to be passé in about a couple of years, the time the new HTML working group is finished with its standard 14:34:59 when IE7 and Firefox 2 are the domintant browsers 14:35:19 bblfish: which stuff? 14:35:37 the new HTML group that is being started 14:36:02 that's why the only real interesting thing is a standardisation of tagsoup 14:36:14 by the time the new HTML WG is finished with its Evolved HTML (if it ever is; I thought evolution just went on and on), I would have thought that just as much bad HTML would be produced. the specifications tend to get ignored. it's browser implementation that drives HTML development 14:36:16 because that will always be needed 14:36:37 yes. But IE7 is out and Firefox 2 is out 14:36:50 they are. what of them? 14:37:01 well they implement the latest standards no? 14:37:14 XHTML I would guess at the very least 14:37:20 hmm. not sure. Firefox 2 gets some CSS wrong, I've seen reported 14:37:37 oh, well, yeah. they're as current as you'll get (apart from Opera) 14:37:49 * sbp wonders what their CSS 3 support is like 14:38:16 now of course, those browsers will always function in backwards compatible mode, so people will be able to ignore the new features 14:38:45 > But isn't tag-soup like HTML 4 evil? 14:38:46 No. HTML 4.01 isn't XML but it's the best option for serving pages at present because all browsers have relatively strong support for it. Serving XHTML as text/html has serious issues and not all browsers (IE <= 7.0) support XHTML as application/xhtml+xml. 14:38:55 the KID faq don't detail what are the "serious issues" 14:39:19 well it is evil, because everyone interprets it differently 14:39:39 but if everyone interpreted it in a similar way, then it would just be a shorthand 14:39:47 and you need to do all these careful measutes to get IE to play along in the correct mode 14:41:43 I think there is also a strong reason to upgrade to newer browsers. A lot of tools and web sites can only work really well with them. Thinking of the more AJAXy sites 14:43:37 zwnj has quit 14:43:50 Also it is easier to produce valid xhtml now 14:44:16 since the libraries come standard with programmer frameworks 14:48:27 hey: firefox 2 does spelling correction. Neat 15:20:27 WikierOFF has quit 15:26:18 libby has quit 15:32:14 ndw has quit 15:46:39 ndw has joined #swig 15:55:05 zwnj has joined #swig 16:10:48 leobard has joined #swig 16:11:31 hi everybody 16:12:50 there is a typo in http://www.w3.org/DesignIssues/HTTP-URI2.html 16:13:05 the title is still the one of the first doc, seems timbl copy/pasted it :-) 16:13:24 and I still don't understand what the TAG has agreed on, 16:13:33 how to identify concepts using URIs 16:17:20 LotR has joined #swig 16:19:43 danbri_ has joined #swig 16:31:05 xx is now known as oo 16:33:53 bengtf has joined #swig 16:36:55 danbri has quit 16:37:53 zwnj has quit 16:38:08 zwnj has joined #swig 16:39:46 oo is now known as _x 16:47:59 baku_ has joined #swig 16:55:50 You mean about tag soup? 16:56:10 baku has quit 16:59:13 The TAg has only decided that the tag soup issue is an important architectural one for the web. 17:01:21 * danja files TagSoupIntegration-54 under "rather you than me" 17:02:18 there was a related problem bugging me about grddl 17:02:42 basically 99% of the microformat docs in the wild lack a profile uri 17:03:03 so can't be grddl'd (strictly speaking) 17:03:58 a workaround I'll probably use if I ever get around to getting code running on this stuff is 17:05:33 to flag the graphs obtained from "weak" HTML as non-authoritative somehow - basically make a note to the effect that it's been scraped, and may not be as the original creator intended 17:13:12 csarven has joined #swig 17:16:31 bit of an easier problem than deterministically interpreting arbitrary text as the creator intended 17:16:36 timbl: the property of the document at the url http://www.w3.org/DesignIssues/HTTP-URI2.html is wrong, because you probably copy/pasted it 17:17:00 <leobard> timbl: about the TAG - I meant the Technical Architecture Group's solution to the identification issues, last year 17:17:26 <leobard> basically, what came out was that URIs of documents should be retrievable (200) 17:17:51 <leobard> I am writing up something for the Nepomuk project on how to identify concepts in RDF. 17:18:10 <timbl> By the way, if looking for timbl typos is like shooting fish in a barrel. :-/ 17:18:31 <danja> hadn't noticed ;-) 17:19:19 <leobard> yes, its ok. Just give us your password on the W3C box, we will fix them in a community effort 17:19:23 <dajobe> heh 17:19:33 <timbl> I just spellchecked HTTP-URI and HTTP-URI2, found lots of things. Apologies to readers. 17:19:45 <dajobe> that's certainly a novel approach to web architecture - put it on a wiki 17:20:02 <timbl> leobard, Ah , oK, abot HTTP Range 14 17:20:28 <timbl> esw is not a web architecture wiki? 17:21:03 <timbl> leobard: The essential of what the TAG decided, as a bit a compromise in a way, but one which works well I think, was this. 17:21:24 <dajobe> true, I guess things like uri schemes only live there now... 17:21:29 <timbl> If you see an HTTP URI you can't assume that it identifies a document (strictly, information resource) 17:21:44 <timbl> so that people can use them (without hashes) to identify cars. 17:22:21 <timbl> However, if your server serves up 200 OK to an HTTP request, thna means "The URI does identify a document, and here is a digital representaton of its content". 17:22:37 <timbl> So if you use HTTP URIs for other things, then you must return a 303 See Also, 17:23:15 <timbl> which means "I can't give you the bits of a rereentatoin of this, but here is a DIFFERENT resource whose contents may help you in your quest" 17:23:41 <leobard> just a sec, I make an example 17:23:48 <timbl> ok 17:24:41 <leobard> so this states that timbl has written the note: 17:24:45 <leobard> <http://www.w3.org/DesignIssues/HTTP-URI2.html> dc:author <http://www.w3.org/People/Berners-Lee/card#i>. 17:25:08 <timbl> yes 17:25:14 <leobard> note that I already fucked it up: the first uri returns 200, the second I don't know 17:25:34 <leobard> hm, or perhaps put another way: apache has written the document to me 17:25:45 <timbl> The second URI has a hash, which is another way of making a URI for a something which doesn't ahve t be a document. 17:26:03 <leobard> but the document written by you is identified with some hash: http://www.w3.org/DesignIssues/HTTP-URI2.html# 17:26:36 <timbl> The hash architecture is that a hash links the URI of a document and a local identifier within that document to make a global URI for the thing identified by the local identifier within the document. 17:27:29 <leobard> yes, thats clear. 17:27:45 <leobard> I have some other problem, well, I will continue writing the exact problem up, with concise examples .. this will take a while. 17:27:54 <timbl> The web arhitecture has always been that to redererence a URI with a hash, you deref everything up to the hash (the document URI) and then learn from the document about the thing identifierd by the loacl ID 17:28:36 <timbl> ?? """leobard: but the document written by you is identified with some hash: http://www.w3.org/DesignIssues/HTTP-URI2.html#""" no, the document is not identified by that, but by the URI without a hash. 17:29:07 <leobard> ok, point 17:31:21 <Wikier> Wikier has joined #swig 17:31:39 <leobard> I am preparing something together with Max Völkel, the guy that is close to the Semantic Mediawiki people. I hope to get a longer document with examples and a thouroug literature review together. 17:31:54 <leobard> Will come back with end later this week, also I am at ISWC2006 17:32:22 <jargonjustin> jargonjustin has joined #swig 17:32:53 <timbl> Semantic Mediawiki does it right, I think , using HTTP 303s. 17:33:42 <timbl> Examples of URIs which do a 303 redirect (try them in the tabulator) include 17:33:45 <timbl> ... http://id.ecs.soton.ac.uk/person/1650 17:35:42 <timbl> ... http://dig.csail.mit.edu/People/RRS 17:35:55 <leobard> ah, ok. Now I see. the id server redirects to a http rep 17:36:02 <leobard> I meant HTML rep 17:36:12 <timbl> Ah, no, semantic mediawiki .. at least Ontoworld does a 301 17:36:23 <leobard> and this contains a written note: 17:36:23 <leobard> Person URI: http://id.ecs.soton.ac.uk/person/1650 17:36:23 <leobard> Role URI: http://id.ecs.soton.ac.uk/role/1650 17:36:23 <leobard> Person RDF: http://rdf.ecs.soton.ac.uk/person/1650 17:36:23 <leobard> Role RDF: http://rdf.ecs.soton.ac.uk/role/1650 17:36:50 <leobard> which is all written here: http://www.ecs.soton.ac.uk/people/wh 17:37:09 <timbl> * timbl notes <http://wiki.ontoworld.org/index.php/_Martin_Hepp> Redirected: 301 to <http://ontoworld.org/wiki/Martin_Hepp> 17:37:17 <leobard> because I wondered how we can make a 'social' protocol that will allow us to find the correct URI identifying the concept of a person from looking at the website of the person 17:37:22 <timbl> Ah, leobard, that is because you didn't ask fro RDF. 17:37:44 <leobard> yes, I just entered it into a browser. 17:37:50 <timbl> If you do an Accept: application/rdf+xml then it will take you to RDF 17:38:12 <leobard> My fear was, that once we start using these 303 URIs inside RDF, stating things like "x hasAuthor y", they won't be "clickable" anymore 17:38:17 <leobard> but through the redirect, it works 17:38:32 <timbl> Yes. 17:38:40 <leobard> clever trick you came up with, and thanks for explaining, I will include it in my doc. 17:39:02 <timbl> The TAG came up withe the clever trick after many many months of deliberation!! 17:39:26 <timbl> and discussion in www-tag and elsewhere in great depth. 17:40:05 <timbl> leobard, please do include the correct recipes (303 or hash) in your doc, as the words needs to be spread around, many people haven't got it yet. 17:42:05 <leobard> timbl: do you have any document on that? I could reproduce it based on the example from ECS but any TAG note would help 17:42:09 <leobard> logger pointer 17:42:28 <timbl> logger, pointer? 17:42:28 <timbl> See http://chatlogs.planetrdf.com/swig/2006-10-29#T17-42-28 17:43:16 <leobard> thx for the syntax :-) 17:43:31 <leobard> is there a writeup of the approach with 303 and the content negotiation? 17:44:41 <timbl> There is the email http://lists.w3.org/Archives/Public/www-tag/2005Jun/0039.html 17:45:19 <timbl> There is a document by the Sem Web Best Practcies group which explsin how to set up Apache to do a 303 17:45:39 <timbl> Note when writing RDF or N3 by hand, the hash method works much more simply 17:47:15 <leobard> by hand, you mean to just use a plain '#' to reference the current doc, I assuem 17:48:09 <timbl> The # i sonly used to refer to something from outside the document normally 17:48:41 <leobard> oh, I have to stop now, I am too fat sitting around all day, have to go running with my wife :-) 17:48:44 <timbl> In XML you say rdf:id="tbl" or you can say rdf:about="#tbl" 17:48:57 <timbl> In N3 you just use the defualt namespace. 17:49:03 <timbl> Have a good run 17:49:08 <leobard> I will continue tomorrow, during worktime, finally something useful to do at the office 17:49:18 <timbl> * timbl forgets where in the world leobard is based 17:49:26 <leobard> thanks for all the useful pointers, they help me a lot 17:49:28 <leobard> sec. 17:49:32 <timbl> welcome 17:50:06 <leobard> http://beta.plazes.com/plaze/1d28f6bf2bb33ddf30a049aafe2f097d 17:50:09 <dc_swig> C: http://beta.plazes.com/plaze/1d28f6bf2bb33ddf30a049aafe2f097d from leobard 17:50:19 <leobard> C:|leobards whereabouts 17:50:20 <dc_swig> Titled item C. 17:51:58 <timbl> Well, around BED we have windy running conditions. leaves ripped off the trees and swirling all over in drifts. 17:52:33 <leobard> * leobard will drive to fitness center and watch TV while running or listening to iPod, which is chicken but ok 17:52:42 <leobard> and its 19:00 and dark here 17:53:56 <timbl> enjoy anyway 17:54:15 <leobard> thx! cya 17:55:03 <zwnj> zwnj has quit 17:55:50 <leobard> leobard has quit 18:00:17 <bengtf> bengtf has quit 18:40:47 <zwnj> zwnj has joined #swig 18:50:16 <nemequ> nemequ has joined #swig 18:51:24 <nemequ> nemequ has left #swig 18:51:39 <nemequ> nemequ has joined #swig 18:58:27 <bblfish> bblfish has quit 19:03:16 <JustRay> JustRay has joined #swig 19:05:43 <Nurf> Nurf has joined #swig 19:06:00 <Nurf> Nurf has quit 19:20:51 <bengee> bengee has quit 19:29:27 <csarven> csarven has quit 19:49:31 <nephrael> nephrael has joined #swig 19:53:23 <nephrael> nephrael has quit 20:01:13 <csarven> csarven has joined #swig 20:13:58 <bkdelong> bkdelong has joined #swig 20:15:34 <nemequ> nemequ has quit 20:15:37 <bkdelong> bkdelong has quit 20:21:17 <danbri_> danbri_ has quit 20:23:16 <libby> libby has joined #swig 20:23:25 <baku_> baku_ has quit 20:24:23 <kpreid> kpreid has quit 20:28:36 <kpreid> kpreid has joined #swig 20:31:37 <zwnj> zwnj has quit 20:37:23 <zwnj> zwnj has joined #swig 20:54:07 <kpreid> kpreid has quit 20:56:47 <bblfish> bblfish has joined #swig 21:04:24 <kpreid> kpreid has joined #swig 21:05:22 <nemequ> nemequ has joined #swig 21:11:05 <perigrin> perigrin has joined #swig 21:31:08 <Mikelodeon> Mikelodeon has joined #swig 21:56:34 <kpreid> kpreid has quit 21:57:44 <kpreid> kpreid has joined #swig 22:03:48 <Mikelodeon> Mikelodeon has quit 22:06:46 <Mikelodeon> Mikelodeon has joined #swig 22:08:44 <zwnj> zwnj has quit 22:09:51 <zwnj> zwnj has joined #swig 22:10:05 <kpreid> kpreid has quit 22:12:42 <grove> grove has quit 22:13:42 <kpreid> kpreid has joined #swig 22:14:12 <Mikelodeon> Mikelodeon has quit 22:15:19 <perigrin> perigrin has quit 22:15:19 <bblfish> bblfish has quit 22:15:19 <ndw> ndw has quit 22:15:19 <idealm> idealm has quit 22:15:19 <danja> danja has quit 22:15:19 <eikeon> eikeon has quit 22:15:19 <timbl> timbl has quit 22:15:19 <CaptSolo> CaptSolo has quit 22:15:19 <boneill> boneill has quit 22:15:19 <bogonflux> bogonflux has quit 22:15:19 <evanpro> evanpro has quit 22:15:19 <gromgull> gromgull has quit 22:15:19 <dc_swig> dc_swig has quit 22:15:19 <phenny> phenny has quit 22:15:48 <perigrin> perigrin has joined #swig 22:15:48 <bblfish> bblfish has joined #swig 22:15:48 <ndw> ndw has joined #swig 22:15:48 <idealm> idealm has joined #swig 22:15:48 <danja> danja has joined #swig 22:15:48 <eikeon> eikeon has joined #swig 22:15:48 <timbl> timbl has joined #swig 22:15:48 <CaptSolo> CaptSolo has joined #swig 22:15:48 <gromgull> gromgull has joined #swig 22:15:48 <boneill> boneill has joined #swig 22:15:48 <bogonflux> bogonflux has joined #swig 22:15:48 <evanpro> evanpro has joined #swig 22:15:48 <phenny> phenny has joined #swig 22:15:48 <dc_swig> dc_swig has joined #swig 22:17:52 <Wikier> Wikier has quit 22:20:49 <JustRay> JustRay has quit 23:06:33 <stelt> stelt has joined #swig 23:13:05 <zwnj> zwnj has quit 23:13:14 <zwn1> zwn1 has joined #swig 23:28:49 <kpreid> kpreid has quit 23:34:00 <kpreid> kpreid has joined #swig 23:38:40 <danja> danja has quit 23:43:09 <zwn1> zwn1 has quit 23:43:36 <bblfish> bblfish has quit 23:56:43 <karlUshi> karlUshi has joined #swig